CN107040209A - Circuit arrangement, oscillator, electronic equipment and moving body - Google Patents
Circuit arrangement, oscillator, electronic equipment and moving body Download PDFInfo
- Publication number
- CN107040209A CN107040209A CN201611167368.2A CN201611167368A CN107040209A CN 107040209 A CN107040209 A CN 107040209A CN 201611167368 A CN201611167368 A CN 201611167368A CN 107040209 A CN107040209 A CN 107040209A
- Authority
- CN
- China
- Prior art keywords
- oscillator
- control data
- circuit arrangement
- frequency control
- processing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012545 processing Methods 0.000 claims abstract description 271
- 238000001914 filtration Methods 0.000 claims abstract description 152
- 238000012937 correction Methods 0.000 claims abstract description 125
- 230000010355 oscillation Effects 0.000 claims abstract description 67
- 238000000034 method Methods 0.000 claims description 67
- 230000008569 process Effects 0.000 claims description 54
- 238000003860 storage Methods 0.000 claims description 35
- 238000004364 calculation method Methods 0.000 claims description 29
- 230000000052 comparative effect Effects 0.000 claims description 10
- 230000008034 disappearance Effects 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 2
- 230000032683 aging Effects 0.000 description 80
- 230000008859 change Effects 0.000 description 61
- 239000000203 mixture Substances 0.000 description 54
- 238000001514 detection method Methods 0.000 description 50
- 230000007613 environmental effect Effects 0.000 description 30
- 239000003990 capacitor Substances 0.000 description 21
- 230000009471 action Effects 0.000 description 18
- 230000006854 communication Effects 0.000 description 16
- 238000004891 communication Methods 0.000 description 13
- 230000015654 memory Effects 0.000 description 12
- 238000005259 measurement Methods 0.000 description 11
- 238000012986 modification Methods 0.000 description 10
- 230000004048 modification Effects 0.000 description 10
- 238000006243 chemical reaction Methods 0.000 description 8
- 238000010276 construction Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 239000011159 matrix material Substances 0.000 description 8
- 230000001360 synchronised effect Effects 0.000 description 8
- 230000005611 electricity Effects 0.000 description 6
- 238000007689 inspection Methods 0.000 description 6
- 238000004519 manufacturing process Methods 0.000 description 6
- 239000000758 substrate Substances 0.000 description 6
- 239000013078 crystal Substances 0.000 description 5
- 238000011084 recovery Methods 0.000 description 5
- 238000009499 grossing Methods 0.000 description 4
- 238000010438 heat treatment Methods 0.000 description 4
- 239000010703 silicon Substances 0.000 description 4
- 229910052710 silicon Inorganic materials 0.000 description 4
- 230000015572 biosynthetic process Effects 0.000 description 3
- 238000005538 encapsulation Methods 0.000 description 3
- 239000010453 quartz Substances 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 2
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002431 foraging effect Effects 0.000 description 2
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 description 2
- 230000002045 lasting effect Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000007480 spreading Effects 0.000 description 2
- 238000003892 spreading Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- WSMQKESQZFQMFW-UHFFFAOYSA-N 5-methyl-pyrazole-3-carboxylic acid Chemical compound CC1=CC(C(O)=O)=NN1 WSMQKESQZFQMFW-UHFFFAOYSA-N 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 101000802640 Homo sapiens Lactosylceramide 4-alpha-galactosyltransferase Proteins 0.000 description 1
- 102100035838 Lactosylceramide 4-alpha-galactosyltransferase Human genes 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000005452 bending Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000007767 bonding agent Substances 0.000 description 1
- 239000000919 ceramic Substances 0.000 description 1
- 238000013016 damping Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 239000002305 electric material Substances 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000009931 harmful effect Effects 0.000 description 1
- 229910052738 indium Inorganic materials 0.000 description 1
- HFGPZNIAWCZYJU-UHFFFAOYSA-N lead zirconate titanate Chemical compound [O-2].[O-2].[O-2].[O-2].[O-2].[Ti+4].[Zr+4].[Pb+2] HFGPZNIAWCZYJU-UHFFFAOYSA-N 0.000 description 1
- 229910052451 lead zirconate titanate Inorganic materials 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- GQYHUHYESMUTHG-UHFFFAOYSA-N lithium niobate Chemical compound [Li+].[O-][Nb](=O)=O GQYHUHYESMUTHG-UHFFFAOYSA-N 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 230000000116 mitigating effect Effects 0.000 description 1
- 230000003534 oscillatory effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 125000006850 spacer group Chemical group 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000010897 surface acoustic wave method Methods 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03B—GENERATION OF OSCILLATIONS, DIRECTLY OR BY FREQUENCY-CHANGING, BY CIRCUITS EMPLOYING ACTIVE ELEMENTS WHICH OPERATE IN A NON-SWITCHING MANNER; GENERATION OF NOISE BY SUCH CIRCUITS
- H03B5/00—Generation of oscillations using amplifier with regenerative feedback from output to input
- H03B5/02—Details
- H03B5/04—Modifications of generator to compensate for variations in physical values, e.g. power supply, load, temperature
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03B—GENERATION OF OSCILLATIONS, DIRECTLY OR BY FREQUENCY-CHANGING, BY CIRCUITS EMPLOYING ACTIVE ELEMENTS WHICH OPERATE IN A NON-SWITCHING MANNER; GENERATION OF NOISE BY SUCH CIRCUITS
- H03B5/00—Generation of oscillations using amplifier with regenerative feedback from output to input
- H03B5/30—Generation of oscillations using amplifier with regenerative feedback from output to input with frequency-determining element being electromechanical resonator
- H03B5/32—Generation of oscillations using amplifier with regenerative feedback from output to input with frequency-determining element being electromechanical resonator being a piezoelectric resonator
- H03B5/36—Generation of oscillations using amplifier with regenerative feedback from output to input with frequency-determining element being electromechanical resonator being a piezoelectric resonator active element in amplifier being semiconductor device
Landscapes
- Oscillators With Electromechanical Resonators (AREA)
- Stabilization Of Oscillater, Synchronisation, Frequency Synthesizers (AREA)
Abstract
Circuit arrangement, oscillator, electronic equipment and moving body.Circuit arrangement is included:Processing unit, its frequency control data to input carries out signal transacting, output frequency control data;And oscillator signal generative circuit, it uses oscillator, generates the oscillator signal of the frequency of oscillation set by frequency control data.Processing unit carries out passage time step k 1 posterior estimate x in the renewal processing for the priori estimates that Kalman filtering is handled^(k 1) and corrected value D's (k 1) is added processing to obtain time step k priori estimates x^‑(k) processing, and the result handled according to Kalman filtering, carry out the age correction of frequency control data.
Description
Technical field
The present invention relates to circuit arrangement, oscillator, electronic equipment and moving body.
Background technology
All the time, it is known to OCXO (oven controlled crystal oscillator:Constant temperature crystal vibrates
Device), TCXO (temperature compensated crystal oscillator:Temperature compensating crystal oscillator) etc. vibration
Device.Reference signal source in such as OCXO is as base station, network router, measuring apparatus and used.
In the oscillators such as such OCXO, TCXO, higher frequency stability is expected.But, there is problems with:
There is changing with time for referred to as aging in the frequency of oscillation of oscillator, frequency of oscillation is passed through and changed with the time.For example, making
In the case of suppressing to receive the reference signals such as gps signal, turning into so-called holding pattern (hold-over) state
, there is the technology disclosed in Japanese Unexamined Patent Publication 2015-82815 publications in the prior art of the variation of frequency of oscillation.In the existing skill
In art, storage part and elapsed time measurement portion, corrected value and process of the storage part to the control voltage of frequency of oscillation are set
The correspondence relationship information (aging characteristics data) of time is stored.Moreover, in the case where detecting holding pattern, according to
The corrected value and the correspondence relationship information in elapsed time that are stored in storage part and the elapsed time measured by elapsed time measurement portion
To perform age correction.
But, in the prior art, there are the following problems:Believe as storage corrected value and the corresponding relation in elapsed time
The storage part of breath can cause the large-scale of circuit arrangement it is necessary to have the storage part of big memory capacity.For example, for reality
Show the age correction of higher precision, it is necessary to which the correspondence relationship information of bigger data volume is stored in into storage part, cause circuit to fill
Put large-scale.
The content of the invention
According to the present invention several modes, it is possible to provide it is a kind of can be with the aging of more small-scale circuit realiration higher precision
Circuit arrangement, oscillator, electronic equipment and moving body of correction etc..
The mode of the present invention is related to circuit arrangement, and the circuit arrangement is included:Processing unit, it enters to frequency control data
Row signal transacting;And oscillator signal generative circuit, it uses oscillator and the frequency control data from the processing unit,
The oscillator signal of the frequency of oscillation set by the frequency control data is generated, the processing unit is handled in Kalman filtering
In the renewal processing of priori estimates, posterior estimate at the time of pass through last time obtains this with the processing that is added of corrected value
The processing of priori estimates at the time of secondary, and the result handled according to the Kalman filtering, carry out the FREQUENCY CONTROL number
According to age correction.
According to the mode of the present invention, signal transacting is carried out to frequency control data by processing unit, using oscillator and
Frequency control data from processing unit, generates the oscillator signal of the frequency of oscillation set by frequency control data.And
In mode of the present invention, in the renewal processing of the priori estimates of Kalman filtering processing, by the time of last time
Posterior estimate is added processing with corrected value, priori estimates at the time of obtaining this.Moreover, according to the Kalman filtering
The result of processing, carries out the age correction of frequency control data.So, for example the situation with being handled using EKF
Compare, alleviate the processing load of processing unit and inhibit increase of circuit scale of circuit arrangement etc..Therefore, it is possible to more
The age correction of small-scale circuit realiration higher precision.
In addition, in the mode of the present invention, Ke Yishi, during the processing unit is handled according to the Kalman filtering
Observation residual error obtains the corrected value.
In such manner, it is possible to which the corrected value updated using the observation residual error reflected in Kalman filtering processing realizes aging school
Just, the age correction of higher precision can be realized.
In addition, in the mode of the present invention, the processing unit carry out be at the time of the last time time step k-1 institute
That states posterior estimate x^ (k-1) and the corrected value D (k-1) is added processing, passes through x^-(k)=x^ (k-1)+D (k-1) is asked
I.e. time step k priori estimates x^ at the time of described in going out this time-(k)。
In such manner, it is possible to pass through x^-(k) the light simple calculations processing of the such processing loads of=x^ (k-1)+D (k-1) is asked
Go out time step k priori estimates x^-(k) small-scaleization of circuit arrangement etc., is realized.
In addition, in the mode of the present invention, Ke Yishi, the processing unit is according to the school of the time step k-1
Observation residual error in D (k-1) and Kalman filtering processing, obtains the corrected value D (k) of the time step k.
In such manner, it is possible to use the observation residual error reflected in Kalman filtering processing and the school updated in each time step
Age correction is realized on the occasion of D (k).
In addition, in the mode of the present invention, Ke Yishi is setting the situation that the observation residual error is E as ek, constant
Under, the processing unit obtains the corrected value D (k) by D (k)=D (k-1)+Eek.
In such manner, it is possible to obtain corrected value D (k), energy by simple calculations processing as D (k)=D (k-1)+Eek
Enough mitigate the processing load of processing unit.
In addition, in the mode of the present invention, Ke Yishi, the circuit arrangement includes the storage part for storing the constant E.
So, the corresponding appropriate constant E such as can use with each product realize corrected value D (k)=D (k-1)+
Eek renewal processing, can realize the age correction of higher precision.
In addition, in the mode of the present invention, Ke Yishi, the processing unit is to based on input signal and reference signal
The frequency control data of phase comparative result carries out the signal transacting, wherein, the input signal is based on the vibration
Signal, the processing unit during detecting before the disappearance of the reference signal or abnormal caused holding pattern,
It is handled as follows:Handled by Kalman filtering, estimation is directed to the FREQUENCY CONTROL number based on the phase comparative result
According to observation true value, in the case where detecting the holding pattern, at the time of preserving and detect the holding pattern
True value at the time of correspondence, by carrying out the frequency after the calculation process based on the true value, generation age correction
Control data.
In such manner, it is possible to be estimated and corresponding with being kept for the detection moment of pattern according to being handled by Kalman filtering
At the time of the true value that preserves realize age correction.Therefore, it is possible to realize the high-precision age correction that can not be realized in the past.
In addition, in the mode of the present invention, Ke Yishi, the processing unit is by carrying out to the true value plus described
The frequency control data after the calculation process of corrected value, generation age correction.
So, carry out adding compensation for example caused by rate of ageing to handling the true value estimated by Kalman filtering
The calculation process of the corrected value of frequency change, hereby it is achieved that age correction.Therefore, it is possible to realize high accuracy with simple processing
Age correction.
In addition, in the mode of the present invention, Ke Yishi, the processing unit carries out adding filtering process to the true value
The calculation process of the corrected value afterwards.
In such manner, it is possible to effectively suppress following situation:It is added and leads with true value due to there will be the corrected value of the fluctuation of variation
Cause the precise decreasing of age correction.
In addition, in the mode of the present invention, the circuit arrangement can also include storage part, storage part storage is described
The system noise constant of the setting of the system noise of Kalman filtering processing and the observation noise of Kalman filtering processing
Setting observation noise constant.
In such manner, it is possible to realize the age correction of the influence for the component deviation for reducing system noise and observation noise.
In addition, in the mode of the present invention, Ke Yishi, the circuit arrangement, which is included, to be used to monitor the priori estimates
With the digital interface portion of observation.
So, the external device (ED) such as check device can monitor priori estimates and observation via digital interface portion.
The observation error equivalent to the difference value of such as observation and priori estimates is set to turn into setting for smaller value thereby, it is possible to realize
Fixed processing etc..
In addition, another mode of the present invention is related to oscillator, the oscillator is included:Described in any one above-mentioned mode
Circuit arrangement;And the oscillator.
In addition, another mode of the present invention is related to electronic equipment, the electronic equipment includes any one above-mentioned mode institute
The circuit arrangement stated.
In addition, another mode of the present invention is related to moving body, the moving body is comprising described in any one above-mentioned mode
Circuit arrangement.
Brief description of the drawings
Fig. 1 is the explanation figure of the component deviation for aging characteristics.
Fig. 2 is the basic configuration example of the circuit arrangement of present embodiment.
Fig. 3 is the explanation figure of the method for present embodiment.
Fig. 4 is the explanation figure of the method for present embodiment.
Fig. 5 is the explanation figure of age correction when being directed to holding pattern.
Fig. 6 is the explanation figure for holding pattern.
Fig. 7 is the explanation figure for holding pattern.
Fig. 8 is the explanation figure for holding pattern.
Fig. 9 is the detailed construction example of the circuit arrangement of present embodiment.
Figure 10 is the explanation figure for the age correction for having used Kalman filtering to handle.
Figure 11 is the explanation figure for the age correction for having used Kalman filtering to handle.
Figure 12 is the detailed construction example of processing unit.
Figure 13 is the explanation figure of temperature-compensating processing.
Figure 14 is the explanation figure of temperature-compensating processing.
Figure 15 is the explanation figure of temperature-compensating processing.
Figure 16 is the action specification figure of processing unit.
Figure 17 is the action specification figure of processing unit.
Figure 18 is the configuration example in age correction portion.
Figure 19 is the model example of Kalman filtering.
Figure 20 is the configuration example in Kalman filtering portion.
Figure 21 is the figure of the example for the prediction frequency departure and practical frequency deviation for showing present embodiment.
Figure 22 is the configuration example of temperature sensor.
Figure 23 is the configuration example of oscillating circuit.
Figure 24 is the explanation figure of modified embodiment of the present embodiment.
Figure 25 is the explanation figure of modified embodiment of the present embodiment.
Figure 26 is the configuration example of oscillator.
Figure 27 is the configuration example of electronic equipment.
Figure 28 is the configuration example of moving body.
Figure 29 is the detailed construction example of oscillator.
Figure 30 is the configuration example of the base station as one of electronic equipment.
Embodiment
Hereinafter, it is described in detail for the preferred embodiment of the present invention.In addition, the present embodiment illustrated below
Improper restriction, all structures illustrated in the present embodiment are not carried out to the present disclosure described in claims
It all must be not the solution of the present invention.
1. frequency of oscillation changes caused by aging
In the oscillators such as OCXO, TCXO, due to being referred to as changing with time for aging, frequency of oscillation changes.Moreover,
In the characteristic that the aging of frequency of oscillation between the individual of oscillator changes, exist by constituting the performance of the part of oscillator, part
With caused by the individual deviation (hereinafter referred to as component deviation) of the installment state of oscillator or the use environment of oscillator etc.
Difference.
Fig. 1 A1~A5 is the measurement result of the aging characteristics of the multiple oscillators identical or different on shipment lot number
One.As shown in Fig. 1 A1~A5, there is the difference along with component deviation in the mode that aging changes.
Be considered as the dust that is produced in hermetic container the reason for the variation of frequency of oscillation caused by from aging to
Oscillator come off and adhere to, the environmental change based on some emergent gas or the bonding agent that uses in an oscillator at any time
Between change.
As the countermeasure for suppressing such variation of frequency of oscillation caused by aging, there is following method:Going out
It is initial aged during implementing to make oscillator operation certain before goods, make shipment again after frequency of oscillation initial shifts.But, for
It is required that the purposes of high frequency stability, it is inadequate only to take such initial aged countermeasure, and expected compensation is caused by aging
Frequency of oscillation variation age correction.
In addition, in the case where oscillator is used as into the reference signal source of base station, there is so-called holding pattern
Problem.For example in a base station, by using PLL circuit by the oscillator signal (output signal) of oscillator and from GPS or network
Reference signal it is synchronous, suppress frequency variation.But, disappear when reference signal of the generation from GPS or network (internet) turns into
During the holding pattern of mistake or exception, it is impossible to obtain for synchronous reference signal.If by taking GPS as an example, due to GPS days
The set location or setting direction of line and fail to receive positioning signal, because interference ripple and failing is accurately received positioning letter
Number or not from the case that positioning sends positioning signal with satellite, produce holding pattern, it is impossible to which execution has used benchmark letter
Number synchronization process.
When producing such holding pattern, the oscillator signal produced by the self-oscillation of oscillator turns into the base of base station
Calibration signal source.It is therefore desirable to following holding mode performance:The moment is being produced to from keeping pattern recovery from the pattern of holding
During holding pattern untill moment (releasing the moment), suppress the variation of the frequency of oscillation caused by the self-oscillation of oscillator.
But, as described above, because there is the degree that can not ignore caused by aging in the frequency of oscillation of oscillator
Change, therefore, because this and presence can not realize the problem of high holding mode performance.The pattern phase is kept such as at 24 hours
In, in the case where defining the frequency departure (Δ f/f) allowed, if there is caused by aging frequency of oscillation it is larger
Change, then can not meet the regulation of the tolerance frequency deviation.
For example as base station and the communication mode of communication terminal, it is proposed that FDD (Frequency Division Duplex:
FDD), TDD (Time Division Duplex:Time division duplex) etc. various modes.Moreover, in TDD modes, it is up
With descending use identical frequency according to time division way transceiving data, when being set with protection between distributing to the time slot of each equipment
Between.Therefore, in order to realize appropriate communication, it is necessary to carry out timing synchronization in each equipment, it is desirable to there is accurately absolute moment
Timing.That is, in order to provide wireless communication system that mobile phone, received terrestrial digital broadcasting etc. communicate in extensive area, it is necessary to
Multiple base stations are set, when between these base stations deviation occurs for timer time, it is impossible to realize appropriate communication.But, in production
In the case of having given birth to the reference signal disappearance from GPS or network or the holding pattern of exception, need do not having in alternator side
Have and timing is carried out to the absolute moment in the state of reference signal, if deviation, communication failure occur for the timer time.Therefore,
For the oscillator used in base station etc., very high frequency stability is also required that during holding pattern.Therefore, for mending
The age correction of the frequency variation caused by aging is repaid, high-precision correction is also required that.
2. the structure of circuit arrangement
Fig. 2 shows the basic circuit structure of the circuit arrangement of present embodiment.As shown in Fig. 2 the circuit of present embodiment
Device includes processing unit 50 and oscillator signal generative circuit 140.In addition, the structure of the circuit arrangement of present embodiment is not limited to figure
2 structure, can implement to omit the various modifications such as a portion structural element or additional other structures key element.
Processing unit 50 carries out various signal transactings.For example frequency control data DFCI (FREQUENCY CONTROL code) is carried out at signal
Reason.Specifically, processing unit 50 (digital signal processing section) carries out such as age correction processing, Kalman filtering processing, and root
According to needing to carry out the signal transacting (Digital Signal Processing) such as temperature-compensating processing.Also, the FREQUENCY CONTROL after output signal processing
Data DFCQ.Processing unit 50 can be included:Kalman filtering portion 54 (Kalman filtering processing circuit or program module) and always
Change correction unit 56 (circuit or program module of age correction processing).The processing unit 50 can be real by ASIC circuits such as gate arrays
It is existing, it can also be realized by processor (DSP, CPU) and the program worked on a processor (program module).
Oscillator XTAL is, for example, that AT cuts type or SC cuts quartz vibrator of type equal thickness scissoring vibration type etc. or bending
The piezoelectric vibrator of oscillatory type etc..As one, oscillator XTAL is disposed in the thermostat of constant temperature groove profile oscillator (OCXO)
Type, but not limited to this can be the TCXO of the type without thermostat oscillator.Oscillator XTAL can also be humorous
Shake device (electromechanical resonator or the resonance circuit of electric).In addition, as oscillator XTAL, SAW (Surface can be used
Acoustic Wave:Surface acoustic wave) resonator, be used as silicon damping son MEMS (Micro Electro Mechanical
Systems:Microelectromechanical systems) oscillator etc. is used as piezoelectric vibrator.As oscillator XTAL baseplate material, usable quartz,
The piezoelectric ceramics such as the piezoelectric single crystals such as lithium tantalate, lithium niobate, lead zirconate titanate equipressure electric material or silicon semiconductor material etc..As shaking
Sub- XTAL motivator, can both use the means based on piezo-electric effect, can also use the electrostatic drive based on Coulomb force.
The generation oscillator signal of oscillator signal generative circuit 140 OSCK.Come from for example, oscillator signal generative circuit 140 is used
The frequency control data DFCQ (frequency control data after signal transacting) and oscillator XTAL of processing unit 50, generation pass through frequency control
The oscillator signal OSCK of the frequency of oscillation of data DFCQ settings processed.As one, oscillator signal generative circuit 140 makes oscillator XTAL
Vibrated according to by the frequency control data DFCQ frequencies of oscillation set, generation oscillator signal OSCK.
In addition, oscillator signal generative circuit 140 can generate oscillator signal OSCK in direct digital synthesiser mode
Circuit.For example can also be using oscillator XTAL (oscillation source of built-in oscillation frequency) oscillator signal as reference signal, with numeral
Mode generates the oscillator signal OSCK by the frequency control data DFCQ frequencies of oscillation set.
Oscillator signal generative circuit 140 can include D/A converter sections 80 and oscillating circuit 150.But, oscillator signal generation electricity
Road 140 is not limited to such structure, can implement omission a portion structural element or additional other structures key element etc. various
Deformation.
D/A converter sections 80 carry out the D/A of the frequency control data DFCQ (output data of processing unit) from processing unit 50
Conversion.It is input into (such as aging school after the signal transacting that the frequency control data DFCQ of D/A converter sections 80 is processing unit 50
Just, after temperature-compensating or the processing of Kalman filtering) frequency control data (FREQUENCY CONTROL code).It is used as D/A converter sections 80
D/A conversion regimes, can for example use resistance serial type (resistance Splittable).But, D/A conversion regime not limited to this can also be adopted
With various modes such as resistance ladder type (R-2R ladder types etc.), capacitor array type or PWM-types.In addition, D/A converter sections 80 are removed
Beyond D/A converter, circuit, modulation circuit (jitter modulation or PWM etc.), filter circuit can also be controlled comprising its
Deng.
Oscillating circuit 150 uses the output voltage VQ and oscillator XTAL of D/A converter sections 80, generation oscillator signal OSCK.Shake
Swing circuit 150 and be connected to oscillator XTAL via the 1st, the 2nd oscillator terminal (oscillator pad).For example, oscillating circuit 150 is logical
Crossing makes oscillator XTAL (piezoelectric vibrator, resonator etc.) vibrate and generate oscillator signal OSCK.Specifically, oscillating circuit 150 makes
Oscillator XTAL is so that the frequency of oscillation of the output voltage VQ of D/A converter sections 80 as frequency control voltage (vibrational control voltage) to be entered
Row vibration.For example, being to control the circuit (VCO) that the vibration to oscillator XTAL is controlled using voltage in oscillating circuit 150
In the case of, the variable capacitance capacitor (transfiguration that oscillating circuit 150 can change comprising capacitance according to frequency control voltage
Diode etc.).
In addition, as described above, oscillating circuit 150 can be realized by direct digital synthesiser mode, in this case,
Oscillator XTAL frequency of oscillation turns into reference frequency, the frequency as the frequency of oscillation different from oscillator signal OSCK.
So, the circuit arrangement of present embodiment is included:Processing unit 50, it is carried out at signal to frequency control data DFCI
Reason;And oscillator signal generative circuit 140, it uses frequency control data DFCQ and oscillator XTAL from processing unit 50, raw
Into the oscillator signal OSCK by the frequency control data DFCQ frequencies of oscillation set.
Moreover, processing unit 50 was carried out by last time in the renewal processing for the priori estimates that Kalman filtering is handled
The posterior estimate at moment and corrected value are added processing to obtain the processing of priori estimates at the time of this.Moreover, root
The result handled according to the Kalman filtering carries out the age correction of frequency control data.In addition, the addition processing of present embodiment
It is subtraction process comprising the processing plus negative value.
That is, in Kalman filtering processing, it is repeated that observation updates and the time more newly arrives estimated state.Updated in observation
In, kalman gain, posterior estimate, renewal of posteriority covariance etc. are carried out according to the result that observation and time update.
During time updates, following prediction processing is carried out:The result updated according to observation, (time step k) elder generation at the time of obtaining this
Test estimate, priori covariance.
Moreover, in the present embodiment, as shown in figure 3, processing unit 50 is carrying out the prior estimate of Kalman filtering processing
During the time of the renewal processing of value updates, pass through the posterior estimate of (time step k-1) at the time of last time and being added for corrected value
Processing, (time step k) priori estimates at the time of obtaining this.And handled according to Kalman filtering result (true value, or
Person's true value and corrected value etc.) carry out frequency control data age correction.
In this case, the observation residual error during processing unit 50 is handled according to Kalman filtering obtains the corrected value.Observation is residual
Difference is corresponding with the difference value of observation and priori estimates.For example, carrying out following prediction processing:By entering to be about to based on card
The posterior estimate at the time of corrected value obtained in the renewal processing of observation residual error in Kalman Filtering processing was with last time is added
Processing, priori estimates at the time of obtaining this.In such manner, it is possible to residual using the observation reflected in Kalman filtering processing
Corrected value that is poor and being updated realizes age correction, can realize the age correction of higher precision.
More specifically, as shown in figure 4, processing unit 50 carry out last time at the time of be time step k-1 posterior estimate x^
(k-1) processing is added with corrected value D (k-1), passes through x^-(k)=x^ (k-1)+D (k-1) i.e. times at the time of obtain this
Walk k priori estimates x^-(k).In addition, here, by represent be the hat of estimate symbol "^" it is properly arranged to 2 words
Accord with being recorded.
I.e., as described later, in the processing of common Kalman filtering, by based on x^-(k)=Ax^ (k-1) time
Renewal handles to obtain time step k priori estimates x^-(k).Here, the time in the case that A there will be no system noise
The state relation for walking the state and time step k+1 system of k system gets up, and is referred to as sytem matrix.In this case, in order to
The value of the A is accurately obtained, it is necessary to be referred to as the processing of EKF processing.
Sometimes processing load is weighed very much for EKF processing.In this case, when spreading kalman to be realized is filtered
During ripple processing, the circuit area of processing unit 50 easily becomes very large, when the circuit arrangement to being built in oscillator is strongly required
It is inappropriate under the situation of miniaturization.On the other hand, when use fixed value as A value when, realize appropriate age correction
When difficulty improve.
Therefore, the solution in the case of avoiding such situation as needs, in the present embodiment, following article is detailed
Thin narration is such, substitutes x^-(k)=Ax^ (k-1), and use x^-(k)=x^ (k-1)+D (k-1), realization obtains time step k's
Priori estimates x^-(k) time renewal processing.Use the x^-(k)=x^ (k-1)+D (k-1) processing can pass through high-ranking officers
It is added the processing of such light load with posterior estimate x^ (k-1) to realize on the occasion of D (k-1).Therefore, because without being expanded
The processing of the such load weight of Kalman filtering processing is opened up, therefore, the processing load of processing unit 50 is alleviated, inhibits circuit to fill
Increase for the circuit scale put etc..That is, can be by the age correction of more small-scale circuit realiration higher precision.
Moreover, in the present embodiment, as shown in figure 4, corrected value D (k-1) and card of the processing unit 50 according to time step k-1
Observation residual error in Kalman Filtering processing, obtains time step k corrected value D (k).Specifically, to set observation residual error be ek, normal
In the case that number is E, corrected value D (k) is obtained by D (k)=D (k-1)+Eek.In addition, setting kalman gain as G (k)
In the case of, corrected value D (k) can also be obtained by D (k)=D (k-1)+G (k) ek.Here, it is being set to observation y
(k) in the case of, observation residual error can be expressed as ek=y (k)-x^-(k)。
For example as described later, corrected value D (k) is the aging for eliminating and compensating the slope by the C3 equivalent to Figure 10
The corrected value that frequency changes caused by speed.Moreover, from Fig. 1 A1~A5, Figure 10, the rate of ageing passes through over time
And change.Therefore, when use fixed value as age correction corrected value when, it is impossible to realize appropriate age correction.
In this regard, in the present embodiment, as D (k)=D (k-1)+Eek, carrying out using in Kalman filtering processing
Observation residual error update age correction corrected value processing.Therefore, even in A1~A5 such as Fig. 1, Figure 10, aging is fast
Rate is passed through according to the time and in the case of variation, also can obtain compensation base by D (k)=D (k-1)+Eek renewal processing
In the corrected value of the frequency change of the rate of ageing of variation.Therefore, it is possible to the old of more small-scale circuit realiration higher precision
Change correction.
In addition, the constant E in D (k)=D (k-1)+Eek expects to be pre-stored within Fig. 9 described later storage part 34.Example
When manufacture such as in product (oscillator), shipment, by by appropriate constant E write-ins corresponding with product for example by it is non-easily
Stored in the storage parts 34 of realization such as the property lost memory.Thereby, it is possible to be realized using appropriate constant E corresponding with each product
Corrected value D (k)=D (k-1)+Eek renewal processing, can realize the age correction of higher precision.
3. the pattern of holding
Then, holding pattern is described in detail.Fig. 5 is the figure of age correction when illustrating holding pattern.Frequency control data
Generating unit 40 carries out the input signal (input clock signal) based on oscillator signal and the reference signal (base from GPS or network
Clock signal) phase bit comparison (comparison operation), generate frequency control data.In usual action, selector 48 will come from
The frequency control data of frequency control data generating unit 40 is output to oscillator signal generative circuit 140.Oscillator signal generative circuit
The frequency control data is converted to frequency control voltage by 140 D/A converter sections 80, is output to oscillating circuit 150.Oscillating circuit
150 make oscillator XTAL be vibrated with frequency of oscillation corresponding with the frequency control voltage, generate oscillator signal.By FREQUENCY CONTROL
Data generating section 40 and oscillator signal generative circuit 140 form the loop of PLL circuit, can make the input based on oscillator signal
Signal and reference signal are synchronous.
Detect that circuit 47 carries out the detection operation of reference signal, detection reference signal disappears or abnormal holding pattern.
Detect after holding pattern, age correction portion 56 is carried out for compensating the frequency control data being stored in register 49 by old
The age correction of frequency variation caused by changing.Also, oscillator signal generative circuit 140 make oscillator XTAL according to the age correction
The corresponding frequency of oscillation of frequency control data afterwards is vibrated, and generates oscillator signal.Thereby, it is possible to supply in self-oscillation
Oscillator signal, is used as the reference signal source of the electronic equipments such as base station.
Fig. 6 B1 represent to generate holding pattern in the case of preferable frequency of oscillation aging characteristic.The opposing party
Face, B2 (dotted line) represents the characteristic for causing frequency of oscillation to change due to aging.B3 is the change of the frequency of oscillation caused by aging
Dynamic amplitude.In addition, Fig. 7 B4, which represents to generate, is used for the frequency control voltage close to B1 characteristic in the case of holding pattern
Passage.On the other hand, frequency control voltage B5 (dotted line) is represented at the time of generating reference signal and disappearing or be abnormal
For constant state.
In order to enter to exercise the correction that preferable characteristic shown in characteristic and B1 shown in Fig. 6 B2 is close, aging school is carried out
Just.If for example, by age correction, and changing frequency control voltage as shown in Fig. 7 B4, then it can enter enforcement figure
The correction of preferable characteristic shown in characteristic close to B1 shown in 6 B2, if for example, improve correction accuracy, can be by B2
Shown characteristic correction is the preferable characteristic shown in B1.On the other hand, age correction is not carried out shown in the B5 such as Fig. 7
In the case of, as shown in Fig. 6 B2, during holding pattern, frequency of oscillation changes, if for example, to keeping model utility
The requirement specification of energy is the B1 shown in Fig. 6, then can not meet the requirement.
The holding mould of the offset (total amount) of the time of such as variation based on frequency of oscillation during expression holding pattern
Formula time θtotIt can be represented as following formula (1).
Here, T1Represent the elapsed time of the aging caused by holding pattern.f0It is nominal oscillation frequency, Δ f/f0It is frequency
Rate deviation.In above formula (1), T1×f0Total clock number is represented, (Δ f/f0)×(1/f0) represent 1 clock at the time of skew
Amount.Moreover, frequency deviation f/f0Holding mode time θ can be usedtotWith elapsed time T1, represented as above formula (2).
As shown in Fig. 8 B6, it is assumed that frequency deviation f/f0Relative to the elapsed time in 1 function with constant slope
Change.In this case, as shown in Fig. 8 B7, with elapsed time T1It is elongated, keep mode time θtotBecome in 2 functions
It is long.
For example, in the case of TDD modes, in order to prevent from setting the Time Slot Overlap of guard time, it is desirable to keep pattern
Time is such as θtotThe μ s of < 1.5.Therefore, from above formula (2), the frequency deviation f/f allowed as oscillator0, it is desirable to
Very small value.Especially, elapsed time T1It is longer, the smaller value of the tolerance frequency deviation requirement.For example, as from holding
From the generation moment of pattern, to using time of the upkeep operation untill at the time of keeping pattern recovery exemplified by time for assuming
Such as T1In the case of=24 is small, tolerance frequency deviation is used as, it is desirable to very small value.It is additionally, since in frequency deviation f/f0
In the frequency departure comprising such as temperature-independent and the frequency departure caused by aging, therefore, in order to meet above-mentioned requirements, it is necessary to
Very high-precision age correction.
4. the detailed construction example of circuit arrangement
Fig. 9 shows the detailed construction example of the circuit arrangement of present embodiment.In fig .9, the structure to Fig. 2 is further set
Temperature sensor 10, A/D converter sections 20, I/F portions 30, register portion 32, storage part 34 and frequency control data generating unit 40
(being phase comparing section in broad terms).In addition, the structure of circuit arrangement is not limited to Fig. 9 structure, it can implement to omit one portion
Various modifications separation structure key element (such as frequency control data generating unit) or additional other structures key element.It is for instance possible to use
The temperature sensor of outside of circuit arrangement is arranged at as temperature sensor 10.
The output temperature of temperature sensor 10 detection voltage VTD.Specifically, the temperature according to environment (circuit arrangement) is exported
And the temperature-independent voltage changed, it is used as temperature detection voltage VTD.Concrete structure example on temperature sensor 10 will be aftermentioned.
A/D converter sections 20 carry out the A/D conversions of the temperature detection voltage VTD from temperature sensor 10, output temperature inspection
Survey data DTD.For example export digital temperature detection data DTD corresponding with temperature detection voltage VTD A/D transformation results
(A/D result datas)., for example can be using gradually manner of comparison or with gradually comparing as the A/D conversion regimes of A/D converter sections 20
Compared with mode similar mode etc..Also, A/D conversion regimes are not limited to this mode, (attribute, parallel connection in various manners can be adopted
Compare type or Serial-Parallel Type etc.).
Digital I/F portions (interface portion) 30 be used for circuit arrangement and external device (ED) (microcomputer, controller etc.) it
Between input and output numerical data interface.Digital I/F portions 30 for example can be by using serial time clock line and serial data line
The serial communication mode of synchronous mode realize.Specifically, I2C (Inter-Integrated Circuit can be passed through:It is interior
Portion's integrated circuit) mode, 3 lines or 4 lines SPI (Serial Peripheral Interface:Serial Peripheral Interface (SPI)) mode
Deng realization.I2C modes are communicated by this 2 signal wires of serial time clock line SCL and two-way serial data line SDA
Synchronous mode serial communication mode.It can connect multiple from device in I2C bus, main device is specifying what is be individually determined
From the address of device, select after device, communicated with this from device.SPI modes are by serial time clock line SCK and list
To the serial communication mode of synchronous mode that is communicated of 2 serial data lines SDI, SDO.It can be connected in SPI bus
It is multiple from device, and in order to determine these from device, main device needs to use from device selection line to select from device.Digital I/
The input and output buffer circuit and control circuit of F portions 30 by realizing these communication modes etc. is constituted.
Register portion 32 is the circuit being made up of multiple registers such as status register, command register, data register.
The external device (ED) of circuit arrangement accesses each register in register portion 32 via digital I/F portions 30.And external device (ED) can
The state of circuit arrangement is confirmed using the register in register portion 32, order is sent to circuit arrangement, circuit arrangement is transmitted
Data and read data etc. from circuit arrangement.
Various information needed for the various processing and action of the memory circuit arrangement of storage part 34.The storage part 34 for example can
Realized by nonvolatile memory.As nonvolatile memory, such as can use EEPROM.As EEPROM,
MONOS (Metal-Oxide-Nitride-Oxide-Silicon can for example be used:Metal oxidation-silicon oxynitride) type storage
Device etc..The flash memory for the memory that make use of MONOS types can for example be used.Or as EEPROM, floating gate type etc. can be used
Other kinds of memory.In addition, as long as storage part 34 can also preserve the storage of simultaneously storage information even if power supply is not supplied
Device, such as also can by fuse circuit realize.
In this case, processing unit 50 is in addition to Kalman filtering portion 54, age correction portion 56, also with the pattern of holding
Processing unit 52 (circuit or program module that keep mode treatment), (circuit or program of temperature-compensating processing of temperature compensation division 58
Module).Mode treatment portion 52 is kept to carry out the various processing related to keeping pattern.Temperature compensation division 58 (processing unit 50) basis
Temperature detection data DTD from A/D converter sections 20, enters the temperature-compensating processing of line of hitch oscillator frequency.Specifically, temperature-compensating
Portion 58 is according to the temperature detection data DTD's (temperature-independent data) and temperature-compensating processing changed corresponding to temperature
Coefficient data (coefficient data of approximate function) etc., carries out the change for reducing frequency of oscillation in the case where there is temperature change
Dynamic temperature-compensating processing.
Reference signal RFCK is input to circuit via the terminal TRFCK (pad) of the external connection terminals as circuit arrangement
Device.The signal PLOCK notified whether is in the lock state to outside PLL circuit via as the outside of circuit arrangement to connect
The terminal TPLOCK (pad) of connecting terminal is input to circuit arrangement.
Moreover, storage part 34 storage Kalman filtering processing system noise setting system noise constant (V) and
The observation noise constant (W) of the setting of the observation noise of Kalman filtering processing.Such as the manufacture in product (oscillator),
During shipment, the measurement (inspection) for monitoring the various information such as frequency of oscillation is carried out.And system is determined according to the measurement result
Noise constant and observation noise constant, and write such as in the storage part 34 of the realization as nonvolatile memory.In such manner, it is possible to
Realize the setting for reducing dysgenic system noise constant and observation noise constant caused by component deviation.
In this case, the circuit arrangement of present embodiment, which has, is used to monitor priori estimates and the digital I/ of observation
F portions 30.That is, external device (ED) can monitor the priori estimates x in Kalman filtering processing via digital I/F portions 30^-(k) and see
Measured value y (k).
In inspection operation when manufacture specifically, in product, shipment, as the check device of external device (ED) via number
The monitoring priori estimates x^ of word I/F portions 30-(k) with observation y (k).For example in the Kalman filtering portion shown in Figure 20 described later
In 54, check device can monitor priori estimates x^ via digital I/F portions 30-(k) signal value and observation y of node
(k) signal value of node.It will for example be used to monitor priori estimates x^-(k) signal value of node and observation y's (k)
The monitoring register of the signal value of node is arranged at register portion 32.And check device via digital I/F portions 30 by accessing
The monitoring register, can obtain priori estimates x^-(k) with observation y (k).
Moreover, check device can be according to the priori estimates x^ of acquirement-(k) with observation y (k), for example, observation is obtained
Residual error ek=y (k)-x^-(k).And obtain so that observe residual error ek as smaller value system noise constant V and observation make an uproar
Acoustic constant W.Or, obtain the best constant E in corrected value D (k)=D (k-1)+Eek calculation process.Moreover, in product
Manufacture, shipment when inspection operation in, calculated system noise constant V, observation noise constant W or constant E are write
Storage part 34.Thus, in each product, best constant V, W, E can be write storage part 34.Moreover, in usual action, leading to
The Kalman filtering processing for performing the present embodiment for having used these constants V, W, E is crossed, the aging of higher precision can be realized
Correction.
The generation frequency control data of frequency control data generating unit 40 DFCI.For example by the input based on oscillator signal OSCK
Signal is compared with reference signal RFCK, generation frequency control data DFCI.The frequency control data DFCI generated is defeated
Enter to processing unit 50.Here, the input signal based on oscillator signal OSCK can be oscillator signal OSCK in itself or by
The signal (such as the signal after dividing) of oscillator signal OSCK generations.Hereinafter, with input signal it is mainly oscillator signal OSCK sheets
Illustrated in case of body.
Frequency control data generating unit 40 includes phase comparing section 41 and digital filtering part 44.Phase comparing section 41 (compares
Operational part) it is electricity of the progress as the oscillator signal OSCK and reference signal RFCK of input signal phase bit comparison (comparison operation)
Road, includes counter 42, TDC 43 (time-to-digit converter).
Counter 42 generates numerical data, and the numerical data with reference signal RFCK reference frequency (such as 1Hz) with being removed
With the integer portion correspondence of result obtained by oscillator signal OSCK frequency of oscillation.TDC 43 generates the fractional part with the result of division
Corresponding numerical data.TDC 43 for example comprising:Multiple delay elements;Multiple latch cicuits, they are reference signal RFCK's
Regularly the multiple delay clock signals exported by multiple delay elements are latched at edge (height);And circuit, it is more by carrying out
The coding of the output signal of individual latch cicuit, generates numerical data corresponding with the fractional part of result of division.Moreover, phase bit comparison
Portion 41 is by the numerical data corresponding with integer portion from counter 42 and from the digital numbers corresponding with fractional part of TDC 43
According to addition, the phase error between detection and setpoint frequency.Moreover, digital filtering portion 44 is by carrying out the smoothing of phase error
Processing, generation frequency control data DFCI.For example setting frequency of the oscillator signal OSCK frequency as FOS, reference signal RFCK
For FRF, in the case that divider ratio corresponding with setpoint frequency (frequency dividing ratio) is FCW, so that FOS=FCW × FRF relation is set up
Mode generate frequency control data DFCI.Or, counter 42 can be counted to oscillator signal OSCK clock number.
That is, counter 42 carries out counting action by the input signal based on oscillator signal OSCK.Also, phase comparing section 41 can lead to
Cross integer, by reference signal RFCK n cycle (n be may be set to more than 2 integer) in counter 42 count value and
The desired value (n × FCW) of count value is compared.Slave phase bit comparison portion 41 exports the count value of such as desired value and counter 42
Difference, be used as phase error data.
In addition, the structure of frequency control data generating unit 40 is not limited to the structure shown in Fig. 9, various modifications can be implemented.
Phase comparing section 41 can be for example made up of the phase comparator of analog circuit or (loop is filtered by the filtering part of analog circuit
Ripple device) constitute digital filtering portion 44.In addition, processing unit 50 can carry out digital filtering portion 44 processing (phase error data
Smoothing techniques).Sequentially enter line number such as processing unit 50 and other processing (keeping mode treatment, Kalman filtering processing)
The processing of word filtering part 44.For example, carrying out phase comparative result (the phase error number for phase comparing section 41 by processing unit 50
According to) filtering process (smoothing techniques).
In addition, in fig .9, circuit arrangement is the structure for being built-in with frequency control data generating unit 40, but FREQUENCY CONTROL
Data generating section can also be disposed on the circuit of the outside of circuit arrangement.In this case, if from be arranged at outside frequency
Frequency control data DFCI is input to processing unit 50 by rate control data generating unit via digital I/F portions 30.
So, in the present embodiment, processing unit 50 (processor), which be directed to, is based on input signal and reference signal
The frequency control data DFCI of RFCK phase comparative result signal transacting, the input signal is based on oscillator signal OSCK.That is,
Processing unit 50 carries out signal transacting for the frequency control data DFCI based on the phase comparative result in phase comparing section 41.Example
Such as, the frequency control data DFCI from frequency control data generating unit 40, the frequency control data are inputted in processing unit 50
Input signal based on oscillator signal OSCK is compared and generates frequency control data by generating unit 40 with reference signal RFCK
DFCI.Processing unit 50 can be carried out at the filtering for phase comparative result with the phase comparative result of input phase comparing section 41
Manage (processing in digital filtering portion 44).Also, processing unit 50 (processor) is detecting disappearance or exception by reference signal
In a period of before caused holding pattern, it is handled as follows:Estimation is handled by Kalman filtering to be directed to based on phase ratio
The true value of the frequency control data DFCI of relatively result observation.The true value be by Kalman filtering handle estimate it is true
Value, is not limited to real true value.Kalman filtering processing is performed by Kalman filtering portion 54.In addition, based on the detection of holding pattern
Control process performed by holding mode treatment portion 52.
Moreover, processing unit 50 (processor) is in the case where detecting holding pattern, when preserving the detection with keeping pattern
Carve true value at the time of correspondence.Can be to maintain at the time of preserving the true value detection moment of pattern in itself or this when
At the time of before quarter etc..Moreover, processing unit 50 is aging correction by carrying out the calculation process based on the true value preserved, generation
Frequency control data DFCQ afterwards.The frequency control data DFCQ of generation is output to oscillator signal generative circuit 140.The aging
The generation processing of frequency control data DFCQ after correction is performed by age correction portion 56.
In for example during usual action, frequency control data DFCI of 50 pairs of the processing unit based on phase comparative result is carried out
The signal transacting such as being handled temperature-compensating, and the frequency control data DFCQ after signal transacting is output to oscillator signal generation
Circuit 140.Oscillator signal generative circuit 140 uses frequency control data DFCQ and oscillator XTAL from processing unit 50, generation
Oscillator signal OSCK, and it is output to frequency control data generating unit 40 (phase comparing section 41).Thus, form based on frequency control
The loop of the PLL circuit of data generating section 40 (phase comparing section 41) processed, oscillator signal generative circuit 140 etc., so as to life
Into the phase locked accurate oscillator signal OSCK with reference signal RFCK.
And in the present embodiment, interior during the usual action before detecting holding pattern, processing unit 50
Kalman filtering portion 54 also acted, to frequency control data DFCI perform Kalman filtering processing.That is, located as follows
Reason:True value of the estimation for frequency control data DFCI observation is handled by Kalman filtering.
When detecting holding pattern, true value at the time of will be corresponding with the detection moment for keeping pattern is saved in processing unit
In 50.Specifically, age correction portion 56 preserves the true value.Moreover, age correction portion 56 is true based on what is preserved by carrying out
Frequency control data DFCQ after the calculation process of value, generation age correction.
So, due to carrying out age correction according to true value of the detection moment with keeping pattern at the time of corresponding, accordingly, it is capable to
Enough increase substantially the precision of age correction.That is, the aging school for the influence for considering observation noise and system noise can be realized
Just.
In addition, oscillator signal generative circuit 140 is in the case of from holding pattern recovery, according to based on phase comparative result
Frequency control data DFCQ, generation oscillator signal OSCK.For example according to from (the phase comparing section of frequency control data generating unit 40
41) the frequency control data DFCQ inputted via processing unit 50, generation oscillator signal OSCK.For example when eliminating reference signal
When RFCK vanishing state or abnormality, the state of pattern is kept to be released from, from holding pattern recovery.In this case, it is electric
The action of road device returns to usual action.And oscillator signal generative circuit 140 not according to processing unit 50 by carrying out aging
The frequency control data DFCQ for correcting and generating, and according to the frequency inputted from frequency control data generating unit 40 via processing unit 50
Rate control data DFCQ (frequency control data after the signal transacting such as temperature-compensating processing), generation oscillator signal OSCK.
In addition, processing unit 50 (is compensated and led by aging by the calculation process carried out to the true value preserved plus corrected value
The calculation process of the frequency change of cause), generate the frequency control data DFCQ after age correction.For example by each defined
Corrected value corresponding with rate of ageing (gradient of aging, aging coefficient) (is eliminated the frequency caused by rate of ageing by the moment successively
Rate change corrected value) and with keep pattern detection the moment it is corresponding at the time of true value be added, generate age correction after frequency
Rate control data DFCQ.
It is AC (k) that time step k corrected value, which is for example set, as the frequency control data after D (k), time step k age correction.
In this case, processing unit 50 obtains the FREQUENCY CONTROL after time step k+1 age correction by AC (k+1)=AC (k)+D (k)
Data AC (k+1).Processing unit 50 carries out the corrected value D (k) of each such time step addition processing, until from the pattern of holding
Untill (moment being released at the time of recovery).
In addition, processing unit 50 true value is added the calculation process of the corrected value after filtering process.For example, to corrected value
D (k) carries out the filtering process such as low-pass filtering treatment, true value is added successively the fortune of the corrected value D ' (k) after filtering process
Calculation is handled.Specifically, AC (k+1)=AC (k)+D ' (k) calculation process is carried out.
In addition, processing unit 50 handled according to Kalman filtering in observation residual error, obtain corrected value.For example, processing unit 50
During before detecting holding pattern, the processing of the corrected value according to observation residual error estimation age correction is carried out.For example exist
If observing residual error in the case of ek, by carrying out D (k)=D (k-1)+Eek processing, to estimate corrected value D (k).Here E
E.g. constant, but it is also possible to substitute constant E, and use kalman gain.Moreover, preserving the detection moment pair with keeping pattern
Corrected value at the time of answering, the calculation process that the corrected value for being about to preserve of going forward side by side is added with true value, is thus generated after age correction
Frequency control data DFCQ.
In addition, processing unit 50 has the voltage of the input terminal of the detection signal of holding pattern according to input or via numeral
The detection information for the holding pattern that I/F portions 30 are inputted, judges whether to turn into the state of holding pattern.These judge processing by protecting
Mode treatment portion 52 is held to carry out.For example keep mode treatment portion 52 that there is the circuit of state machine, the state transformation of the state machine is
Performed according to various signals and information.Moreover, the electricity of the input terminal when the detection signal for having holding pattern according to input
Pressure and detection information of holdings pattern etc. for being inputted via digital I/F portions 30 and when judging to be in the state of holding pattern,
The state of state machine is changed into the state of holding pattern.Then the various processing (age correction etc.) during holding pattern are performed.
For example it can be assumed that reference signal RFCK and signal PLOCK, is used as the detection signal of the pattern of holding.In the situation
Under, processing unit 50 has reference signal RFCK terminal TRFCK voltage according to input, inputs the terminal for having signal PLOCK
TPLOCK voltage, judges whether to turn into the state of holding pattern.
For example, the feelings of the formation PLL circuit of frequency control data generating unit 40 in the inside by being arranged at circuit arrangement
Under condition, there can be reference signal RFCK terminal TRFCK voltage according to input, judge whether the state in holding pattern.
Such as processing unit 50 detects states of the reference signal RFCK in disappearance or exception in the voltage according to terminal TRFCK
In the case of, judge the state in holding pattern.
On the other hand, the frequency control data generating unit in the outside by being arranged at circuit arrangement forms the feelings of PLL circuit
Under condition, there can be signal PLOCK terminal TPLOCK voltage according to input, judge whether to turn into the state of holding pattern.
For example external device (ED) (device of the outside PLL circuit of control) will notify whether outside PLL circuit turns into the signal of lock-out state
PLOCK is output to circuit arrangement.And for example it is being judged as that outside PLL circuit does not turn into lock-out state by signal PLOCK
In the case of, processing unit 50 is judged as the state in holding pattern.In addition, in addition to signal PLOCK, benchmark can also be used
Signal RFCK, to judge whether to turn into the state of holding pattern.In addition, outside PLL circuit is, for example, by being arranged at circuit dress
The PLL circuit of the composition such as the frequency control data generating unit for the outside put and the oscillator signal generative circuit 140 of circuit arrangement.
In addition, the situation of the frequency control data generating unit formation PLL circuit in the outside by being arranged at circuit arrangement
Under, it can judge whether to turn into the shape of holding pattern according to the detection information of the holding pattern inputted via digital I/F portions 30
State.For example controlling disappearance or exception of the external device (ED) (such as microcomputer) of outside PLL circuit according to reference signal
And in the case of being judged as the state as holding pattern, the detection information of the pattern of holding is set via digital I/F portions 30
Due to the register (notice register) in register portion 32.Processing unit 50 is set in the holding pattern of the register by reading
Detection information come judge whether turn into holding pattern state.This way it is not necessary to newly set the end of the detection of holding pattern
Son, realizes reduction of the number of terminals of circuit arrangement etc..
5. the age correction for having used Kalman filtering to handle
In the present embodiment, the burn-in correction method handled using Kalman filtering is employed.Specifically, in this reality
Apply in mode, in a period of before detecting holding pattern, estimation is handled by Kalman filtering and is directed to frequency control data
The true value of the observation of (frequency of oscillation).Moreover, in the case where detecting holding pattern, when preserving the detection with keeping pattern
True value at the time of carving correspondence under (time point), and the calculation process based on the true value preserved is carried out, hereby it is achieved that aging school
Just.
Figure 10 is the figure for the measurement result example for showing the variation of frequency of oscillation caused by aging.Transverse axis is the elapsed time
(ageing time), the longitudinal axis is frequency departure (the Δ f/f of frequency of oscillation0).As shown in Figure 10 C1, in the measurement as observation
There is deviation big as caused by system noise, observation noise in value.Also included as caused by environment temperature partially in the deviation
Difference.
When so existing in the measurement under the situation of big deviation, in order to correctly obtain true value, in present embodiment
In, handled based on Kalman filtering the state estimation of (such as linear Kalman filter processing).
Figure 11 shows the state-space model of time series, the discrete time state equation formula of the model by following formula (3),
(4) equation of state, observation equation are provided.
X (k+1)=Ax (k)+v (k) (3)
Y (k)=x (k)+w (k) (4)
X (k) is moment k state, and y (k) is observation.V (k) is system noise, and w (k) is observation noise, and A is system
Matrix.In the case where x (k) is frequency of oscillation (frequency control data), A is for example equivalent to rate of ageing (aging coefficient).Always
Change rate representation frequency of oscillation relative to the rate of change during process.
For example, generating holding pattern at the time of being set to shown in the C2 in Figure 10.In this case, according to reference signal
Time of day x (k) at the time of the C2 that RFCK is interrupted and rate of ageing (A) execution equivalent to the slope shown in Figure 10 C3
Age correction.Specifically, as reducing the compensation (correction) changed as frequency caused by the rate of ageing shown in C3, example
Such as to eliminate the corrected value that (counteractings) frequency changes, the true value of the frequency of oscillation (frequency control data) at the time of entering to exercise C2
The age correction that x (k) changes successively.That is, the frequency change under the rate of ageing shown in Fig. 6 B2 is eliminated, so as to B1
The corrected value of shown preferable characteristic changes true value x (k).So, for example the feelings during holding pattern for 24 hours
Under condition, Figure 10 of variation as the frequency of oscillation after 24 hours FDV can be compensated by age correction.
Here, include and become as caused by temperature change in the variation of frequency of oscillation (frequency departure) shown in the C1 in Figure 10
The dynamic and variation as caused by aging.Therefore, in the present embodiment, for example by using the constant temperature slot structure with thermostat
Oscillator (OCXO), the variation of the frequency of oscillation as caused by temperature change is suppressed to Min..In addition, using Fig. 9's
The grade of temperature sensor 10 performs the temperature-compensating processing of the variation of reduction frequency of oscillation as caused by temperature change.
Moreover, during PLL circuit (internal PLL circuit, outside PLL circuit) is synchronous with reference signal RFCK (generally
During action) in, monitoring frequency control data (FREQUENCY CONTROL code) is obtained after removal error (system noise, observation noise)
True value, and it is stored in register.Moreover, in the disappearance due to reference signal RFCK or lock that is abnormal and relieving PLL circuit
In the case of fixed, held according to the true value (for the true value of the observation of frequency control data) preserved at the time of latch-release
Row age correction.For example, as reduce by Figure 10 C3 slope be rate of ageing caused by frequency change compensation, enter
The true value of frequency control data of the row to being preserved adds the processing for the corrected value for for example eliminating frequency change successively, thus,
Generate holding pattern during self-oscillation when frequency control data DFCQ, make oscillator XTAL vibrate.So, due to can be with
Minimal error is obtained into true value at the time of holding pattern, and performs age correction, will be changed therefore, it is possible to realize by aging
Caused harmful effect is suppressed to minimal holding mode performance.
6. the structure of processing unit
Figure 12 shows the detailed construction example of processing unit 50.In addition, the structure of processing unit 50 is not limited to Figure 12 structure, can
Implement to omit the various modifications such as a portion structural element or additional other structures key element.
As shown in figure 12, processing unit 50 includes Kalman filtering portion 54, age correction portion 56, temperature compensation division 58, selection
Device 62,63 and adder 65.
The input of Kalman filtering portion 54 has frequency control data DFCI (to eliminate the FREQUENCY CONTROL number of environmental turbulence composition
According to), perform Kalman filtering processing.Moreover, output is estimated equivalent to the posteriority that the true value estimated is handled by Kalman filtering
Evaluation x^ (k).In addition, in this manual, by represent be the hat of estimate symbol "^" it is properly arranged to 2 characters
To be recorded.
Kalman filtering processing refers to following processing:Made an uproar assuming that being included in the variable of observation and the state for representing system
Sound (error), uses the optimum state for carrying out estimating system from the past to the observation obtained now.Specifically, sight is repeated
Survey and update (observation process) and time renewal (prediction process), estimated state.Observation renewal is to be updated using observation with the time
Result update the process of kalman gain, estimate, error covariance.Time is updated the result for being to be updated using observation and come
Predict estimate, the process of error covariance of subsequent time.In addition, in the present embodiment, primarily illustrating using linear
The method of Kalman filtering processing, but also can be using EKF processing.Kalman on present embodiment filters
The details of ripple processing, will be described later.
Age correction portion 56 inputs posterior estimate x^ (k) and corrected value D ' (k) from Kalman filtering portion 54.Moreover, logical
The calculation process for carrying out that corrected value D ' (k) is added to the posterior estimate x^ (k) of the true value equivalent to frequency control data is crossed, it is raw
It is AC (k) into the frequency control data after age correction.Here D ' (k) is the correction (after low-pass filtering treatment) after filtering process
Value D (k).That is, setting time step k, (moment k) corrected value (corrected value after filtering process) is D ' (k), time step k aging
In the case that frequency control data after correction is AC (k), age correction portion 56 is obtained by AC (k+1)=AC (k)+D ' (k)
Frequency control data AC (k+1) after time step k+1 (moment k+1) age correction.
The input of temperature compensation division 58 has temperature detection data DTD, carries out temperature-compensating processing, generates for making frequency of oscillation
Stationary temperature offset data TCODE (temperature-compensating code) is remained relative to temperature change.Temperature detection data DTD is to pass through
The data as obtained from the temperature detection voltage VTD from temperature sensor 10 is carried out A/D conversions by Fig. 9 A/D converter sections 20.
For example, showing the example of initial oscillation temperature characterisitic in Figure 13, Figure 14, Figure 15.In these figures, transverse axis
It is environment temperature, the longitudinal axis is the frequency departure of frequency of oscillation.As shown in Figure 13~Figure 15, the temperature characterisitic of frequency of oscillation is according to every
The sample of individual product and have relatively large deviation.In inspection operation when manufacture therefore, in product (oscillator), shipment, measurement is shaken
Swing the temperature characterisitic of frequency and the variation characteristic of temperature detection data corresponding with environment temperature.And according to measurement result come
Obtain the coefficient A of the multinomial (approximate function) of following formula (5)0~A5, by the coefficient A tried to achieve0~A5Information be written to Fig. 9's
Stored in storage part 34 (nonvolatile memory).
TCODE=A5·X5+A4·X4+A3·X3+A2·X2+A1·X+A0···(5)
In above formula (5), X is equivalent to the temperature detection data DTD (A/D conversion values) obtained by A/D converter sections 20.Due to
The temperature detection data DTD changed relative to environment temperature change is also measured, therefore, passes through the multinomial institute of above formula (5)
Environment temperature, can be mapped by the approximate function of expression with frequency of oscillation.Temperature compensation division 58 is read from storage part 34
Number A0~A5Information, according to coefficient A0~A5The calculation process of above formula (5) is carried out with temperature detection data DTD (=X), it is raw
Into temperature compensation data TCODE (temperature-compensating code).Thereby, it is possible to realize for making frequency of oscillation relative to the change of environment temperature
Change remains stationary temperature compensation deals.
Selector 62,63 is selected " 1 " in the case where the logic level of selection terminal S input signal is " 1 " (effective)
The input signal of the terminal of side, and exported as output signal.In addition, the logic level of the input signal in selection terminal S
In the case of " 0 " (invalid), the input signal of the terminal of " 0 " side is selected, and is exported as output signal.
Signal KFEN is the enable signal of Kalman filtering processing.Kalman filtering portion 54 is logic level in signal KFEN
Kalman filtering processing is performed in the case of " 1 " (following, to be abbreviated as " 1 ").Signal PLLLOCK be PLL circuit for locking shape
Turn into the signal of " 1 " in the case of state.Signal HOLDOVER is to turn into " 1 " during the holding pattern of holding pattern is detected
Signal.These signals PLLLOCK, HOLDOVER are the circuit evolvings of the state machine in the holding mode treatment portion 52 by Fig. 9.
Signal TCEN is the enable signal of temperature-compensating processing.Hereinafter, mainly using signal TCEN as " 1 " and selector 63
Illustrated in case of the input signal for selecting " 1 " side.In addition, signal KFEN is also " 1 ".
During usual action, because signal HOLDOVER is that ((following, to be abbreviated as " 0 ") therefore, is selected logic level " 0 "
Select the frequency control data DFCI that device 62 selects " 0 " terminals side.Moreover, being added by 65 couples of frequency control data DFCI of adder
Upper temperature compensation data TCODE, the frequency control data DFCQ after temperature-compensating processing are output to the oscillator signal life of rear class
Into circuit 140.
On the other hand, during holding pattern, signal HOLDOVER is " 1 ", and selector 62 selects the AC of " 1 " terminals side
(k).AC (k) is the frequency control data after age correction.
Figure 16 is the truth table for the action for illustrating Kalman filtering portion 54.All it is the feelings of " 1 " in signal PLLLOCK, KFEN
Under condition, Kalman filtering portion 54 performs true value estimation processing (Kalman filtering processing).That is, PLL electricity in during usual action
In the case that road (PLL circuit either internally or externally) is in the lock state, the lasting FREQUENCY CONTROL number carried out as observation
Handled according to DFCI true value estimation.
Moreover, in the state as the pattern of holding, the locking of PLL circuit is released, so that signal PLLLOCK is the feelings of " 0 "
Under condition, Kalman filtering portion 54 keeps the output state of last time.For example in fig. 12, preserve and continue to export the inspection of holding pattern
The value gone out the moment under (at the time of the latch-release of PLL circuit), is used as the posteriority for the true value for being estimated as frequency control data DFCI
Estimate x^ (k) and age correction corrected value D ' (k).
Age correction portion 56 during holding pattern in, using the posterior estimate x^ (k) from Kalman filtering portion 54,
Corrected value D ' (k) carries out age correction.Specifically, posterior estimate x^ (k), the correction at the detection moment of holding pattern are preserved
Value D ' (k), carries out age correction.
In addition, in fig. 12, it (is inventionbroadly environment that input, which eliminates temperature change composition, in Kalman filtering portion 54
Variance components) and temperature change composition in aging variance components frequency control data DFCI.54 pairs of Kalman filtering portion is gone
Except the frequency control data DFCI of temperature change composition (environmental turbulence composition) carries out Kalman filtering processing, estimation is for frequency
Rate control data DFCI true value.That is, posterior estimate x^ (k) is obtained.Moreover, age correction portion 56 is according to the true value estimated
That is posterior estimate x^ (k) carries out age correction.More specifically, according to the posterior estimate x^ from Kalman filtering portion 54
(k) the frequency control data AC (k) after age correction is obtained with corrected value D ' (k).Moreover, the FREQUENCY CONTROL number after age correction
Adder 65 is input to via selector 62 according to i.e. AC (k), adder 65 carries out adding temperature compensation data TCODE to AC (k)
The processing of (the compensation data of environmental turbulence composition).
For example, as shown in Figure 17 schematic diagram, when temperature change, as shown in E1, frequency control data is also corresponding
Ground changes.Therefore, when carrying out Kalman filtering using the frequency control data changed as E1 along with temperature change
During processing, the true value at holding pattern detection moment also produces fluctuation.
Therefore, in the present embodiment, the frequency control data for eliminating temperature change composition is obtained, and is input to karr
Graceful filtering part 54.That is, by the temperature change eliminated in temperature change composition (environmental turbulence composition) and aging variance components into
The frequency control data divided is input to Kalman filtering portion 54.That is, the frequency control data shown in input Figure 17 E2.E2 frequency
Rate control data is eliminates the frequency control data that temperature change composition remains aging variance components.
Kalman filtering portion 54 to so eliminating temperature change composition by remaining the frequencies of aging variance components
Control data DFCI carries out Kalman filtering processing, obtains the correction of the posterior estimate x^ (k) for being estimated true value, age correction
Value D ' (k).Moreover, the true value i.e. posterior estimate x^ (k), the corrected value D ' (k) that will be estimated at the detection moment of the pattern of holding
Age correction portion 56 is saved in, for performing age correction.
The processing plus temperature compensation data TCODE is for example carried out by adder 65, frequency control data DFCQ turns into
By the frequency control data after temperature-compensating.Therefore, input has frequency control data DFCQ oscillator signal generative circuit 140 defeated
The oscillator signal OSCK of the frequency of oscillation gone out after temperature-compensating.Therefore, PLL is constituted together with the oscillator signal generative circuit 140
Fig. 9 of circuit frequency control data generating unit 40 will eliminate the frequency control of temperature change composition shown in the E2 such as Figure 17
Data DFCI processed is supplied to processing unit 50.Moreover, as shown in Figure 17 E2, eliminating the FREQUENCY CONTROL of the temperature change composition
The aging variance components changed with the elapsed time are remained in data DFCI.Therefore, the Kalman filtering portion of processing unit 50
54 couples of frequency control data DFCI for remaining the aging variance components carry out Kalman filtering processing, if age correction portion 56
The result handled according to Kalman filtering carries out age correction, then can realize high-precision age correction.
In addition, as Figure 12 variation, can be without the place plus temperature compensation data TCODE in adder 65
Reason, and the calculation process of the temperature change composition (environmental turbulence composition) for removing frequency control data DFCI is carried out, and will
Frequency control data DFCI after calculation process is input to Kalman filtering portion 54.For example omit Figure 12 adder 65 and choosing
The structure of device 63 is selected, the prime in Kalman filtering portion 54 is set subtracts temperature compensation data from frequency control data DFCI
TCODE subtracter, Kalman filtering portion 54 is input to by the output of the subtracter.In addition, in age correction portion 56 and selection
The adder that the output in age correction portion 56 is added with temperature compensation data TCODE is set between device 62, by the defeated of adder
Go out to be input to the terminal of " 1 " side of selector 62.By such structure, also temperature change composition can will be eliminated and only residual
The frequency control data DFCI for leaving aging variance components is input to Kalman filtering portion 54.
Figure 18 shows the detailed construction example in age correction portion 56.During usual action, signal HOLDOVER is
" 0 ", therefore, selection " 0 " terminals side of selector 360,361.Thus, in during usual action, transported by Kalman filtering portion 54
Posterior estimate x^ (k), the corrected value D ' (k) (corrected value after filtering process) calculated be saved in respectively register 350,
351。
When detecting holding pattern, so that when signal HOLDOVER is " 1 ", selector 360,361 selects " 1 " terminals side.
Thus, during selector 361 is during holding pattern, lasting output is stored in register 351 at the detection moment of the pattern of holding
Corrected value D ' (k).
Moreover, adder 340 is handled as follows:According to each time step, to being stored at the detection moment of the pattern of holding
The posterior estimate x^ (k) of register 350 is successively plus the corrected value D ' for being stored in register 351 and being exported from selector 361
(k) (corrected value).Hereby it is achieved that the age correction shown in following formula (6).
AC (k+1)=AC (k)+D ' (k) (6)
That is, it is handled as follows to realize age correction:To the true value i.e. Posterior estimator preserved at the time of Figure 10 C2
Value x^ (k) adds corrected value D ' (k) successively, and the corrected value D ' (k) is used to eliminate the aging of (compensation) by the slope equivalent to C3
Frequency changes caused by speed.
7. Kalman filtering processing
Next, the details to the Kalman filtering processing of present embodiment are illustrated.Figure 19 shows Kalman filtering
Model example.The equation of state of Figure 19 model, observation equation are represented as following formula (7), (8).
X (k+1)=Ax (k)+v (k) (7)
Y (k)=CT·x(k)+w(k)···(8)
K is denoted as the time step of discrete time.X (k) is time step k (state of moment k) system, e.g. n
The vector of dimension.A is referred to as sytem matrix.Specifically, A is n × n matrix, there will be no in the case of system noise when
The state relation of the state of spacer step k system and time step k+1 system gets up.V (k) is system noise.Y (k) is observation,
W (k) is observation noise.C is that observed differential is vectorial (n dimensions), and T represents transposed matrix.
In the Kalman filtering processing of above formula (7), the model of (8), the processing of following formula (9)~(13) is carried out, estimation is true
Value.
P-(k)=AP (k-1) AT+v(k)···(10)
P (k)=(1-G (k) CT)·P-(k)···(13)
x^(k):Posterior estimate
x^-(k):Priori estimates
P(k):Posteriority covariance
P-(k):Priori covariance
G(k):Kalman gain
Above formula (9), (10) are the formulas for the time updating (prediction process), and above formula (11)~(13) are that observation updates (observation
Process) formula.Often advance 1 as the time step k of discrete time, then carry out the time of 1 Kalman filtering processing more
Newly (formula (9), (10)) and observation update (formula (11)~(13)).
X^ (k), x^ (k-1) are the posterior estimates of time step k, k-1 Kalman filtering processing.x^-(k) it is to be seen
The priori estimates predicted before measured value.P (k) be Kalman filtering processing posteriority covariance, P- (k) be obtain observation it
The priori covariance of preceding prediction.G (k) is kalman gain.
In Kalman filtering processing, in observation updates, kalman gain G (k) is obtained by above formula (11).In addition,
According to observation y (k), by above formula (12), posterior estimate x^ (k) is updated.In addition, by above formula (13), updating error
Posteriority covariance P (k).
In addition, in Kalman filtering processing, in the time updates, such as shown in above formula (9), after time step k-1
Estimate x^ (k-1) and sytem matrix A is tested, prediction future time walks k priori estimates x^-(k).In addition, such as above formula (10)
It is shown, according to time step k-1 posteriority covariance P (k-1), sytem matrix A, system noise v (k), prediction future time step k's
Priori covariance P- (k).
In addition, when the Kalman filtering processing of above formula to be performed (9)~(13), the processing load mistake of processing unit 50 sometimes
Greatly, the large-scale of circuit arrangement is caused.For example for the x^ for obtaining above formula (9)-(k)=Ax^ (k-1) A is, it is necessary to expansion card
Kalman Filtering processing.Moreover, the processing load of EKF processing is weighed very much, when will be by that can be extended karr
The hardware of graceful filtering process realizes during processing unit 50 that the circuit area of processing unit 50 easily becomes very large.Therefore, when internal
It is placed in the circuit arrangement of oscillator to be strongly required under the situation of miniaturization, is inappropriate.On the other hand, when use fixed value
When scalar value is as sytem matrix A, realize that difficulty during appropriate age correction is improved.
Therefore, as solution when needing to avoid such situation, in the present embodiment, above formula (9) are not passed through
~(13), and Kalman filtering is realized by the processing based on following formula (14)~(19) and is handled.That is, (the Kalman of processing unit 50
Filtering part 54) perform the Kalman filtering processing based on following formula (14)~(19).
P-(k)=P (k-1)+v (k) (15)
P (k)=(1-G (k)) P-(k)···(18)
In addition, be frequency control data as the x (k) of the object of the estimation processing of true value in the present embodiment, observation
Value y (k) is also frequency control data, therefore, C=1.Further, since A scalar value is infinitely close to 1, therefore, it is possible to use
Above formula (15) substitutes above formula (10).
As described above, compared with the situation for being used as Kalman filtering processing is handled using EKF, at this
In the Kalman filtering processing of embodiment, such as shown in above formula (14), passage time step k-1 posterior estimate x^ (k-1) with
Corrected value D (k-1) addition handles to obtain time k priori estimates x^-(k).It therefore, there is no need to use spreading kalman
Filtering process is excellent in terms of realizing that the mitigation of processing load of processing unit 50, the increase of circuit scale suppress.
In the present embodiment, above formula (14) is exported by the deformation of following formulas.
For example above formula (20) can be deformed as above formula (21).Here, because (A-1) of above formula (21) is very small
Number, therefore, such as shown in above formula (22), (23), can use (A-1) x^ (k-1) being replaced into (A-1) F0It is approximate.So
Afterwards, by (A-1) F0It is replaced into corrected value D (k-1).
And as shown in above formula (19), when being updated from time step k-1 to time step k time, it is corrected value D (k)
=D (k-1)+E (y (k)-x^-(k))=D (k-1)+Eek renewal processing.Here, ek=y (k)-x^-(k) it is referred to as card
Observation residual error in Kalman Filtering processing.In addition, E is constant.In addition, can also substitute constant E, and implement to increase using Kalman
The deformation of beneficial G (k).I.e., it is possible to be D (k)=D (k-1)+G (k) ek.
So, in formula (19), set observation residual error be E as ek, constant in the case of, pass through D (k)=D (k-1)+E
Ek obtains corrected value D (k).In such manner, it is possible to carry out reflecting observation residual error ek, the corrected value D (k) in Kalman filtering processing
Renewal processing.
Figure 20 shows the configuration example in Kalman filtering portion 54.Kalman filtering portion 54 comprising adder 300,301,302,
303rd, 304, multiplier 305, register 310,311,312,313, selector 320,321, wave filter 330,331 and arithmetic unit
332、333.In addition, the structure in Kalman filtering portion 54 is not limited to the structure shown in Figure 20, it can implement to omit a portion
The various modifications such as structural element or additional other structures key element.For example, can handle to realize by the time-division of 1 arithmetic unit
The processing of the grade of adder 300~304.
By adder 304 and register 312, the calculation process of above formula (14) is performed.In addition, the setting of system noise is used
System noise constant V and the observation noise constant W information of setting of observation noise read from Fig. 9 storage part 34,
And it is input to Kalman filtering portion 54 (processing unit 50).Also, by adder 300 and register 310, perform above formula (15)
Calculation process.In addition, arithmetic unit 332 performs the calculation process of above formula (16), kalman gain G (k) is obtained.Also, according to asking
The kalman gain G (k) gone out, by adder 301, multiplier 305 and adder 302, performs the calculation process of above formula (17).
In addition, arithmetic unit 333 performs the calculation process of above formula (18), posteriority covariance P (k) is obtained.
In addition, by adder 303, register 311 and wave filter 330, performing the calculation process of above formula (19).It is input to
The information of constant E in wave filter 330 is read from Fig. 9 storage part 34.Correction coefficient of the constant E equivalent to rate of ageing
(filter constant).For example, wave filter 330 carries out Zeng Yi Tone according to constant E whole etc., the E thus, it is possible to realize above formula (19)
(y(k)-x^-(k))。
In the case where signal PLLLOCK, KFEN are respectively " 1 ", selector 320,321 selects the defeated of the terminal of " 1 " side
Enter signal.The output signal of selector 320 is saved in register 313.Therefore, in state and signal as the pattern of holding
PLLLOCK is changed into after " 0 " from " 1 ", and the true value by the detection moment of the pattern of holding is that x^ (k) is saved in register 313.
Wave filter 331 is filtered processing to corrected value D (k).Specifically, digital lowpass filter is carried out to corrected value D (k)
Ripple processing, the corrected value D ' (k) after filtering process is input to Figure 18 age correction portion 56.Constant J is the filter of wave filter 331
Wave constant.According to constant J, the optimal cut-off frequency of wave filter 331 is set.
For example, it can be seen from Figure 10, existing in the compensation corrected value D (k) that frequency changes caused by rate of ageing thin
The fluctuation of micro- variation.Therefore, after it so will be added in the presence of the corrected value D (k) fluctuated with true value, under the precision of age correction
Drop.
On this point, in the present embodiment, the corrected value D ' (k) after filtering process is added with true value, so can
Realize the age correction of higher precision.
As described above, in the present embodiment, such as shown in above formula (14), the priori that processing unit 50 is handled in Kalman filtering
In the renewal processing (time renewal) of estimate, it is handled as follows:By the posterior estimate x^ (k-1) at the time of last time with
Corrected value D (k-1) addition processing, priori estimates x^ at the time of obtaining this-(k).Moreover, at according to Kalman filtering
The result of reason, carries out the age correction of frequency control data.That is, carry out last time at the time of be time step k-1 posterior estimate x
^ (k-1) is added processing with corrected value D's (k-1), passes through x^-(k)=x^ (k-1)+D (k-1) i.e. times at the time of obtain this
Walk k priori estimates x^-(k)。
Moreover, the result (true value, corrected value) that is handled according to the Kalman filtering of processing unit 50 (age correction portion 56) come
Carry out age correction.That is, time step k corrected value is being set as the frequency after D (k) (or D ' (k)), time step k age correction
In the case that rate control data is AC (k), time step k is obtained by AC (k+1)=AC (k)+D (k) (or AC (k)+D ' (k))
Frequency control data AC (k+1) after+1 age correction.
In addition, processing unit 50 is filtered as shown in above formula (19) according to the corrected value D (k-1) at the time of last time and Kalman
Observation residual error ek in ripple processing, the corrected value D (k) at the time of obtaining this.For example, by carrying out school at the time of to last time
The processing that the value based on observation residual error is Eek (or G (k) ek), school at the time of obtaining this are added on the occasion of D (k-1)
On the occasion of D (k).Specifically, according to last time it is during time step k-1 corrected value D (k-1) and Kalman filtering is handled at the time of
Observe residual error ek, i.e. time step k corrected value D (k) at the time of obtaining this.For example, being E setting observation residual error as ek, constant
In the case of, corrected value D (k) is obtained by D (k)=D (k-1)+Eek.
For example in the present embodiment, as illustrated in fig. 17, the environmental turbulences such as temperature change composition information are obtained
Composition information, and using the environmental turbulence composition information obtained, acquirement is eliminated in environmental turbulence composition and aging variance components
Environmental turbulence composition frequency control data.Here, environmental turbulence composition information can be power supply voltage variation composition, air pressure
Variance components or gravity variance components etc..Then, according to the frequency control data for eliminating environmental turbulence composition, aging is carried out
Correction.Specifically, if environmental turbulence composition is temperature.According to temperature detection data DTD, obtain and believe as environmental turbulence composition
The temperature change composition information of breath, temperature detection data DTD is by from the work for obtaining environmental turbulence composition information
Obtained for the temperature detection voltage VTD of Fig. 9 of environmental turbulence information acquiring section temperature sensor 10.Moreover, using taking
The temperature change composition information obtained, obtains the frequency control data for eliminating temperature change composition.Such as Figure 12 temperature-compensating
Portion 58 obtains temperature compensation data TCODE, and the addition for carrying out temperature compensation data TCODE by adder 65 is handled, and thus, is gone
Except the frequency control data DFCI of temperature change composition is inputted from frequency control data generating unit 40, and taken by processing unit 50
.That is, as shown in Figure 17 E2, the frequency control data for eliminating temperature change composition and remaining aging variance components is obtained
DFCI, and it is input to Kalman filtering portion 54.
In addition, eliminating the frequency control data of environmental turbulence composition except comprising completely eliminated environmental turbulence composition
Outside the frequency control data of appropriate state, there is the environmental turbulence for the degree that can ignore that also in frequency control data
The frequency control data of the state of composition.
For example, can be by detecting the TEMP as environmental turbulence information acquiring section of environmental turbulence composition information
Device, voltage detecting circuit etc. obtain the environmental turbulence such as temperature change composition information or power supply voltage variation composition information composition letter
Breath.On the other hand, aging variance components are the variance components for the frequency of oscillation passed through with the time and changed, it is difficult to pass through sensor
Deng the information for directly detecting the aging variance components.
Therefore, in the present embodiment, obtaining can be become by environment such as the temperature change composition informations of the detections such as sensor
Dynamic composition information, and the environmental turbulence composition information is utilized, acquirement is eliminated in environmental turbulence composition and aging variance components
The frequency control data of environmental turbulence composition.That is, by carrying out removing environmental turbulence from the variance components of frequency control data
The processing (such as addition process of adder 65) of composition, can be obtained as shown in Figure 17 E2 and only remain aging change
The frequency control data of dynamic composition.Then, if carrying out Kalman according to the frequency control data for remaining aging variance components
Filtering process etc., then can estimate the true value of the aging variance components for frequency control data.If moreover, according to so estimating
The true value counted carries out age correction, then can realize the high-precision age correction that can not be realized in the prior embodiment.
So, in the present embodiment, input eliminates temperature change composition (environmental turbulence in Kalman filtering portion 54
Composition) and remain the frequency control data DFCI of aging variance components.And as shown in Figure 1, Figure 10 shows, if during limiting,
Then within this period, it can be assumed that frequency of oscillation is changed with constant rate of ageing.It can be assumed that for example with shown in Figure 10 C3
Constant slope variation.
In the present embodiment, by D (k)=D (k-1)+Eek formula, obtained for compensating (elimination) by this
The corrected value of frequency change caused by the aging variance components of sample under constant rate of ageing.That is, obtained for compensate by
The corrected value D (k) that frequency changes caused by the rate of ageing of the slope of C3 equivalent to Figure 10.Here, rate of ageing is not permanent
Fixed, but as shown in Figure 1, Figure 10 shows, change with the elapsed time.
In this regard, in the present embodiment, as D (k)=D (k-1)+Eek, the observation handled according to Kalman filtering
Residual error ek=y (k)-x^-(k) the renewal processing of corrected value D (k) corresponding with rate of ageing, is carried out.Therefore, it is possible to realize also
Reflect change, corrected value D (k) the renewal processing of rate of ageing corresponding with the elapsed time.It is higher therefore, it is possible to realize
The age correction of precision.
In such as Figure 21, practical frequency deviation and prediction frequency departure are contrasted and shown.D1 is shaking for actual measurement
The frequency departure of frequency is swung, D2 is the frequency for the frequency of oscillation predicted by the estimation processing of the Kalman filtering of present embodiment
Rate deviation.Prediction frequency departure shown in D2 is dropped into range of allowable error, table relative to the practical frequency deviation shown in D1
Show realizes high-precision age correction by present embodiment.
8. temperature sensor, oscillating circuit
Figure 22 shows the configuration example of temperature sensor 10.Figure 22 temperature sensor 10 has current source IST and current collection
Pole is provided to the bipolar transistor TRT from current source IST electric current.Bipolar transistor TRT turns into its colelctor electrode and base stage quilt
The diode connection of connection, temperature detection voltage of the node output with temperature characterisitic of bidirectional bipolar transistor TRT colelctor electrode
VTDI.Temperature detection voltage VTDI temperature characterisitic be due to bipolar transistor TRT emitter-to-base voltage temperature according to
Rely property and produce.The temperature detection voltage VTDI of the temperature sensor 10 for example (has negative ladder with negative temperature characterisitic
1 temperature characterisitic of degree).
Figure 23 shows the configuration example of oscillating circuit 150.The oscillating circuit 150 have current source IBX, bipolar transistor TRX,
Resistance RX, variable capacitance capacitor CX1, capacitor CX2, CX3.
Current source IBX bidirectional bipolar transistors TRX colelctor electrode provides bias current.Resistance RX is arranged at bipolar transistor
Between TRX colelctor electrode and base stage.
The variable capacitance capacitor CX1 of variable capacitance one end is connected with oscillator XTAL one end.Specifically, it is variable
Capacitive battery container CX1 one end is connected to oscillator XTAL via the 1st oscillator terminal (oscillator pad) of circuit arrangement
One end.Capacitor CX2 one end is connected with the oscillator XTAL other end.Specifically, capacitor CX2 one end is via circuit
The 2nd oscillator terminal (oscillator pad) of device and the other end for being connected to oscillator XTAL.Capacitor CX3 one end and oscillator
XTAL one end connection, the other end is connected with bipolar transistor TRX colelctor electrode.
The base emitter interpolar electric current that oscillator XTAL vibration is flowed through in bipolar transistor TRX and is produced.Also,
When base emitter interpolar electric current increases, bipolar transistor TRX colelctor electrode-transmitting electrode current increase, from current source IBX
Reduce to the bias current of resistance RX branches, therefore, collector voltage VCX reductions.On the other hand, when bipolar transistor TRX's
When base emitter interpolar electric current reduces, colelctor electrode-transmitting electrode current reduces, from current source IBX to the biasing of resistance RX branches
Electric current increases, therefore, and collector voltage VCX rises.Collector voltage VCX feeds back to oscillator XTAL via capacitor CX3.
Oscillator XTAL frequency of oscillation has temperature characterisitic, the output voltage VQ that the temperature characterisitic passes through D/A converter sections 80
(frequency control voltage) is compensated.That is, output voltage VQ is input into variable capacitance capacitor CX1, and utilizes output
Voltage VQ is controlled to variable capacitance capacitor CX1 capacitance.Occur in variable capacitance capacitor CX1 capacitance
During change, the resonant frequency of oscillating loop can change, therefore the oscillator XTAL change of frequency of oscillation that causes of temperature characterisitic
It is dynamic to be compensated.Variable capacitance capacitor CX1 can be by such as varicap (varactor:Varactor) etc. it is real
It is existing.
In addition, the oscillating circuit 150 of present embodiment is not limited to Figure 23 structure, various modifications can be implemented.For example in figure
It is illustrated in 23 in case of CX1 is variable capacitance capacitor, but it is also possible to which CX2 or CX3 are set into profit
The variable capacitance capacitor controlled with output voltage VQ.In addition it is also possible to which multiple in CX1~CX3 are set to utilize VQ controls
The variable capacitance capacitor of system.
In addition, oscillating circuit 150 can be without including whole circuit elements for vibrating oscillator XTAL.For example,
Following structure can be used:The circuit element of a part is made up of the discrete part for the outside for being arranged at circuit arrangement 500, and
It is connected via external connection terminals with oscillating circuit 150.
9. variation
Then, the various modifications example of present embodiment is illustrated.Figure 24 shows the circuit arrangement of modified embodiment of the present embodiment
Configuration example.
In fig. 24, it is different from Fig. 2, Fig. 9, D/A converter sections 80 are not provided with oscillator signal generative circuit 140.Also,
The oscillator signal OSCK generated by oscillator signal generative circuit 140 frequency of oscillation is according to the FREQUENCY CONTROL number from processing unit 50
It is directly controlled according to DFCQ.That is, oscillator signal OSCK frequency of oscillation is controlled not via D/A converter sections.
For example in fig. 24, oscillator signal generative circuit 140 has variable capacitance circuit 142 and oscillating circuit 150.At this
Fig. 2, Fig. 9 D/A converter sections 80 are not provided with oscillator signal generative circuit 140.Also, replace Figure 23 variable capacitance electric capacity
Device CX1 and the variable capacitance circuit 142 is set, one end of variable capacitance circuit 142 is connected with oscillator XTAL one end.
The capacitance of the variable capacitance circuit 142 is controlled according to the frequency control data DFCQ from processing unit 50.
For example, variable capacitance circuit 142 has multiple capacitors (array of capacitors), controls each switch according to frequency control data DFCQ
The switched on and off multiple switch element (switch arrays) of element.Each switch element of this multiple switch element and multiple electric capacity
Each capacitor electrical connection of device.Also, by being switched on or switched off in this multiple switch element, multiple capacitors, one end with shaking
The number of the capacitor of sub- XTAL one end connection changes.Thus, the capacitance of variable capacitance circuit 142 is controlled, and is shaken
The capacitance of sub- XTAL one end changes.Therefore, variable capacitance circuit is directly controlled using frequency control data DFCQ
142 capacitance, control oscillator signal OSCK frequency of oscillation.
In addition, in the case where constituting PLL circuit using the circuit arrangement of present embodiment, can also turn into Direct Digital
The PLL circuit of synthesizer mode.Figure 25 direct digital synthesiser mode is shown in the case of circuit structure example.
Phase comparing section 380 (comparing and computing unit) carries out reference signal RFCK and oscillator signal OSCK and (is based on oscillator signal
Input signal) phase bit comparison (comparison operation).Digital filtering portion 382 carries out the smoothing techniques of phase error.Phase ratio
Structure, action compared with portion 380 is identical with Fig. 9 phase comparing section 41, can include counter and TDC (time-to-digital converters
Device).Digital filtering portion 44 of the digital filtering portion 382 equivalent to Fig. 9.Numerical Control type oscillator 384 is to use to come to have to shake
The reference oscillator signal of sub- XTAL reference oscillator 386, the circuit of digit synthesis is carried out to arbitrary frequency and waveform.That is,
It is not to control frequency of oscillation according to the control voltage from D/A converter as VCO, but uses the FREQUENCY CONTROL of numeral
Data and reference oscillator 386 (oscillator XTAL), the oscillator signal for generating arbitrary frequency of oscillation is handled by digital operation
OSCK.By Figure 25 structure, the ADPLL circuits of direct digital synthesiser mode can be realized.
In addition, in the method for present embodiment described above, to by from the variance components of frequency control data
Remove aging variance components obtained by environmental turbulence composition and carry out Kalman filtering processing, estimate the true value removed after deviation.
Moreover, further calculating the corrected value (oscillating characteristic coefficient of alteration) for age correction, carried out using true value and corrected value old
Change correction, hereby it is achieved that making frequency of oscillation be constant control.But, the method for present embodiment is not limited to such side
Method, can implement various modifications.
In the method for the 1st variation of such as present embodiment, it will be removed from measured value after environmental turbulence composition
The value of aging variance components is saved in storage part (memory).Furthermore, it is possible to according to the multiple aging variance components stored
Value and pre-prepd 1 sublinear formula or repeatedly polynomial approximate expression, control frequency of oscillation.
In addition, in the method for the 2nd deformation of present embodiment, such as can be, become removing environment from measured value
The value of aging variance components after dynamic composition is another by switch to what is set as preparation in the case of more than specified value
One oscillator, the influence of aging is suppressed in some constant range.
In addition, the circuit arrangement of present embodiment serves not only as constituting the circuit arrangement of the loop of PLL circuit, it can also make
Circuit arrangement for the oscillator of self-excitation is used.
10. oscillator, electronic equipment, moving body
Figure 26 shows the configuration example of the oscillator 400 of the circuit arrangement 500 comprising present embodiment.As shown in figure 26, shake
Swing device 400 and include oscillator 420 and circuit arrangement 500.Oscillator 420 and circuit arrangement 500 are installed on the encapsulation 410 of oscillator 400
It is interior.Also, the terminal of oscillator 420 and the terminal (pad) of circuit arrangement 500 (IC) are electrically connected using the internal wiring of encapsulation 410
Connect.
Figure 27 shows the configuration example of the electronic equipment of the circuit arrangement 500 comprising present embodiment.The electronic equipment is included
The oscillators such as circuit arrangement 500, the quartz vibrator of present embodiment 420, antenna ANT, communication unit 510 and processing unit 520.In addition,
Operating portion 530, display part 540 and storage part 550 can also be included.Oscillator 400 is constituted by oscillator 420 and circuit arrangement 500.
In addition, electronic equipment is not limited to Figure 27 structure, it is possible to implement omit the structural element or additional other structures of a portion
The various modifications such as key element.
As Figure 27 electronic equipment, such as it can be assumed that network relevant device base station or router, high-precision
Measuring apparatus, GPS onboard clocks, biological information detecting equipment (sphygmometer, pedometer etc.) or head-mount formula display device
It is whole etc. mobile informations such as wearable device, smart mobile phone, mobile phone, portable type game device, notebook PC or tablet PCs
Hold the various equipment such as the image documentation equipments such as (mobile terminal), content providing terminal, digital camera or the video camera of issuing content.
Communication unit 510 (radio-circuit) carries out sending data via antenna ANT from external reception data or to outside
Processing.Processing unit 520 carries out the control process of electronic equipment and to the various numbers for the data received and dispatched via communication unit 510
Word processing etc..The function of the processing unit 520 such as can by microcomputer processor and realize.
Operating portion 530 is used to carry out input operation for user, can be by operation button, touch panel display etc. come real
It is existing.Display part 540 is used to show various information, can be realized by liquid crystal, organic EL etc. display.In addition, using tactile
Touch in the case that panel display is used as operating portion 530, the touch panel display has operating portion 530 and display part concurrently
540 function.Storage part 550 is used for data storage, and its function can (hard disk drives by the semiconductor memories such as RAM, ROM or HDD
Dynamic device) etc. realize.
Figure 28 shows the example of the moving body of the circuit arrangement comprising present embodiment.The circuit arrangement of present embodiment
In (oscillator) various moving bodys such as can be assembled into vehicle, aircraft, motorcycle, bicycle or ship.Mobile style
Have drive mechanism, steering wheel or the Duo Deng steering mechanism such as engine or motor and various electronic equipments (vehicle-mounted to set in this way
It is standby), and on land, aerial or marine mobile device.Figure 28 summary is shown as the concrete example of moving body
Automobile 206.The oscillator (not shown) of circuit arrangement and oscillator with present embodiment is assembled in automobile 206.Control dress
208 bases are put to be acted by the clock signal that the oscillator is generated.Control device 208 according to such as car body 207 posture
Soft durometer to suspension is controlled, or the braking of each wheel 209 is controlled.Control device can for example be utilized
208 realize the automatic operating of automobile 206.In addition, the equipment for the circuit arrangement or oscillator for being assembled with present embodiment is not limited to
This control device 208, can also be assembled in the various equipment (mobile unit) set by the grade moving body of automobile 206.
Figure 29 is the detailed construction example of oscillator 400.Figure 29 oscillator 400 is that double constant temperature slot structures (are inventionbroadly
Constant temperature slot structure) oscillator.
Encapsulation 410 is made up of substrate 411 and housing 412.Various electronic units (not shown) are equipped with substrate 411.
The 2nd container 414 is provided with the inside of housing 412, the 1st container 413 is provided with the inside of the 2nd container 414.Also, in the 1st container
The medial surface (downside) of 413 upper surface is provided with oscillator 420.In addition, the upper surface of the 1st container 413 lateral surface (on
The circuit arrangement 500, heater 450 and temperature sensor 460 of present embodiment are installed sideways).Heater 450 can be passed through
(heater element), adjusts the temperature of such as inside of the 2nd container 414.Further, it is possible to which by temperature sensor 460, detection is for example
The temperature of the inside of 2nd container 414.
2nd container 414 is arranged on substrate 416.Substrate 416 is can to carry the circuit substrate of various electronic units.
The reverse side in face in substrate 416, being provided with the 2nd container 414 installs having heaters 452 and temperature sensor 462.It can pass through
The temperature in such as space between heater 452 (heater element), the adjustment container 414 of housing 412 and the 2nd.Further, it is possible to pass through
The temperature in the space between temperature sensor 462, the detection container 414 of housing 412 and the 2nd.
As the heater element of heater 450,452, for example, heating power bipolar transistor, heating type can be used to add
Hot device MOS transistor, heating resistor, Peltier element etc..The control example of the heating of these heaters 450,452 is if logical
The thermostat of oversampling circuit device 500 controls circuit to realize.As temperature sensor 460,462, for example, it can use temperature-sensitive electricity
Resistance, diode etc..
In Figure 29, due to that can realize that the temperature of the grade of oscillator 420 is adjusted by the thermostat of double constant temperature slot structures, because
This, realizes stabilisation of the frequency of oscillation of oscillator 420 etc..
Figure 30 is the configuration example of the base station (base station apparatus) as one of electronic equipment.Physical layer circuit 600 carry out via
The processing of physical layer in the communication process of network.Network processing unit 602 leaned on than physical layer the processing (link layer of upper layer
Deng).Switch portion 604 carries out the various hand-off process of communication process.DSP 606 carries out the various data signals needed for communication process
Processing.RF circuits 608 are included:The receiving circuit being made up of low-noise amplifier (LNA);The transmission electricity being made up of power amplifier
Road;D/A converter and A/D converter etc..
Selector 612 is by the reference signal RFCK1 from GPS 610, the reference signal from physical layer circuit 600
Any one in RFCK2 (clock signal for carrying out automatic network) is output to the circuit of present embodiment as reference signal RFCK
Device 500.It is synchronous with reference signal RFCK that circuit arrangement 500 enters to exercise oscillator signal (input signal based on oscillator signal)
Processing.And generation different various clock signal CK1, CK2, CK3, CK4, CK5 of frequency, and be supplied to physical layer circuit 600,
Network processing unit 602, switch portion 604, DSP 606, RF circuits 608.
According to the circuit arrangement 500 of present embodiment, in the base station shown in Figure 30, oscillator signal can be made to believe with benchmark
Number RFCK synchronizations, base station is supplied to by the high clock signal CK1~CK5 of the frequency stability generated according to the oscillator signal
Each circuit.
In addition, present embodiment is described in detail as described above, and to those skilled in the art, should be able to
It is readily appreciated that the various deformation of the new item and effect that do not actually detach the present invention.Therefore, such variation is integrally incorporated in
In the scope of the present invention.For example, in specification or accompanying drawing, at least one times from more broad sense or synonymous different terms
The term (temperature change composition etc.) that (environmental turbulence composition etc.) is together described all may be used in the arbitrary portion of specification or accompanying drawing
To be replaced into the different terms.In addition, all combinations of present embodiment and variation are also included within the scope of the present invention.This
Outside, circuit arrangement, oscillator, electronic equipment, the structure of moving body or action, age correction processing, Kalman filtering processing, guarantor
The content that mode treatment, temperature-compensating processing etc. are also not necessarily limited to illustrate in present embodiment is held, various modifications can be implemented.
Claims (14)
1. a kind of circuit arrangement, wherein, the circuit arrangement is included:
Processing unit, its frequency control data to input carries out signal transacting, output frequency control data;And
Oscillator signal generative circuit, it uses oscillator, generates the vibration of the frequency of oscillation set by the frequency control data
Signal,
The processing unit is carried out after passage time step k-1 in the renewal processing for the priori estimates that Kalman filtering is handled
Obtain time step k priori estimates x in the phase Calais for testing estimate x^ (k-1) and corrected value D (k-1)^-(k) processing, and root
The result handled according to the Kalman filtering, carries out the age correction of the frequency control data.
2. circuit arrangement according to claim 1, wherein,
Observation residual error of the processing unit in Kalman filtering processing, obtains the corrected value D (k-1).
3. circuit arrangement according to claim 1, wherein,
The processing unit carries out the phase of the posterior estimate x^ (k-1) and the corrected value D (k-1) of the time step k-1
Plus, pass through x^-(k)=x^ (k-1)+D (k-1), obtains the priori estimates x of the time step k^-(k)。
4. circuit arrangement according to claim 3, wherein,
Observation of the processing unit in the corrected value D (k-1) of the time step k-1 and Kalman filtering processing
Residual error ek, obtains the corrected value D (k) of the time step k.
5. circuit arrangement according to claim 4, wherein,
In the case where setting constant as E, the processing unit obtains the corrected value D (k) by D (k)=D (k-1)+Eek.
6. circuit arrangement according to claim 5, wherein,
The circuit arrangement is also comprising the storage part for storing the constant E.
7. circuit arrangement according to claim 1, wherein,
Frequency control data of the processing unit to the input based on input signal and the phase comparative result of reference signal
The signal transacting is carried out, wherein, the input signal is based on the oscillator signal,
During detecting before the disappearance of the reference signal or abnormal caused holding pattern, located as follows
Reason:Handled by the Kalman filtering, estimation is directed to the true value of the observation of the frequency control data of the input,
In the case where detecting the holding pattern, institute at the time of with detecting corresponding at the time of the holding pattern is preserved
True value is stated, the calculation process based on the true value is carried out, thus, the frequency control data after generation age correction.
8. circuit arrangement according to claim 7, wherein,
The processing unit generates the age correction by carrying out the calculation process to the true value plus the corrected value
Frequency control data afterwards.
9. circuit arrangement according to claim 8, wherein,
The processing unit true value is added the calculation process of the corrected value after filtering process.
10. circuit arrangement according to claim 1, wherein,
The circuit arrangement also includes storage part, and the storage part stores the setting of the system noise of the Kalman filtering processing
The observation noise constant of the setting of system noise constant and the observation noise of Kalman filtering processing.
11. circuit arrangement according to claim 1, wherein,
The circuit arrangement also includes the digital interface portion for being used for monitoring the priori estimates and observation.
12. a kind of oscillator, wherein, the oscillator is included:
Circuit arrangement described in claim 1;And
The oscillator.
13. a kind of electronic equipment, wherein, the electronic equipment includes the circuit arrangement described in claim 1.
14. a kind of moving body, wherein, the moving body includes the circuit arrangement described in claim 1.
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2016001302 | 2016-01-06 | ||
JP2016-001302 | 2016-01-06 | ||
JP2016-137665 | 2016-07-12 | ||
JP2016137665A JP6753181B2 (en) | 2016-01-06 | 2016-07-12 | Circuit devices, oscillators, electronic devices and mobile units |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107040209A true CN107040209A (en) | 2017-08-11 |
CN107040209B CN107040209B (en) | 2022-03-22 |
Family
ID=59305792
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611167368.2A Active CN107040209B (en) | 2016-01-06 | 2016-12-16 | Circuit device, oscillator, electronic apparatus, and moving object |
Country Status (2)
Country | Link |
---|---|
JP (1) | JP6753181B2 (en) |
CN (1) | CN107040209B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2019005192A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005204A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005194A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005189A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005190A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005191A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005201A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005202A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005196A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005193A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005188A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005195A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP2019005203A (en) * | 2017-06-23 | 2019-01-17 | 株式会社三洋物産 | Game machine |
JP7210891B2 (en) * | 2018-03-29 | 2023-01-24 | セイコーエプソン株式会社 | Circuit devices, oscillators, electronic devices and moving bodies |
JP2020010206A (en) | 2018-07-10 | 2020-01-16 | セイコーエプソン株式会社 | Circuit device, oscillator, clock signal generation device, electronic apparatus, and mobile body |
JP6980199B2 (en) * | 2020-02-21 | 2021-12-15 | 株式会社東京測振 | Estimator, vibration sensor system, method performed by the estimator, and program |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5697082A (en) * | 1993-10-01 | 1997-12-09 | Greer; Steven Craig | Self-calibrating frequency standard system |
US6711230B1 (en) * | 2002-09-27 | 2004-03-23 | Nortel Networks Limited | Reference timing signal oscillator with frequency stability |
US20040239560A1 (en) * | 2001-09-28 | 2004-12-02 | Jacques Coatantiec | Hybrid inertial navigation system with improved integrity |
JP2010193337A (en) * | 2009-02-20 | 2010-09-02 | Furuno Electric Co Ltd | Reference signal generating apparatus |
US20110181463A1 (en) * | 2009-12-18 | 2011-07-28 | Thales | Satellite-based positioning receiver |
CN104076373A (en) * | 2013-03-27 | 2014-10-01 | 北京大学 | Receiver carrier wave tracking implementation method and system based on multi-information fusion assistance |
CN104407362A (en) * | 2014-11-03 | 2015-03-11 | 中国人民解放军63961部队 | Carrier wave phase-locked loop based on four-path signal processing |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7915962B2 (en) * | 2009-07-06 | 2011-03-29 | Nortel Networks Limited | System and method for built in self test for timing module holdover |
-
2016
- 2016-07-12 JP JP2016137665A patent/JP6753181B2/en active Active
- 2016-12-16 CN CN201611167368.2A patent/CN107040209B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5697082A (en) * | 1993-10-01 | 1997-12-09 | Greer; Steven Craig | Self-calibrating frequency standard system |
US20040239560A1 (en) * | 2001-09-28 | 2004-12-02 | Jacques Coatantiec | Hybrid inertial navigation system with improved integrity |
US6711230B1 (en) * | 2002-09-27 | 2004-03-23 | Nortel Networks Limited | Reference timing signal oscillator with frequency stability |
JP2010193337A (en) * | 2009-02-20 | 2010-09-02 | Furuno Electric Co Ltd | Reference signal generating apparatus |
US20110181463A1 (en) * | 2009-12-18 | 2011-07-28 | Thales | Satellite-based positioning receiver |
CN104076373A (en) * | 2013-03-27 | 2014-10-01 | 北京大学 | Receiver carrier wave tracking implementation method and system based on multi-information fusion assistance |
CN104407362A (en) * | 2014-11-03 | 2015-03-11 | 中国人民解放军63961部队 | Carrier wave phase-locked loop based on four-path signal processing |
Also Published As
Publication number | Publication date |
---|---|
JP2017123630A (en) | 2017-07-13 |
JP6753181B2 (en) | 2020-09-09 |
CN107040209B (en) | 2022-03-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107040209A (en) | Circuit arrangement, oscillator, electronic equipment and moving body | |
JP6680121B2 (en) | Circuit devices, oscillators, electronic devices and mobile units | |
CN106953633A (en) | Circuit arrangement, oscillator, electronic equipment and moving body | |
CN106953631A (en) | Circuit arrangement, oscillator, electronic equipment and moving body | |
JP6766427B2 (en) | Circuits, oscillators, electronics and mobiles | |
US10177770B2 (en) | Circuit device, oscillator, electronic apparatus, and vehicle | |
US10084462B2 (en) | Circuit device, oscillator, electronic apparatus, and vehicle | |
US10171094B2 (en) | High accuracy clock synchronization circuit | |
US10348309B2 (en) | Circuit device, oscillator, electronic apparatus, and vehicle | |
CN106817080A (en) | The manufacture method of circuit arrangement, oscillator, electronic equipment, moving body and oscillator | |
CN106953597A (en) | Circuit arrangement, oscillator, electronic equipment and moving body | |
US10298175B2 (en) | Circuit device, oscillator, electronic apparatus, and vehicle | |
US10305488B2 (en) | Circuit device, oscillator, electronic apparatus, and vehicle | |
US10305489B2 (en) | Circuit device, oscillator, electronic apparatus, and vehicle | |
CN106961251A (en) | Circuit arrangement, oscillator, electronic equipment and moving body |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |