CN101414822A - All-digital phase-locked loop, loop bandwidth calibration method, and loop gain calibration method for the same - Google Patents

Abstract

For decreasing errors within an analog phase-locked loop, an all-digital phase-locked loop (ADPLL) with only digital components and digital operations is used. The ADPLL may also be used for direct frequency modulation (DFM). By modulating certain parameters within the ADPLL by following an all-pass frequency response, a loop gain of the ADPLL may be precisely modulated, and an available bandwidth of the ADPLL is also significantly broadened.

Description

All-digital phase-locked loop, loop bandwidth calibration method and loop gain calibration steps

Technical field

The invention relates to a kind of all-digital phase-locked loop and a kind of loop bandwidth calibration method and loop gain calibration steps that is used for all-digital phase-locked loop.

Background technology

Phase-locked loop is a kind of electronic control system that the signal of fixed relationship is arranged with the phase place (Phase) of reference signal that is used for producing.Phase-locked loop circuit is in response to the frequency and the phase place of input signal, and improves or reduce the frequency of controlled oscillator automatically, until phase-locked loop circuit and reference signal till being consistent on frequency and the phase place.The prior art analog phase-locked look comprise phase detectors, voltage controlled oscillator (Voltage-Controlled Oscillator, VCO), and feedback path.Feedback path is used for the output signal of voltage controlled oscillator is fed back to the input of phase detectors, to improve or to reduce the frequency of the input signal of analog phase-locked look.Therefore, the frequency of analog phase-locked look always can keep catching up with the reference frequency of reference signal, and wherein reference signal is used by phase detectors, that is to say, the frequency of the input signal of analog phase-locked look is always locked by the reference frequency of reference signal.In addition, in the prior art, frequency divider (Frequency divider) is used for feedback path, so that the integer multiple frequency of reference frequency or reference frequency always can be captured.In the prior art, low pass filter (Low-pass filter) is connected in after the phase detectors, is able to filtering so that be positioned at high-frequency noise.

As known to persons skilled in the art dawn, because analog phase-locked look is used simulated assembly, and use the analog form operation, above-mentioned analog phase-locked look very easily produces error, even be error propagation (Errorpropagation).Therefore, digital phase-locked loop is just arisen at the historic moment, and to reduce above-mentioned error under the support of part digit manipulation and digital assembly, wherein digital phase-locked loop is used the frequency divider with variable divisor number on feedback path.In addition, all-digital phase-locked loop also helps chip area to reduce and the processing procedure migration very much.For instance, (Digital-Controlled Oscillator DCO) can be used to replace the voltage controlled oscillator of the employed simulated assembly of prior art to the digital controlled oscillator of all-digital phase-locked loop.Also can (Time-to-Digital Converter TDC) replaces with the time-to-digit converter of all-digital phase-locked loop with phase detectors.Therefore, in wireless communication field, using all-digital phase-locked loop has been a kind of trend.

Summary of the invention

Very easily produce error for solving above-mentioned analog phase-locked look, even be the problem of error propagation, the loop bandwidth calibration method and the loop gain calibration steps that the invention provides a kind of all-digital phase-locked loop and be used for all-digital phase-locked loop can be eliminated the error of analog phase-locked look by digit manipulation and digital assembly.

The present invention discloses a kind of loop bandwidth calibration method that is used for all-digital phase-locked loop.The loop bandwidth calibration method comprises according to the two combination of the gain of the gain of the gain of the gain of the time-to-digit converter of all-digital phase-locked loop, digital controlled oscillator or time-to-digit converter and this digital controlled oscillator, the Amplifier Gain of the proportion expression path module by adjusting all-digital phase-locked loop is calibrated the loop frequency range of this all-digital phase-locked loop.

The present invention discloses a kind of all-digital phase-locked loop.All-digital phase-locked loop comprises digital loop filters.Digital loop filters comprises the proportion expression path module.The proportion expression path module be used for following the trail of with from the relevant phase change of the output signal of time-to-digit converter module.The proportion expression path module comprises proportion expression path module amplifier.Proportion expression path module Amplifier Gain is adjusted according to the gain of the time-to-digit converter of time-to-digit converter module.

The present invention discloses a kind of loop gain calibration steps that is used for all-digital phase-locked loop.The loop gain calibration steps comprises the frequency response of digital controlled oscillator of the digital controlled oscillator of reference frequency, all-digital phase-locked loop of sign indicating number variable quantity according to the half period in the reference cycle of the reference signal that is received corresponding to the time-to-digit converter module of all-digital phase-locked loop, reference signal that the time-to-digit converter module is received and ∑ Delta modulator module or the combination of above-mentioned each condition, and the ∑ Delta modulator compensating module Amplifier Gain of ∑ Delta modulator compensating module is modulated; And according to the frequency variation of the input of the ∑ Delta modulator of the feedback path module of all-digital phase-locked loop, corresponding to the sign indicating number variable quantity of frequency variation, divide the reference frequency of this reference signal that digital variable quantity, time-to-digit converter module received or the combination of above-mentioned each condition, the gain of the modulator amplifier of modulator is modulated.

Above-mentioned all-digital phase-locked loop, loop bandwidth calibration method and loop gain calibration steps come resize ratio formula path module Amplifier Gain by the gain according to the time-to-digit converter of the time-to-digit converter module in the all-digital phase-locked loop, to be used for the calibration of loop frequency range; Gain according to the time-to-digit converter of the time-to-digit converter module in the all-digital phase-locked loop comes the all-digital phase-locked loop of resize ratio formula path module Amplifier Gain; Reach the reference frequency of the reference signal that is received according to the time-to-digit converter module, come the gain of the modulator amplifier of the ∑ Delta modulator compensating module Amplifier Gain of ∑ Delta modulator compensating module and modulator is modulated, to be used for the loop gain calibration, thereby by digit manipulation and digital assembly, avoid the error in the analog phase-locked look, and reached the effect of accurate alignment loop gain and loop frequency range.

Description of drawings

Fig. 1 is the schematic diagram of the disclosed all-digital phase-locked loop of the present invention.

Fig. 2 is the schematic diagram of the all-digital phase-locked loop of direct frequency modulated among the present invention.

Fig. 3 be among Fig. 1 and Fig. 2 illustrated digital controlled oscillator at the disclosed detailed maps of the present invention.

Fig. 4 follows the trail of the schematic diagram of unit that groove comprises for prior art.

Fig. 5 is the associated voltage-frequency inverted curve synoptic diagram of unit shown in Figure 4.

The detailed maps of the unit that Fig. 6 is comprised for tracking groove shown in Figure 3.

Fig. 7 is the relevant folding transformation curve schematic diagram of voltage-frequency in unit shown in Figure 6.

Fig. 8 is in order to explain the digital loop bandwidth calibration method of the present invention at all-digital phase-locked loop shown in Figure 1, the rough schematic view of employed all-digital phase-locked loop.

Fig. 9 is the rough schematic that is used for explaining the fractional phase error that how to compensate the prior art analog phase-locked look.

Figure 10 is by being disclosed the schematic diagram of the digital phase error cancellation module that comprises in addition in the ∑ Delta modulator compensating module according to an embodiment of the present invention.

Figure 11 when implementing loop gain calibration steps shown in Figure 8, phase-frequency detector shown in Figure 1 and circulating time-to-digit converter module and time figure switch decoders shown in Figure 1 and the simple and easy schematic diagram of first adder.

Figure 12 is the generalized schematic of circulating time-to-digit converter shown in Figure 11.

Figure 13 is relevant to the schematic flow sheet of the circulating time-to-digit converter calibration procedure of Figure 11 and Figure 12 for enforcement.

Embodiment

The present invention discloses a kind of all-digital phase-locked loop that is used for direct frequency modulated and has precise gain calibration (Fine gaincalibration), and wherein all-digital phase-locked loop uses some assembly that is disclosed side of the present invention (digital controlled oscillator that side for example of the present invention discloses) and technical characterictic.By the disclosed all-digital phase-locked loop of the present invention, switching noise can be reduced significantly, and the loop gain of all-digital phase-locked loop also can be finely tuned accurately.By the disclosed digital controlled oscillator of the present invention, can in disclosed all-digital phase-locked loop, reach accurate frequency resolution.

See also Fig. 1, it is the schematic diagram of the disclosed all-digital phase-locked loop 100 of an embodiment of the present invention.As shown in Figure 1, all-digital phase-locked loop 100 comprises time-to-digit converter (Time-to-Digitalconverter, TDC) module 102, digital macroblock (Digital macro module) 120, digital controlled oscillator and ∑ Delta modulator (Sigma-Delta Modulator, SDM) module 110, and feedback path module 112.

Time-to-digit converter module 102 comprises phase-frequency detector (Phase-Frequency Detector, PFD) and circulating time-to-digit converter (Cyclic Time-to-Digital Converter, CTDC) module 1021 and time-to-digit converter state machine (TDC state machine) 1023.Though circulating time-to-digit converter is applied to the present invention's disclosed each execution mode afterwards, in other execution mode of the present invention, still can use the time-to-digit converter of any other kind to replace circulating time-to-digit converter.

Numeral macroblock 120 comprises time figure switch decoders 1022, first adder 104, proportion expression path (Proportional path) module 106, wave digital lowpass filter (Digital low passfilter) 108, second adder 105 and ∑ Delta modulator compensating module 114.Proportion expression path module 106 comprises infinite impulse response (Infinite Impulse Response, IIR) module 1061 and proportion expression path module amplifier (PPM amplifier) 1062.The gain that note that proportion expression path module amplifier 1062 is assumed to be a at this.Wave digital lowpass filter 108 is used for being used as the path of integration (Integral path) in the all-digital phase-locked loop 100.Proportion expression path module 106 can be regarded as digital loop filters with wave digital lowpass filter 108 combining of the two.∑ Delta modulator compensating module 114 comprises first accumulator (Accumulator) 1141, has ∑ Delta modulator compensating module amplifier (Sigma-deltamodulator compensation module amplifier) the 1142 and the 3rd adder 1143 of gain b.Note that ∑ Delta modulator compensating module 114 also can be considered the error compensation module at this.

Digital controlled oscillator and ∑ Delta modulator module 110 comprise numerical control vibration decoder 1101, the first ∑ Delta modulator 1102, ∑ Delta modulator filter 1103, digital controlled oscillator 1104 and first frequency divider 1105.Please note, though in Fig. 1, first frequency divider, 1105 employed divisors are 4, in other execution mode of the present invention, first frequency divider 1105 also can use other numerical value beyond 4 to be used as its divisor, that is to say that first frequency divider, 1105 employed divisors are not limited to the employed numerical value 4 of Fig. 1.Feedback path module 112 comprises the second ∑ Delta modulator 1121 and second frequency divider 1122.Note that as shown in Figure 1 second frequency divider, 1122 employed divisors are assumed to be M, and M is a variable.Wherein, numerical control vibration decoder 1101, digital controlled oscillator 1104 can be regarded as the digital controlled oscillator module with combining of first frequency divider 1105, to be used for the integer signal of track digital loop filter.

As shown in Figure 1, time-to-digit converter module 102 receives reference signal REF and feedback signal FB, and produces cycle signal C and data-signal D.Cycle signal C all comprises phase information and the frequency information relevant with feedback signal FB with data-signal D.Note that cycle signal C points out the present employed circulation of circulating time-to-digit converter in phase-frequency detector and the circulating time-to-digit converter module 1021.Note that data-signal D points out the data that a plurality of d type flip flops (D flip-flop) in phase-frequency detector and the circulating time-to-digit converter module 1021 are produced.Please note, cycle signal C and data-signal D can be decoded by time figure switch decoders 1022 subsequently, in digital macroblock 120, to produce output signal TDC, wherein output signal TDC also comprises phase information and the frequency information relevant with feedback signal FB, and output signal TDC is also referred to as decoded output signal.First adder 104 is reduced to output signal TDC and error signal Err addition to a certain degree with the error that will may comprise among the output signal TDC, and wherein error signal Err is essentially error compensating signal.First adder 104 also exports signal X to proportion expression path module 106 and wave digital lowpass filter 108.Please note, phase-frequency detector and circulating time-to-digit converter module 1021 are produced, and test oneself signal Bbcomp and symbol (Sign) signal L also added up, to carry the information that indication about whether is improved the output signal frequency of digital controlled oscillator and ∑ Delta modulator module 110 or reduce.Note that phase-frequency detector and circulating time-to-digit converter module 1021 go back clock signal dlyfbclk, operate with built-in clock (built-in clock) to digital macroblock 120.Time-to-digit converter state machine 1023 also produces divisor-signal (divider signal) Div, so that the information relevant with divisor is sent to digital macroblock 120.

Proportion expression path module 106 is used for the variation of phase place of trace signals X; And wave digital lowpass filter 108 (being above-mentioned path of integration) is used for the long-term frequency drift (Long-term frequencydrift) of trace signals X.Numeral macroblock 120 exports integer signal (Integer signal) Integ and fractional signal (Fractionalsignal) Frac to digital controlled oscillator and ∑ Delta modulator module 110.

In digital controlled oscillator and ∑ Delta modulator module 110, the first input end of numerical control vibration decoder 1101 receives integer signal Integ; The first input end of the first ∑ Delta modulator 1102 receives fractional signal Frac; The input of ∑ Delta modulator filter 1103 is coupled to the output of the first ∑ Delta modulator 1102, and in an embodiment of the present invention, ∑ Delta modulator filter 1103 receives the ∑ Δ modulation signal SDM of the first ∑ Delta modulator, 1102 outputs; The first input end of digital controlled oscillator 1104 is coupled to the output of numerical control vibration decoder 1101, and second input of digital controlled oscillator 1104 is coupled to the output of ∑ Delta modulator filter 1103; And the input of first frequency divider 1105 is coupled to the output of digital controlled oscillator 1104, and the output of first frequency divider 1105 is coupled to second input of numerical control vibration decoder 1101 and second input of the first ∑ Delta modulator 1102.Note that first loop process numerical control vibration decoder 1101, digital controlled oscillator 1104, reach first frequency divider 1105.First loop is used for integer signal Integ is adjusted or modulates.Second loop is through the first ∑ Delta modulator 1102, ∑ Delta modulator filter 1103, digital controlled oscillator 1104 and first frequency divider 1105.Second loop is used for fractional signal Frac is adjusted or modulates.

Feedback path module 112 and ∑ Delta modulator compensating module 114 operate together, wherein ∑ Delta modulator compensating module 114 is contained in the digital macroblock 120.Second frequency divider 1122 is used for the signal that digital controlled oscillator and ∑ Delta modulator module 110 are exported is carried out frequency division.Second frequency divider 1122 and the second ∑ Delta modulator, 1121 operate together.∑ Delta modulator compensating module 114 is used for predicting the error that may comprise in the signal that digital controlled oscillator and ∑ Delta modulator module 110 exported.∑ Delta modulator compensating module 114 also is used in feedforward (Feed-forward) mode the error of above-mentioned prediction being inputed to first adder 104, wherein above-mentioned error compensating signal comprises the error of prediction, thus, output signal TDC with error just can significantly be reduced.In an embodiment of the present invention, the error of prediction is by 1142 outputs of ∑ Delta modulator compensating module amplifier.Please note, the positive input terminal of the 3rd adder 1143 is coupled to the input of the second ∑ Delta modulator 1121, the negative input end of the 3rd adder 1143 is coupled to the output of the second ∑ Delta modulator 1121, and the output of the 3rd adder 1143 is coupled to the input of first accumulator 1141.

Because proportion expression path module 106, wave digital lowpass filter 108, with ∑ Delta modulator compensating module 114 all with the fine setting height correlation of the loop gain of all-digital phase-locked loop 100, so the feature of the structure of all-digital phase-locked loop 100 mainly is the existence of said modules.Yet, each assembly that above-mentioned all-digital phase-locked loop 100 is comprised, module, be all numeral, so all-digital phase-locked loop 100 is to operate under complete numerically controlled prerequisite with signal.By all-digital phase-locked loop 100 complete numerically controlled mechanism, can reach frequency range control accurately.All-digital phase-locked loop 100 also can effectively reduce switching noise, and relevant detailed technology can after exposure separately.

The main application of all-digital phase-locked loop 100 is for realizing the digital framework of direct frequency modulated.See also Fig. 2, it is the schematic diagram of the all-digital phase-locked loop 200 of direct frequency modulated among the present invention, and it is designed that wherein all-digital phase-locked loop 200 is based on all-digital phase-locked loop shown in Figure 1 100.As shown in Figure 2, except each assembly that all-digital phase-locked loop 100 is comprised, all-digital phase-locked loop 200 comprises the second accumulator (Accumulator in addition, ACC) 202, accumulator amplifier (ACC amplifier) 204 and modulated amplifier (Modulator amplifier) 206, above-mentioned second accumulator 202, accumulator amplifier 204 can be regarded as modulator with combining of modulated amplifier 206.Accumulator amplifier 204 and second accumulator, 202 operate together, and the gain of accumulator amplifier 204 is gain b, ∑ Delta modulator compensating module amplifier 1142 employed gains just.Modulated amplifier 206 employed gains are assumed to be gain c.The message MSG that is actually modulation signal is input to second accumulator 202 and modulated amplifier 206, with after with feed-forward mode feed-in first adder 104 and second adder 105.Note that for message MSG second accumulator 202 can be regarded as high pass filter (High-pass filter) with the combination of accumulator amplifier 204.Note that the lowpass response that digital controlled oscillator and ∑ Delta modulator module 110 also provide message MSG, wherein the voltage controlled oscillator in the prior art phase-locked loop can give the upper frequency limit of the frequency domain of message MSG; That is to say that for message MSG, voltage controlled oscillator is a low pass filter, make the frequency domain of message MSG be low pass filtering device and limit.By making up above-mentioned high pass response and lowpass response, can obtain all-pass response (All-pass response), make wide-band modulation (Wide bandmodulation) be achieved, or the frequency range that makes the frequency range of message MSG no longer be subjected to phase-locked loop limit or restrain.For above-mentioned all-pass response being operated above-mentioned gain b of adjustment that must be accurate and gain c.Note that because by the all-pass response, the frequency domain of message MSG is not restricted again or is relevant with all-digital phase-locked loop 200, so above-mentioned wide-band modulation is achieved.Use the technology of predistortion (Pre-distortion) in the phase-locked loop of prior art, making noise distortion in advance, yet the assembly of implementing pre-distortion technology can occupy bigger chip area.Avoided using this kind pre-distortion technology at the disclosed all-digital phase-locked loop 200 of the present invention.

Correcting gain b discloses as follows with the technology of the value of gain c among the present invention.See also Fig. 2, the loop gain of all-digital phase-locked loop 200 can respond m[n by using the input among the message MSG] obtain, to obtain corresponding output frequency response V _Out[n] is with the output response as digital controlled oscillator and ∑ Delta modulator module 110.The loop gain of all-digital phase-locked loop 200 with

Represent, and when all-digital phase-locked loop 200 was modulated with the all-pass responsive state, the response of this loop gain can be expressed as follows:

$\frac{V_{\mathrm{out}}\left[n\right]}{m\left[n\right]}=\frac{c\&CenterDot;\mathrm{Kv}+b\&CenterDot;L\left(z\right)\&CenterDot;\mathrm{<figure-callout\; id="1"\; label="Kv"\; filenames="A200810169984E00361.png,A200810169984E00381.png"\; state="\{\{state\}\}">Kv</figure-callout>}}{1+\frac{1}{\mathrm{TDC}\&CenterDot;\mathrm{<figure-callout\; id="2"\; label="F\; ref"\; filenames="A200810169984E00401.png"\; state="\{\{state\}\}">F</figure-callout>}{\mathrm{ref}}^{}{}^2}\&CenterDot;L\left(z\right)\&CenterDot;\mathrm{Kv}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}}=1---\left(1\right).$

The employed partial condition simplicity of explanation of equation (1) is as follows.CKv represents to comprise the response in modulated amplifier 206 and the path of digital controlled oscillator and ∑ Delta modulator module 110, and wherein Kv is the gain of digital controlled oscillator and ∑ Delta modulator module 110, that is to say that Kv is the gain of digital controlled oscillator 1104.Item bL (z) Kv represents combination, the wave digital lowpass filter 108 that comprises second accumulator 202 and the amplifier 204 that adds up, the response that reaches the path of digital controlled oscillator and ∑ Delta modulator module 110, and wherein the response of wave digital lowpass filter 108 is assumed to be L (z).

For representing the gain of time-to-digit converter module 102, wherein Fref is the reference frequency of reference signal REF, and TDC is meant the gain of the circulating time-to-digit converter that phase-frequency detector and circulating time-to-digit converter module 1021 are comprised.

It is the response of second frequency divider 1122.

Be meant the frequency response of digital controlled oscillator 1104.

Observation equation formula (1) as can be known, in order to satisfy above-mentioned all-pass responsive state, gain b needs according to following two equations decision with the value of gain c:

c·Kv＝1 (2)；

$b\&CenterDot;L\left(z\right)\&CenterDot;\mathrm{Kv}=\frac{1}{\mathrm{TDC}\&CenterDot;{\mathrm{Tref}}^2}\&CenterDot;L\left(z\right)\&CenterDot;\mathrm{Kv}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}---\left(3\right).$

After equation (2) and the further derivation of (3) do, the value of gain b and c can be expressed as follows:

$c=\frac{1}{\mathrm{Kv}}---\left(4\right)$

$b=\frac{1}{\mathrm{TDC}\&CenterDot;{\mathrm{Tref}}^2}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}---\left(5\right)$

In order to reach the purpose of digital controlling mechanism, it is necessary that gain b is carried out complete operation with the value of gain c.Observation equation formula (5) as can be known, for to the gain b value operate, the value of the gain TDC of circulating time-to-digit converter also needs for controlled.The gain TDC of circulating time-to-digit converter may be defined as the resolution of time-to-digit converter module 102, that is to say, the gain TDC can be expressed as the merchant of time variation amount Δ t divided by sign indicating number variation delta N, make circulating time-to-digit converter gain TDC value can by the decision as follows:

$\mathrm{TDC}=\frac{\mathrm{\&Delta;t}}{\mathrm{\&Delta;N}}=\frac{\frac{1}{2}\mathrm{<figure-callout\; id="1"\; label="Tref\; N"\; filenames="A200810169984E00361.png,A200810169984E00381.png"\; state="\{\{state\}\}">Tref</figure-callout>}}{N_{}}=\frac{1}{2\mathrm{Fre}\&CenterDot;N_1}---\left(6\right).$

Sign indicating number variable quantity N wherein ₁Half period corresponding to the reference cycle Tref of reference signal that is to say, sign indicating number variable quantity N ₁Be the sign indicating number variable quantity in the half period of reference cycle Tref, and in single reference cycle Tref, positive status and negative state occupy the Cycle Length of half in turn.In an embodiment of the present invention, sign indicating number variable quantity N ₁Produce according to the reference signal that receives by the time-to-digit converter module.According to equation (6), the derivation of the value of gain b can be rewritten as follows:

$b=\frac{1}{\frac{1}{2\&CenterDot;\mathrm{Fref}\&CenterDot;{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="A200810169984E00361.png,A200810169984E00381.png"\; state="\{\{state\}\}">N</figure-callout>}}_1}\&CenterDot;{\mathrm{Fref}}^2}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}=\frac{{2N}_1}{M}\&CenterDot;\frac{1}{\mathrm{Fref}}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}---\left(7\right).$

Observation equation formula (4) as can be known, for to the gain c value operate, the value of the gain Kv of digital controlled oscillator need be for controlled.Equation (4) can be derived as follows separately:

$c=\frac{1}{\mathrm{Kv}}=\frac{\mathrm{\&Delta;I}}{\mathrm{\&Delta;N}\&CenterDot;\mathrm{Fref}}---\left(8\right);$

Its discipline Δ NFref represents the frequency variation of the signal that the input of the second ∑ Delta modulator 1121 imports, and the sign indicating number variation delta I corresponding to frequency variation Δ NFref can obtain in the output signal of wave digital lowpass filter 108, wherein Δ N is meant the sign indicating number variable quantity of branch number (Fractional code), and promptly Δ N is meant the digital variable quantity of branch.In an embodiment of the present invention, wave digital lowpass filter 108 is to come output code variation delta I according to a minute digital variation delta N.Because it is controlled that a frequency variation Δ NFref and a sign indicating number variation delta I are all, so the value of gain c also should be controlled.According to above-mentioned exposure, can realize accurate calibration to the loop gain of all-digital phase-locked loop 200.

Digital controlled oscillator 1104 is used for the frequency band of following the trail of output signal according to the integer signal in the output signal in the digital macroblock 120 and fractional signal.The integer signal is decoded by numerical control vibration decoder 1101, and fractional signal is to handle by ∑ Delta modulator 1102 and 1103 runnings of ∑ Delta modulator filter.The running of ∑ Delta modulator 1102 and ∑ Delta modulator filter 1103 is similar in appearance to the ∑ Delta modulator and the ∑ Delta modulator filter of prior art, so relevant running is given unnecessary details no longer separately.Use the digital controlled oscillator of prior art also to belong to covering scope of the present invention in the present invention, but digital controlled oscillator 1104 is what be designed and propose especially In some embodiments of the present invention, being used for realizing that frequency band follows the trail of, and be used for avoiding conspicuous frequency discontinuous (Frequency discontinuity).

See also Fig. 3, its be among Fig. 1 and Fig. 2 illustrated digital controlled oscillator 1104 at the disclosed detailed maps of the present invention.Digital controlled oscillator 1104 comprises brilliant (On-chip) low dropout voltage regulator (Low-drop-out regulator of carrying, LDO regulator) 302, inductance and resistive module 304, process voltage temperature groove (Process/Voltage/Temperature tank, PVT tank) 306, gather groove (Acquisitiontank) 308, and follow the trail of groove (Tracking tank) 310.If target application allows, then low dropout voltage regulator 302 can be excluded from outside the digital controlled oscillator 1104.Inductance and resistive module 304 are coupled to low dropout voltage regulator 302.Process voltage temperature groove 306 is coupled to inductance and resistive module 304.Gather groove 308 and be coupled to process voltage temperature groove 306.Follow the trail of groove 310 and be coupled to collection groove 308.In the said modules, except following the trail of groove 310, all can be implemented by the corresponding assembly of prior art, therefore only said modules is briefly described as follows.The brilliant low dropout voltage regulator 302 that carries is used for producing the required voltage VCCreg that is used for digital controlled oscillator 1104 according to main voltage VCC.Inductance and resistive module 304 comprise a plurality of inductance, a plurality of changeable resistance 3033,3034 and negative trnasducing element (Negative gm cell, wherein gm is BJT or the employed transduction of MOS transistor (Transconductance) parameter) 3042.Inductance and resistive module 304 are used for setting the current drain and the oscillation amplitude of digital controlled oscillator 1104, inject phenomenon (common-mode injection) to be used for improving common mode, and reduce the noise that earth terminal produced and the surging (Spur) of digital controlled oscillator 1104.Process voltage temperature groove 306 is used for compensate for process, voltage, variation of temperature.Gathering groove 308 is used to provide fast frequency and obtains (Frequency acquisition).

The principal character of digital controlled oscillator 1104 is to follow the trail of groove 310.Before in detail disclosing the details of following the trail of groove 310, must introduce the tracking groove that prior art is used, with the advantage of interpretive tracing groove 310 more in advance.See also Fig. 4, Fig. 5, Fig. 6 and Fig. 7.Fig. 4 follows the trail of the schematic diagram of the unit 400 of groove for prior art.Fig. 5 is the associated voltage-frequency inverted curve synoptic diagram of unit 400 shown in Figure 4.Fig. 6 is the detailed maps of the unit 600 of tracking groove 310 shown in Figure 3.Fig. 7 is that unit 600 relevant voltage-frequencies shown in Figure 6 fold the transformation curve schematic diagram.

As shown in Figure 4, the unit 400 of prior art tracking groove comprises reverser (Inverter) 402, the first P-type mos field-effect transistor (P-type MOSFET) 404, the one N type metal oxide semiconductor field-effect transistor 406, the second P-type mos field-effect transistor 408, the 2nd N type metal oxide semiconductor field-effect transistor 410, the 3rd N type metal oxide semiconductor field-effect transistor 412, the 4th N type metal oxide semiconductor field-effect transistor 414, first electric capacity 416, second electric capacity 418, first resistance 420, and second resistance 422.The coupling mode of said modules has been illustrated in Fig. 4, so locate to give unnecessary details no longer in detail.Voltage VCCreg is imported first resistance 420 and second resistance 422.The input of one bit is comprised the set of the first P-type mos field-effect transistor 404 and a N type metal oxide semiconductor field-effect transistor 406, wherein this bit can be odd number or even number, to be used to refer to the digital integer signal from numerical control vibration decoder 1101.In an embodiment of the present invention, this bit is relevant to the combination of integer signal, fractional signal or integer signal and fractional signal.Also the fractional signal input is comprised the set of the second P-type mos field-effect transistor 408 and the 2nd N type metal oxide semiconductor field-effect transistor 410, wherein this fractional signal also can be regarded as main voltage (Primary voltage), and main voltage can receive from ∑ Δ low pass filter.In an embodiment of the present invention, fractional signal is a ∑ Delta modulator fractional signal.In another execution mode of the present invention, fractional signal is the signal from ∑ Δ low pass filter, i.e. ∑ Δ low pass filter signal.The voltage that will comprise high-level output voltage (being high level voltage) Vo+ and low-level output voltage (being low level voltage) Vo-is to output, follows the trail of the vibration in the groove to be used for representing above-mentioned prior art.Briefly, whenever the value of relative integers signal was coupled with 1 o'clock, the value of ∑ Delta modulator fractional signal is reduced 1, makes the mean value of ∑ Delta modulator fractional signal be maintained at below 1, or even near 0.Yet because the bit that is transfused to continues between 0 and 1 to change, whenever the value of integer signal is increased at once at 1 o'clock, the value of ∑ Delta modulator fractional signal reduces by 1 running speed and can't catch up with the value of integer signal and increase by 1 speed.Therefore, as shown in Figure 5, V _{The Δ ∑}Expression voltage, when the value of integer signal increases to (N+1) by N,, the value of ∑ Delta modulator fractional signal can't be adjusted (or correspondingly reducing) before timely because of being increased to (N+1) in the value of integer signal to desired value Targ, so can the occurrence frequency non-continuous event.Transformation curve when Fig. 5 gives the integer signal and is N-1 and N+2.

The unit 600 that tracking groove 310 is comprised is disclosed at this, to solve the said frequencies non-continuous event.Unit 600 is separated to two different set with the strange bit and the running of even bit, just strange bit set and even bit set, make the voltage-frequency transformation curve that is illustrated in Fig. 7 not producing frequency agility, be to present folding shape under the discontinuous situation of frequency, just represent the value at the integer signal to reach (N+1) afterwards, fractional signal arrives the program that desired value Targ is carried out.

As shown in Figure 6, unit 600 comprises the first tracking set and second and follows the trail of set, and wherein the first tracking set is used for handling strange bit, and the second tracking set is used for handling even bit.Note that in other execution mode of the present invention first follows the trail of set also can be used to handle even bit, and the second tracking set also can be used to handle strange bit simultaneously.First follows the trail of set comprises first reverser 602, first digital module 603, first analog module 605 and first capacitance module 611.First digital module 603 is used for handling the odd bits bit of being exported by numerical control vibration decoder 1101 (odd bits signal).First analog module 605 is used for handling the ∑ Delta modulator fractional signal that ∑ Delta modulator filter 1103 is exported.First capacitance module 611 is used to provide required capacitance and gives high-level output voltage Vo+ and low-level output voltage Vo-.First digital module 603 comprises the first P-type mos field-effect transistor 604 and a N type metal oxide semiconductor field-effect transistor 606.First analog module 605 comprises the second P-type mos field-effect transistor 608 and the 2nd N type metal oxide semiconductor field-effect transistor 610.First capacitance module 611 comprises the 3rd N type metal oxide semiconductor field-effect transistor 612 and the 4th N type metal oxide semiconductor field-effect transistor 614.First follows the trail of set comprises first electric capacity 616, second electric capacity 618, first resistance 620 and second resistance 622 in addition.Note that first digital module 603, first analog module 605, and the assembly that comprised of first capacitance module 611 or form and in other execution mode of the present invention, be not subjected to restriction shown in Figure 6.Second follows the trail of set comprises second reverser 652, second digital module 653, second analog module 655, reaches second capacitance module 661.Second digital module 653 is used for handling the even bit bit (even bit signal) that numerical control vibration decoder 1101 is exported.Second analog module 655 is used for handling the ∑ Delta modulator fractional signal that ∑ Delta modulator filter 1103 is exported.Second capacitance module 661 is used to provide high-level output voltage Vo+ and the required capacitance of low-level output voltage Vo-.In an embodiment of the present invention, first capacitance module is opposite with the polarity of the capacitance that second capacitance module is provided.Second digital module 653 comprises the 3rd P-type mos field-effect transistor 654 and the 5th N type metal oxide semiconductor field-effect transistor 656.Second analog module 655 comprises the 4th P-type mos field-effect transistor 658 and the 6th N type metal oxide semiconductor field-effect transistor 660.Second capacitance module 661 comprises the 7th N type metal oxide semiconductor field-effect transistor 662 and the 8th N type metal oxide semiconductor field-effect transistor 664.Second follows the trail of set comprises the 3rd electric capacity 666, the 4th electric capacity 668, the 3rd resistance 670 and the 4th resistance 672 in addition.

The anode of first reverser 602 is used for receiving the selection signal.The grid of the first P-type mos field-effect transistor 604 is coupled to the anode of first reverser 602; And the source electrode of the first P-type mos field-effect transistor 604 receives strange bit.The drain electrode of the one N type metal oxide semiconductor field-effect transistor 606 is coupled to the source electrode of the first P-type mos field-effect transistor 604, and the source electrode of a N type metal oxide semiconductor field-effect transistor 606 is coupled to the drain electrode of the first P-type mos field-effect transistor 604.The grid of the second P-type mos field-effect transistor 608 is coupled to the negative terminal of first reverser 602 and the grid of a N type metal oxide semiconductor field-effect transistor 606.The drain electrode of the 2nd N type metal oxide semiconductor field-effect transistor 610 is coupled to the source electrode of the second P-type mos field-effect transistor 608, to receive the signal that ∑ Δ low pass filter (sigma-delta low-pass filter) is exported, i.e. ∑ Delta modulator fractional signal.The source electrode of the 2nd N type metal oxide semiconductor field-effect transistor 610 is coupled to the drain electrode of the second P-type mos field-effect transistor 608 and the source electrode of a N type metal oxide semiconductor field-effect transistor 606.The grid of the 2nd N type metal oxide semiconductor field-effect transistor 610 is coupled to the grid of the first P-type mos field-effect transistor.The source electrode of the 3rd N type metal oxide semiconductor field-effect transistor 612 is coupled to the source electrode of a N type metal oxide semiconductor field-effect transistor 606.The drain electrode of the 3rd N type metal oxide semiconductor field-effect transistor 612 is coupled to the source electrode of the 3rd N type metal oxide semiconductor field-effect transistor 612.The drain electrode of the 4th N type metal oxide semiconductor field-effect transistor 614 is coupled to the source electrode of the 3rd N type metal oxide semiconductor field-effect transistor 612.The source electrode of the 4th N type metal oxide semiconductor field-effect transistor 614 is coupled to the drain electrode of the 3rd N type metal oxide semiconductor field-effect transistor 612.First end of first electric capacity 616 is coupled to the grid of the 3rd N type metal oxide semiconductor field-effect transistor 612, and second end of first electric capacity 616 is used for exporting first high-level output voltage, for example high-level output voltage Vo+.First end of second electric capacity 618 is coupled to the grid of the 4th N type metal oxide semiconductor field-effect transistor 614, and second end of second electric capacity 618 is used for exporting first low-level output voltage, for example low-level output voltage Vo-.First end of first resistance 620 is coupled to first end of first electric capacity 616, and second end of first resistance 620 is used for receiving the required voltage VCCreg that low dropout voltage regulator produces.First end of second resistance 622 is coupled to first end of second electric capacity 618, and second end of second resistance 622 is used for receiving the required voltage VCCreg that low dropout voltage regulator produces.

Second follows the trail of set comprises second reverser 652, the 3rd P-type mos field-effect transistor 654, the 5th N type metal oxide semiconductor field-effect transistor 656, the 4th P-type mos field-effect transistor 658, the 6th N type metal oxide semiconductor field-effect transistor 660, the 7th N type metal oxide semiconductor field-effect transistor 662, the 8th N type metal oxide semiconductor field-effect transistor 664, the 3rd electric capacity 666, the 4th electric capacity 668, the 3rd resistance 670, and the 4th resistance 672.The anode of second reverser 652 is used for receiving the selection signal.The 3rd P-type mos field-effect transistor 654 grids are coupled to the anode of second reverser 652, and the source electrode of the 3rd P-type mos field-effect transistor 654 is used for receiving even bit.The drain electrode of the 5th N type metal oxide semiconductor field-effect transistor 656 is coupled to the source electrode of the 3rd P-type mos field-effect transistor 654, the source electrode of the 5th N type metal oxide semiconductor field-effect transistor 656 is coupled to the drain electrode of the 3rd P-type mos field-effect transistor 654, and the grid of the 5th N type metal oxide semiconductor field-effect transistor 656 is coupled to the negative terminal of second reverser 652.The grid of the 4th P-type mos field-effect transistor 658 is coupled to the grid of the 5th N type metal oxide semiconductor field-effect transistor 656, the source electrode of the 4th P-type mos field-effect transistor 658 is used for receiving the signal from ∑ Δ low pass filter, and the drain electrode of the 4th P-type mos field-effect transistor 658 is coupled to the source electrode of the 5th N type metal oxide semiconductor field-effect transistor 656.The drain electrode of the 6th N type metal oxide semiconductor field-effect transistor 660 is coupled to the source electrode of the 4th P-type mos field-effect transistor 658, the source electrode of the 6th N type metal oxide semiconductor field-effect transistor 660 is coupled to the drain electrode of the 4th P-type mos field-effect transistor 658, and the grid of the 6th N type metal oxide semiconductor field-effect transistor 660 is coupled to the grid of the 3rd P-type mos field-effect transistor 654.The grid of the 7th N type metal oxide semiconductor field-effect transistor 662 is coupled to the source electrode of the 5th N type metal oxide semiconductor field-effect transistor 656, and the drain electrode of the 7th N type metal oxide semiconductor field-effect transistor 662 is coupled to the source electrode of the 7th N type metal oxide semiconductor field-effect transistor 662.The grid of the 8th N type metal oxide semiconductor field-effect transistor 664 is coupled to the grid of the 7th N type metal oxide semiconductor field-effect transistor 662, and the drain electrode of the 8th N type metal oxide semiconductor field-effect transistor 664 is coupled to the source electrode of the 8th N type metal oxide semiconductor field-effect transistor 664.First end of the 3rd electric capacity 666 is coupled to the drain electrode of the 7th N type metal oxide semiconductor field-effect transistor 662, and second end of the 3rd electric capacity 666 is used for exporting second high-level output voltage, for example high-level output voltage Vo+.First end of the 4th electric capacity 668 is coupled to the drain electrode of the 8th N type metal oxide semiconductor field-effect transistor 664, and second end of the 4th electric capacity 668 is used for exporting second low-level output voltage, for example low-level output voltage Vo-.First end of the 3rd resistance 670 is coupled to first end of the 3rd electric capacity 666, and second end of the 3rd resistance 670 is used for receiving the required voltage VCCreg of low dropout voltage regulator.First end of the 4th resistance 672 is coupled to first end of the 4th electric capacity 668, and second end of the 4th resistance 672 is used for receiving the required voltage VCCreg of low dropout voltage regulator.First high-level output voltage and first low-level output voltage are used to refer to the vibration in the strange bit of following the trail of groove 310, and second high-level output voltage and second low-level output voltage are used to refer to the vibration in the even bit of tracking groove 310.

Negative trnasducing element 3042 is fed into each unit 600 with control signal, stablizes the oscillatory occurences of high-level output voltage Vo+ and low-level output voltage Vo-so that required positive feedback (Positive Feedback) to be provided.As shown in Figure 6, predetermined control signal and first reverser 602 (perhaps second reverser 652) by node SEL and b place, at one time, (perhaps between second digital module 653 and second analog module 655) only has one of them and is unlocked between first digital module 603 and first analog module 605, promptly introduce the control mutual idol of voltage (Control voltage parity) herein, that is to say that control signal makes to have alternative between above-mentioned wantonly two modules at this.First reverser 602 and second reverser 652 are used for promoting the mutual idol of control voltage between first digital module and first analog module and between second digital module and second analog module respectively.Therefore, the running that is relevant to integer signal and fractional signal can separatedly be come and independently of one another, to realize frequency shown in Figure 7 mechanism continuously.Note that in other execution mode of the present invention the composition mode of negative trnasducing element 3042 is not subjected to restriction shown in Figure 3 with the composition assembly.

Please note, first capacitance module 611 is opposite with the polarity of the capacitance that second capacitance module 661 is produced, with corresponding strange bit of difference and even bit, and opposite like this polarity also can make to win and follow the trail of in the set and the second tracking set, and the formed voltage-frequency transformation curve of corresponding high-level output voltage Vo+ and low-level output voltage Vo-all becomes reciprocal curve.As shown in Figure 7, when the value of integer signal was coupled with 1, the trend of curve presented the trend opposite with curve shown in Figure 5, made that the said frequencies non-continuous event is eliminated.Therefore, cause the surging of interference and noise all can disappear, and relevant phase place also can be locked continuously.

Then disclose the digital loop bandwidth calibration method of all-digital phase-locked loop 100 of the present invention or 200.In order to explain details, need to use the simple and easy diagram of all-digital phase-locked loop 100 to describe earlier at this at the digital loop bandwidth calibration method of all-digital phase-locked loop 100.See also Fig. 8, it is in order to explain the digital loop bandwidth calibration method of the present invention at all-digital phase-locked loop shown in Figure 1 100, the rough schematic view of employed all-digital phase-locked loop 100.Wherein, the second ∑ Delta modulator, 1121 received signal Δ F.Note that all-digital phase-locked loop 100 can be considered this moment high resolution frequency to digital quantizer (Frequency-to-digital converter, FDC).The key of implementing the digital loop bandwidth calibration method is the value of calibration-gain a, and this is because other relevant variable is all controllable variable, relevant details will after proved.The loop frequency range is defined as the proportional path gain of proportion expression path module 106 be multiply by

Therefore, the proportional path of proportion expression path module 106 gain Pgain can be expressed as follows:

$\mathrm{Pgain}=\frac{\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{\mathrm{Fref}}---\left(9\right)$

Its discipline BW represents the initial loop frequency range of all-digital phase-locked loop.Simple and easy schematic diagram by observing Fig. 8 as can be known, proportional path gain Pgain also can be expressed as follows:

$\mathrm{Pgain}=\frac{1}{\mathrm{TDC}}\&CenterDot;\mathrm{aDCO}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{1}{F{\mathrm{ref}}^2}---\left(10\right).$

The definition of variable shown in the equation (10) is identical with the variable of same names in above-mentioned each equation, so do not repeat to give unnecessary details in this definition with regard to each variable.

Be illustrated in the unit interval sign indicating number variable quantity from phase-frequency detector and circulating time-to-digit converter module 1021.The gain a of proportion expression path module amplifier 1062 can be considered the gain of wave digital lowpass filter 108 at this moment.

Represent the sign indicating number variable quantity of the output of wave digital lowpass filter 108, sign indicating number variation delta I just shown in Figure 8.

Representative comes from the frequency variation Δ f of yard variation delta I _c Representative with frequency variation Δ f divided by second frequency divider, 1122 employed divisors (Dividing ratio) M.At last, please note the proportion expression path gain Pgain representative sign indicating number caused time drift of variable quantity (Time drift) Δ t in the unit interval _cNote that the reference cycle satisfied $\mathrm{Tref}=\frac{1}{\mathrm{Fref}},$ It is as follows then can to get equation:

$\frac{{\mathrm{\&Delta;t}}_c}{\mathrm{Tref}}=\frac{{\mathrm{\&Delta;f}}_c}{\mathrm{Fref}}---\left(11\right).$

So time drift Δ t _cIt is as follows to derive:

$\mathrm{Pgain}={\mathrm{\&Delta;t}}_c=\frac{{\mathrm{\&Delta;t}}_c}{\mathrm{Tref}}\&CenterDot;\frac{1}{\mathrm{Fref}}=\frac{{\mathrm{\&Delta;f}}_{\mathrm{<figure-callout\; id="2"\; label="c\; Fre\; f"\; filenames="A200810169984E00401.png"\; state="\{\{state\}\}">c</figure-callout>}}}{\mathrm{Fre}{f}^{}}=\frac{\mathrm{\&Delta;f}}{M}\&CenterDot;\frac{1}{{\mathrm{Fref}}^2}=\frac{1}{\mathrm{TDC}}\&CenterDot;a\&CenterDot;\mathrm{DCO}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{1}{{\mathrm{Fref}}^2}---\left(12\right).$

Equation (12) is explained the step of deriving equation (10).Note that gain DCO also can be considered gain Kv.By merging equation (9) and (10), and with reference to equation (6) and (8), gain a can derive as follows:

$\frac{1}{\mathrm{TDC}}\&CenterDot;a\&CenterDot;\mathrm{Kv}\&CenterDot;\frac{1}{M}\&CenterDot;\frac{1}{{\mathrm{Fref}}^2}=\mathrm{Pgain}=\frac{\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{\mathrm{Fref}}---\left(13\right);$ And

$a=\frac{\mathrm{TDC}\&CenterDot;M\&CenterDot;{\mathrm{<figure-callout\; id="2"\; label="Fref"\; filenames="A200810169984E00401.png"\; state="\{\{state\}\}">Fref</figure-callout>}}^2\&CenterDot;\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{\mathrm{Kv}\&CenterDot;\mathrm{Fref}}=\frac{\mathrm{TDC}\&CenterDot;M\&CenterDot;\mathrm{Fref}\&CenterDot;\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{\mathrm{Kv}}$

$=\frac{1}{2\mathrm{Fref}\&CenterDot;{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="A200810169984E00361.png,A200810169984E00381.png"\; state="\{\{state\}\}">N</figure-callout>}}_1}\&CenterDot;M\&CenterDot;\mathrm{Fref}\&CenterDot;\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}\&CenterDot;\frac{\mathrm{\&Delta;I}}{\mathrm{\&Delta;N}\&CenterDot;\mathrm{Fref}}$

$=\frac{\mathrm{\&Delta;I}\&CenterDot;M\&CenterDot;\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{{2\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="A200810169984E00361.png,A200810169984E00381.png"\; state="\{\{state\}\}">N</figure-callout>}}_1\&CenterDot;\mathrm{Fref}\&CenterDot;\mathrm{\&Delta;N}}---\left(14\right).$

Each variable relevant with gain a has been proved to be in above-listed narration and has been controlled variable in equation (14), and a that therefore gains is also for controlled.That is to say,, can realize the loop bandwidth calibration method of all-digital phase-locked loop 100 by adjust gain a according to equation (14).

In Fig. 1, error compensating signal Err is produced by ∑ Delta modulator compensating module 114, with the error that may exist in make-up time digital quantizer module 102 and the time figure switch decoders 1022.Error compensating signal Err mainly produces according to the fractional phase error.See also Fig. 9, it is the rough schematic that is used for explaining the fractional phase error that how to compensate the prior art analog phase-locked look.In Fig. 9, provided the schematic diagram of clock edge and phase error.The fractional phase error can be represented with the difference between actual clock position N+e (n) and the desirable clock position N+aa, wherein actual clock position N+e (n) is produced by the ∑ Delta modulator, and e (n) is an integer, aa is a mark, and desirable clock position N+aa is positioned between clock position N+e (n)-1 and the actual clock position N+e (n).Therefore, the fractional phase error phase_error by the correspondence that phase-frequency detector produced can be expressed as:

phase_error＝[N+e(n)-(N+aa)]·T _VCO＝[e(n)-aa]·T _VCO (15)；

Because of equation (15) is to derive according to analog phase-locked look, and the value of equation (15) approximates

So T _VCOThe cycle of expression voltage controlled oscillator.By the corresponding fractional phase error phase_error that adds up, can obtain the compensating error compensation_error that adds up, and can be expressed as:

compensation_error＝∑[e(n)-aa]·T _VCO (16)；

By digital quantizer service time, the compensating error that adds up compensation_error also can be quantified as:

compensation_error＝∑[e(n)-aa]·T _VCO/TDC

≈∑[e(n)-aa]/[TDC·Fref·(N+a)] (17)。

Yet service time, digital quantizer can cause a yard variable quantity (N for example ₁) with time-to-digit converter in a large amount of delay lines (Delay line), and the bigger circuit area that accounts for, consume higher power etc.Therefore, the present invention also discloses a kind of circulating time-to-digit converter that is arranged at phase-frequency detector and circulating time-to-digit converter module 1021 inside, with tap (Tap) quantity of a large amount of saving delay line (delay line).Circulating time-to-digit converter will after disclose separately.Moreover in the disclosed all-digital phase-locked loop 100 of the present invention, digital controlled oscillator 1104 is used for replacing the voltage controlled oscillator of prior art.Under the running of the second ∑ Delta modulator 1121, can obtain the difference between actual clock position N+e (n) and the desirable clock position N+aa, this difference at this with e _{The Δ ∑}Represent, and difference e in fact _{The Δ ∑}Be quantization error (Quantization error).According to Error Compensation Algorithm used in the present invention, the compensating error e of the circulating time-to-digit converter in phase-frequency detector and the circulating time-to-digit converter module 1021 _CTDC[k] can be expressed as:

$e_{\mathrm{CTDC}}\left[k\right]=\underset{n=0}{\overset{k-1}{\mathrm{\&Sigma;}}}{e}_{\mathrm{\&Delta;\&Sigma;}}\&CenterDot;\frac{T_{\mathrm{DCO}}}{\mathrm{TDC}}---\left(18\right);$

Its discipline T _DCORepresent the cycle of digital controlled oscillator 1104.And the period T of digital controlled oscillator 1104 _DCOCan be expressed as:

$T_{\mathrm{DCO}}=\frac{1}{\mathrm{Fref}\&CenterDot;(M+F)}---\left(19\right);$

According to equation (19), circulation timei digital quantizer compensating error e _CTDC[k] can further derive as follows:

$e_{\mathrm{CTDC}}\left[k\right]=\underset{n=0}{\overset{k-1}{\mathrm{\&Sigma;}}}{e}_{\mathrm{\&Delta;\&Sigma;}}\&CenterDot;\frac{1}{\mathrm{TDC}\&CenterDot;\mathrm{Fref}\&CenterDot;(M+F)}$

$=\underset{n=0}{\overset{k-1}{\mathrm{\&Sigma;}}}{e}_{\mathrm{\&Delta;\&Sigma;}}\&CenterDot;\frac{\mathrm{\&Delta;N}}{\mathrm{\&Delta;t}\&CenterDot;\mathrm{Fref}\&CenterDot;(M+F)}=\underset{n=0}{\overset{k-1}{\mathrm{\&Sigma;}}}{e}_{\mathrm{\&Delta;\&Sigma;}}\&CenterDot;\frac{2\mathrm{Fref}\&CenterDot;{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="A200810169984E00361.png,A200810169984E00381.png"\; state="\{\{state\}\}">N</figure-callout>}}_1}{\mathrm{Fref}\&CenterDot;(M+F)}$

$=\underset{n=0}{\overset{k-1}{\mathrm{\&Sigma;}}}{e}_{\mathrm{\&Delta;\&Sigma;}}\&CenterDot;\frac{2\&CenterDot;N_1}{(M+F)}---\left(20\right);$

Wherein, F is meant the mark relevant with quantization error.Observation equation formula (20) as can be known, the compensating error e of circulating time-to-digit converter _CTDC[k] is numeral, and controlled fully, and is applied in the digital phase error elimination of the present invention (Digital phase error cancellation).See also Figure 10, it is the schematic diagram of the digital phase error cancellation module 1144 that comprises in addition in the disclosed according to an embodiment of the present invention ∑ Delta modulator compensating module 114.Digital phase error cancellation module 1144 operates based on equation (20).Digital phase error cancellation module 1144 comprises ∑ Delta modulator 702, first adder 704, second adder 706, first d type flip flop (D Flip-Flop, DFF) 708, second d type flip flop 710, divider 712, multiplier 714 and d type flip flop and truncation module (DFF/Truncation module) 716.∑ Delta modulator 702 is implemented with the multistage noise shaping 1-1-1 modulator (Multi-stage noise shaping 1-1-1 modulator, MASH 1-1-1modulator) that comprises a plurality of single order modulators (First-order modulator).The multistage noise shaping n-1-1 modulator that use comprises a n rank modulator and a plurality of single order modulators comparatively significantly advantage is to reduce the mismatch phenomenon of (Coefficient mismatch) of coefficient, and this is because most of noise can be eliminated easily in inside.∑ Delta modulator 702, first adder 704, second adder 706, be used for producing with first d type flip flop 708 and quantize error e _{The Δ ∑}(quantization error e as shown in figure 10 _{The Δ ∑}[n]).∑ Delta modulator 702 received signal F, and output signal F_ Δ ∑.Divider 712 received signal 2N ₁With signal M+F.Second d type flip flop 710 is used for producing the item shown in the equation (20) with divider 712

At last, compensating error e _CTDC[k] can be output to first adder 104.

The present invention has used special technology on time figure switch decoders 1022, for example mistake prevents method (error protection method).In this technology, the output signal TDC of time-to-digit converter 1022 can be added an error protection sign indicating number (Error protection code) in addition, to improve the accuracy of output signal TDC.The input signal of supposing time figure switch decoders 1022 comprises data-signal D[0:2 ^m-1] with cycle signal C[0:(m-1)], data-signal D[0:2 wherein ^m-1] comprises 2 ^mIndividual bit, signal C[0:(m-1)] comprise m bit, and m is a positive integer.In an embodiment of the present invention, the value of positive integer m is 5, so cycle signal C comprises 5 bits, and data-signal D comprises 32 bits.Briefly, error protection sign indicating number err_protect can via first bit of last bit of data-signal D and cycle signal C is carried out mutual exclusion or (Exclusive-or) logical operation realize.Therefore error protection sign indicating number err_protect can be expressed as:

err_protect＝XOR(D[2 ^m-1]，C[0]) (21)。

In one embodiment of the present invention, the output signal TDC[0:2 of time figure switch decoders 1022 (m-1)+1] comprise 10 bits, and output signal TDC[0:2 (m-1)+1] can be expressed as:

TDC[0:2·(m-1)+1]＝(C[0:(m-1)]+err_protect)*2 ^m+output1[0:(m-1)] (22)；

Note that an output1 represents the decoded signal of time figure switch decoders 1022, with the bit 0 that comprised among the expression data-signal D or the quantity of bit 1.By error protection sign indicating number (or bit) being added cycle signal C, and by cycle signal C being improved m bit (this is to be 2 because of multiplier ^m, just cycle signal C be multiply by 2 ^mOr with cycle signal C m the bit that move to left), the accuracy of the output signal TDC of time figure switch decoders 1022 significantly can be improved.

See also Figure 11, Figure 12, reach Figure 13.Figure 11 when implementing loop gain calibration steps shown in Figure 8, phase-frequency detector shown in Figure 1 and circulating time-to-digit converter module 1021 and the time figure switch decoders 1022 shown in Figure 1 and the simple and easy schematic diagram of first adder 104.Figure 12 is the generalized schematic of circulating time-to-digit converter shown in Figure 11.Figure 13 is the schematic flow sheet that is relevant to the circulating time-to-digit converter calibration procedure of Figure 11 and Figure 12.

As shown in figure 11, phase-frequency detector and circulating time-to-digit converter module 1021 comprise multiplexer 10211, phase-frequency detector 10212, logical block 10213, circulating time-to-digit converter 10214, reach time digital quantizer controller calibration 10215.Multiplexer 10211 is used for receiving reference signal REF shown in Figure 1 and feedback signal FB.Phase-frequency detector 10212 receives from two output signal A of multiplexer 10211 and B, and wherein output signal A and B are corresponding to reference signal REF or feedback signal FB.As the description about Fig. 1, phase-frequency detector 10212 is gone back output frequency and is promoted signal Up and frequency reduction signal Dn, to improve or to reduce the frequency of the output signal TDC of first adder 104.Logical block 10213 receive frequencies promote signal Up and frequency reduces signal Dn, and send enabling signal St or stop signal Sp, to start or stop the running of circulating time-to-digit converter 10214 at any time.Logical block 10213 is gone back output symbol signal L to time figure switch decoders 1022.Time figure switch decoders 1022 output symbol signal S and prediction signal TDC_pre.In an embodiment of the present invention, mark signal S produces according to mark signal L, and prediction signal TDC_pre comprises the information of feedback signal FB.Circulating time-to-digit converter 10214 also produces data-signal D and cycle signal C, wherein data-signal D is corresponding to the d type flip flop of circulating time-to-digit converter 10214 inside, and cycle signal C is corresponding to circulating time-to-digit converter 10214 inner employed circulations.Time-to-digit converter controller calibration 10215 produces the shifted signal Offs that process is calculated according to output signal TDC, and produces a sign indicating number variation delta N.In an embodiment of the present invention, time-to-digit converter controller calibration 10215 can use shifted signal Offs control multiplexer 10211 receive reference signals and feedback signal one of them.

As shown in figure 12, circulating time-to-digit converter 10214 comprises circulation module 102146 and data module 102148.With respect to circulation module 102146 and data module 102148, circulating time-to-digit converter 10214 also can be regarded as circulating time-to-digit converter module.Circulation module 102146 comprises double edge detector (Double-edge detector) 102141 and counter, N bit count-up counter for example shown in Figure 12 (N-bit up counter) 102142.Cycle signal C in the circulation module 102146 generation time digital quantizer modules 102.Data module 102148 comprise the first d type flip flop array 102143, the second d type flip flop array 102144, with circular buffering array (Cyclic bufferarray) 102145.Data-signal D in the data module 102148 generation time digital quantizer modules 102.Triggering signal Trig+ and Trig-that double edge detector 102141 receives in the data module 102148 are to detect rising edge (Rising edge) and drop edge (Falling edge).Double edge detector 102141 receives triggering signal Trig+ or Trig-from data module 102148.Whenever triggering signal Trig+ or Trig-rising one of at least triggers the edge or descends when triggering the edge and being received, can output signal Incr, so that the count increments of N bit count-up counter 102142.When the counting of N bit count-up counter 102142 surpasses predetermined value, will start the new circulation that begins among the N bit count-up counter 102142, and finish the old circulation of N bit count-up counter 102142.The replacement interface received signal stopb of N bit count-up counter 102142.At this moment, the number of the current circulation of record N bit count-up counter 102142 can be output with the form of cycle signal C.In an embodiment of the present invention, the bit number among the cycle signal C is 5, and cycle signal C is expressed as C[0:4].The first d type flip flop array 102143, the second d type flip flop array 102144, with the circular buffering array 102145 common circulation frameworks that form.Please note, circular buffering array 102145 comprises delay line buffer (Delay linebuffer) Binv of a plurality of series connection, and the input of first delay line buffer and the output of last delay line buffer interconnect among a plurality of delay line buffer Binv.In an embodiment of the present invention, the number of a plurality of delay line buffer Binv is 32, just a plurality of delay line buffer Binv0, Binv1 as shown in figure 12, Binv2 ..., Binv15, Binv16 ..., Binv29, Binv30, Binv31, and other logical block that a plurality of delay line buffer Binv can use reverser or be fit to implement delay line tap (Delay line tap) is implemented.Preceding half section operate together of the first d type flip flop array 102143 and a plurality of delay line buffer Binv, and the second half section operate together of the second d type flip flop array 102144 and a plurality of delay line buffer Binv.When the number of a plurality of delay line buffer Binv is 32, preceding half 16 bits of the first d type flip flop array, 102143 outputting data signals D, and later half 16 bits of the second d type flip flop array, 102144 outputting data signals D.Among Figure 12, preceding half 16 table of bits of data-signal D are shown D[0:15], later half 16 table of bits of data-signal D are shown D[16:31].In an embodiment of the present invention, the positive input terminal of double edge detector 102141 is coupled in a plurality of delay line buffer the negative input end of preceding delay line buffer, and is coupled to the positive output end of last delay line buffer, to receive first triggering signal; The negative input end of double edge detector is coupled to the positive input terminal of preceding delay line buffer, and is coupled to the negative output terminal of last delay line buffer, to receive second triggering signal.Note that present embodiment N bit count-up counter only is used to illustrate the present invention, and also unrestricted the present invention.In other execution mode of the present invention, also can use the counter of other type, the scope that this does not break away from the present invention is yet protected.

Figure 13 describes circulating time-to-digit converter 10214 employed calibration procedures, and calibration procedure is used for calibrating the loop gain of above-mentioned all-digital phase-locked loop 100 or 200.

As shown in figure 13, in step 1302, carrying out the offset calibration program, is reference signal REF to be used for specifying input signal A and B by direct control multiplexer 10211.Moreover, also be designated as the prediction signal TDC_pre of time figure switch decoders 1022 from the shifted signal Offs of time-to-digit converter controller calibration 10215.Note that prediction signal TDC_pre comprises the information of feedback signal FB, make that the predicated error that is comprised among the output signal TDC can be compensated in advance by the running of first adder 104.At this moment, the value of output signal TDC should be logical zero, and this moment, the offset calibration program was finished.

In step 1304, implement normalization (Normalization) program, and the enforcement of regular program is by keeping input signal A identical with reference signal REF, and input signal B reassigned to back-reference signal REFB finished, promptly fill strip and indicate (pad a bar), with indication inverted reference signal REFB.At this moment, the digital variation delta N of above-mentioned branch is produced by time-to-digit converter controller calibration 10215, and represent with the form of time-to-digit converter prediction drift signal TDC_pre-Offs, in the loop gain calibration procedure, to realize the normalization of all-digital phase-locked loop 100 or 200.

Step 1306 is represented the normal operation program of all-digital phase-locked loop 100 or 200.At this moment, it is identical with reference signal REF that input signal A still is held, and input signal B redesignated as identical with feedback signal FB, in the next one postpones, to measure the characteristic of the new output signal that digital controlled oscillator and ∑ Delta modulator module 110 produced.

By with the divisor of the frequency range in the all-digital phase-locked loop and reference frequency, time-to-digit converter gain, digital controlled oscillator gain, frequency divider, and Amplifier Gain define the proportion expression path gain of all-digital phase-locked loop, Amplifier Gain can obtain suitable adjustment, so that the optimal loop frequency range can accurately be adjusted in the all-digital phase-locked loop.By reaching the fully digitalization of all-digital phase-locked loop, can further adjust the gain of time-to-digit converter and digital controlled oscillator with digital form.

By the disclosed all-digital phase-locked loop of the present invention and other relevant assembly and method, because the employed all component of all-digital phase-locked loop is digitized all with operation, so avoided using the shortcoming of prior art analog phase-locked look.In addition, by being used for the above-mentioned disclosed pinpoint accuracy loop gain calibration steps of all-digital phase-locked loop, the available bandwidth of the all-digital phase-locked loop that is captured will be increased considerably because of the application of all-pass response.

The above only is a better embodiment of the present invention, and all equalizations of doing according to claim of the present invention change and modify, and all should belong to covering scope of the present invention.

Claims (25)

1. a loop bandwidth calibration method is used for all-digital phase-locked loop, and described loop bandwidth calibration method comprises:

According to the two combination of the gain of the gain of the gain of the gain of the time-to-digit converter of described all-digital phase-locked loop, digital controlled oscillator or described time-to-digit converter and described digital controlled oscillator, the Amplifier Gain of the proportion expression path module by adjusting described all-digital phase-locked loop is calibrated the loop frequency range of described all-digital phase-locked loop.

2. loop bandwidth calibration method as claimed in claim 1, it is characterized in that, according to the two combination of the gain of the gain of the gain of the gain of the described time-to-digit converter of described all-digital phase-locked loop, described digital controlled oscillator or described time-to-digit converter and described digital controlled oscillator, the described Amplifier Gain of the described proportion expression path module by adjusting described all-digital phase-locked loop, the loop frequency range of calibrating described all-digital phase-locked loop comprises:

According to the gain of the gain of described time-to-digit converter, described digital controlled oscillator, divisor, reference frequency, the initial loop frequency range of described all-digital phase-locked loop or the combination of above-mentioned each condition of frequency divider, the described Amplifier Gain of the described proportion expression path module by adjusting described all-digital phase-locked loop is calibrated the loop frequency range of described all-digital phase-locked loop.

3. loop bandwidth calibration method as claimed in claim 2 is characterized in that, described loop bandwidth calibration method comprises in addition:

According to the reference cycle of reference signal, determine described reference frequency.

4. loop bandwidth calibration method as claimed in claim 3 is characterized in that, described all-digital phase-locked loop comprises:

The time-to-digit converter module comprises described time-to-digit converter, and described time-to-digit converter module is used for receiving described reference signal;

The numeral macroblock comprises described proportion expression path module; And

Described frequency divider.

5. loop bandwidth calibration method as claimed in claim 2 is characterized in that, described Amplifier Gain is adjusted according to following formula:

$a=\frac{\mathrm{TDC}\&CenterDot;M\&CenterDot;\mathrm{Fref}\&CenterDot;\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{\mathrm{Kv}}$

Wherein a refers to controlled described Amplifier Gain; TDC refers to the gain of described time-to-digit converter; M refers to the described divisor of described frequency divider; Fref refers to described reference frequency; BW refers to the described initial loop frequency range of described all-digital phase-locked loop; And Kv refers to the gain of described digital controlled oscillator.

6. loop bandwidth calibration method as claimed in claim 3 is characterized in that, described loop bandwidth calibration method comprises in addition:

Determine the gain of described time-to-digit converter according to following formula:

$\mathrm{TDC}=\frac{\frac{1}{2}\mathrm{Tref}}{N_1}=\frac{1}{2\mathrm{Fref}\&CenterDot;N_1}$

Wherein TDC refers to the gain of described time-to-digit converter; Tref refers to the described reference cycle of described reference signal; Fref refers to described reference frequency; And N ₁Finger is corresponding to the sign indicating number variable quantity of the half period of described reference cycle Tref.

7. loop bandwidth calibration method as claimed in claim 3 is characterized in that, described loop bandwidth calibration method comprises in addition:

Determine the gain of described digital controlled oscillator according to following formula:

$\frac{1}{\mathrm{Kv}}=\frac{\mathrm{\&Delta;I}}{\mathrm{\&Delta;N}\&CenterDot;\mathrm{Fref}}$

Wherein Kv refers to the gain of described digital controlled oscillator; Fref refers to described reference frequency; Δ N refers to the digital variable quantity of branch; And Δ I refers to the sign indicating number variable quantity corresponding to frequency variation Δ NFref.

8. loop bandwidth calibration method as claimed in claim 3 is characterized in that, described loop bandwidth calibration method comprises in addition:

Determine the gain of described time-to-digit converter according to following formula:

$\mathrm{TDC}=\frac{\frac{1}{2}\mathrm{Tref}}{N_1}=\frac{1}{2\mathrm{Fref}\&CenterDot;N_1};$ And

Determine the gain of described digital controlled oscillator according to following formula:

$\frac{1}{\mathrm{Kv}}=\frac{\mathrm{\&Delta;I}}{\mathrm{\&Delta;N}\&CenterDot;\mathrm{Fref}};$

Wherein TDC refers to the gain of described time-to-digit converter; Tref refers to the described reference cycle of described reference signal; Fref refers to described reference frequency; And N ₁Finger is corresponding to second yard variable quantity of the half period of described reference cycle Tref; Kv refers to the gain of described digital controlled oscillator; Δ N refers to the digital variable quantity of branch; And Δ I refers to first yard variable quantity corresponding to frequency variation Δ NFref.

9. loop bandwidth calibration method as claimed in claim 8, it is characterized in that, according to the described initial loop frequency range of the described divisor of the gain of the gain of described time-to-digit converter, described digital controlled oscillator, described frequency divider, described reference frequency, described all-digital phase-locked loop or the combination of above-mentioned each condition, by the described Amplifier Gain in the described proportion expression path module of adjusting described all-digital phase-locked loop, the loop frequency range of calibrating described all-digital phase-locked loop comprises:

According to the combination of the described initial loop frequency range of described first yard variable quantity, the described divisor of described frequency divider, described all-digital phase-locked loop, described second yard variable quantity, described reference frequency, the digital variable quantity of described branch or above-mentioned each condition, adjust the described Amplifier Gain of the described proportion expression path module of described all-digital phase-locked loop.

10. loop bandwidth calibration method as claimed in claim 9 is characterized in that, described all-digital phase-locked loop comprises:

The time-to-digit converter module comprises described time-to-digit converter, and described time-to-digit converter module is used for receiving described reference signal;

The numeral macroblock comprises described proportion expression path module;

Wave digital lowpass filter is used for exporting described first yard variable quantity according to the digital variable quantity of described branch; And

The feedback path module comprises described frequency divider,

Wherein said time-to-digit converter module is used for producing described second yard variable quantity according to the described reference signal that is received.

11. loop bandwidth calibration method as claimed in claim 9, it is characterized in that, according to the combination of the described initial loop frequency range of described first yard variable quantity, the described divisor of described frequency divider, described all-digital phase-locked loop, described second yard variable quantity, described reference frequency, the digital variable quantity of described branch or above-mentioned each condition, the described Amplifier Gain of adjusting the described proportion expression path module of described all-digital phase-locked loop comprises:

Adjust described Amplifier Gain according to following formula:

$a=\frac{\mathrm{\&Delta;I}\&CenterDot;M\&CenterDot;\mathrm{BW}\&CenterDot;2\mathrm{\&pi;}}{2N_1\&CenterDot;\mathrm{Fref}\&CenterDot;\mathrm{\&Delta;N}};$

Wherein a refers to controlled described Amplifier Gain; Δ I refers to described first yard variable quantity; M refers to the described divisor of described frequency divider; BW refers to the described initial loop frequency range of described all-digital phase-locked loop; N ₁Refer to described second yard variable quantity; Fref refers to described reference frequency; Δ N refers to the digital variable quantity of described branch.

12. an all-digital phase-locked loop is characterized in that, described all-digital phase-locked loop comprises:

Digital loop filters comprises the proportion expression path module, described proportion expression path module be used for following the trail of with from the relevant phase change of the output signal of time-to-digit converter module, and described proportion expression path module comprises proportion expression path module amplifier,

Wherein said proportion expression path module Amplifier Gain is adjusted according to the gain of the time-to-digit converter of described time-to-digit converter module.

13. all-digital phase-locked loop as claimed in claim 12, it is characterized in that, described all-digital phase-locked loop more comprises the digital controlled oscillator module, be used for following the trail of the integer signal from described digital loop filters, wherein said proportion expression path module Amplifier Gain is more adjusted according to the gain of the digital controlled oscillator in the described digital controlled oscillator module.

14. all-digital phase-locked loop as claimed in claim 13 is characterized in that, described time-to-digit converter module receives corresponding to the feedback signal from the output signal of described digital controlled oscillator module; And described time-to-digit converter module also comprises the time-to-digit converter controller calibration, to produce the digital variable quantity of branch corresponding to the gain of described time-to-digit converter.

15. all-digital phase-locked loop as claimed in claim 14 is characterized in that, described time-to-digit converter module comprises in addition:

Phase-frequency detector is used for receiving two input signals corresponding to reference signal or described feedback signal, and is used for output frequency and promotes signal and frequency reduction signal; And

Logical block is used for receiving described frequency upgrading signal and described frequency and reduces signal, and is used for sending enabling signal or stop signal,

Wherein said time-to-digit converter is used for receiving described enabling signal or described stop signal, and is used for producing cycle signal and data-signal.

16. all-digital phase-locked loop as claimed in claim 15 is characterized in that, described all-digital phase-locked loop comprises in addition:

The time figure switch decoders is used for receiving described cycle signal and described data-signal, and produces prediction signal; And

Adder is used for the shifted signal that the described prediction signal that produced according to described time figure switch decoders and described time-to-digit converter controller calibration exported to produce output signal, and imports described digital loop filters,

Wherein said frequency upgrading signal is used for improving the described output signal frequency of described adder, and described frequency reduces the described output signal frequency that signal is used for reducing described adder, described enabling signal is used for starting described time-to-digit converter, and described stop signal is used for stopping described time-to-digit converter

Wherein first predetermined bit of described data-signal and second predetermined bit of described cycle signal are implemented the exor computing, to produce the error protection sign indicating number; And by described error protection sign indicating number being added described cycle signal and with the bit number of described cycle signal displacement predetermined number, the error in the described cycle signal is corrected.

17. all-digital phase-locked loop as claimed in claim 13 is characterized in that, all-digital phase-locked loop comprises in addition:

The error compensation module is used for predicting the error from the output signal of described digital controlled oscillator module, and is used for described predicated error is imported described digital loop filters with feed-forward mode.

18. all-digital phase-locked loop as claimed in claim 13 is characterized in that, described digital controlled oscillator module comprises:

Numerical control vibration decoder, the first input end of described numerical control vibration decoder receives the described integer signal from described digital loop filters;

Described digital controlled oscillator, the input of described digital controlled oscillator are coupled to the output of described numerical control vibration decoder; And

Frequency divider, the input of described frequency divider is coupled to the output of described digital controlled oscillator, and the output of described frequency divider is coupled to second input of described numerical control vibration decoder,

Wherein first loop is passed through described numerical control vibration decoder, described digital controlled oscillator, is reached described frequency divider, and is used for described integer signal is modulated.

19. all-digital phase-locked loop as claimed in claim 13 is characterized in that, described all-digital phase-locked loop comprises in addition:

Frequency divider is coupled between described time-to-digit converter module and the described digital controlled oscillator module, is used for the output signal from described digital controlled oscillator module is carried out frequency division.

20. a loop gain calibration steps is used for all-digital phase-locked loop, described loop gain calibration steps comprises:

The frequency response of the digital controlled oscillator of the reference frequency of the described reference signal that the sign indicating number variable quantity of the half period in the reference cycle of the reference signal that is received according to the time-to-digit converter module corresponding to described all-digital phase-locked loop, described time-to-digit converter module are received, the digital controlled oscillator of described all-digital phase-locked loop and ∑ Delta modulator module or the combination of above-mentioned each condition are modulated the ∑ Delta modulator compensating module Amplifier Gain of ∑ Delta modulator compensating module; And

According to the frequency variation of the input of the ∑ Delta modulator of the feedback path module of described all-digital phase-locked loop, corresponding to the sign indicating number variable quantity of described frequency variation, divide the described reference frequency of the described reference signal that digital variable quantity, described time-to-digit converter module received or the combination of above-mentioned each condition, the gain of the modulator amplifier of modulator is modulated.

21. path gain calibration steps as claimed in claim 20, it is characterized in that, the described sign indicating number variable quantity of the half period in the described reference cycle of the described reference signal that is received according to described time-to-digit converter module corresponding to described all-digital phase-locked loop, the described reference frequency of the described reference signal that described time-to-digit converter module is received, the described frequency response of the described digital controlled oscillator of the described digital controlled oscillator of described all-digital phase-locked loop and ∑ Delta modulator module or the combination of above-mentioned each condition, the described ∑ Delta modulator compensating module Amplifier Gain of described ∑ Delta modulator compensating module modulated comprise:

According to following formula described ∑ Delta modulator compensating module Amplifier Gain is modulated:

$b=\frac{2N_1}{M}\&CenterDot;\frac{1}{\mathrm{Fref}}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}$

Wherein b is described ∑ Delta modulator compensating module Amplifier Gain; N ₁Described sign indicating number variable quantity for half period in described reference cycle of the described reference signal that received corresponding to described time-to-digit converter module; Response for frequency divider; Fref is the described reference frequency of the described reference signal that received of described time-to-digit converter module;

Described frequency response for described digital controlled oscillator.

22. loop gain calibration steps as claimed in claim 21 is characterized in that, described method comprises in addition:

According to following formula the accumulator Amplifier Gain corresponding to accumulator is modulated:

$b=\frac{2N_1}{M}\&CenterDot;\frac{1}{\mathrm{Fref}}\&CenterDot;\frac{Z^{-1}}{1-Z^{-1}}$

Wherein b is described accumulator Amplifier Gain; N ₁Described sign indicating number variable quantity for half period in reference cycle of the described reference signal that received corresponding to described time-to-digit converter module;

Described response for described frequency divider; Fref is the described reference frequency of the described reference signal that received of described time-to-digit converter module;

Described frequency response for described digital controlled oscillator.

23. loop gain calibration steps as claimed in claim 20, it is characterized in that, the described reference frequency of the described reference signal that is received according to the described frequency variation of the described input of the described ∑ Delta modulator of the described feedback path module of described all-digital phase-locked loop, corresponding to described sign indicating number variable quantity, the digital variable quantity of described branch, the described time-to-digit converter module of described frequency variation or the combination of above-mentioned each condition, the gain of the described modulator amplifier of described modulator modulated comprise:

According to following formula the gain of described modulator amplifier is modulated:

$c=\frac{\mathrm{\&Delta;I}}{\mathrm{\&Delta;N}\&CenterDot;\mathrm{Fref}}$

Wherein c is the gain of described modulator amplifier; Δ NFref refers to the described frequency variation of the described input of described ∑ Delta modulator; Δ I refers to the described sign indicating number variable quantity corresponding to described frequency variation Δ NFref, and described sign indicating number variation delta I is obtained in the output signal of wave digital lowpass filter; Δ N refers to the digital variable quantity of described branch; Fref is the described reference frequency of the described reference signal that received of described time-to-digit converter module.

24. loop gain calibration steps as claimed in claim 22, it is characterized in that, the described reference frequency of the described reference signal that is received according to the described frequency variation of the described input of the described ∑ Delta modulator of the described feedback path module of described all-digital phase-locked loop, corresponding to described sign indicating number variable quantity, the digital variable quantity of described branch, the described time-to-digit converter module of described frequency variation or the combination of above-mentioned each condition, the gain of the described modulator amplifier of described modulator modulated comprise:

According to following formula the gain of described modulator amplifier is modulated:

$c=\frac{\mathrm{\&Delta;I}}{\mathrm{\&Delta;N}\&CenterDot;\mathrm{Fref}};$

Wherein c is the gain of described modulator amplifier; Δ NFref refers to the described frequency variation of described input of the described ∑ Delta modulator of described feedback path module; Δ I refers to the described sign indicating number variable quantity corresponding to described frequency variation Δ NFref, and described sign indicating number variation delta I takes out in the output signal of wave digital lowpass filter; Δ N refers to the digital variable quantity of described branch; Fref is the described reference frequency of the described reference signal that received of described time-to-digit converter module.

25. loop gain calibration steps as claimed in claim 24 is characterized in that, described loop gain calibration steps comprises in addition:

Described all-digital phase-locked loop is provided, and described all-digital phase-locked loop comprises:

Described time-to-digit converter module is used for receiving described reference signal;

The numeral macroblock comprises described wave digital lowpass filter;

The error compensation module comprises first accumulator and first amplifier;

Described digital controlled oscillator and ∑ Delta modulator module comprise described digital controlled oscillator;

Described feedback path module is used for and described error compensation module operate together;

Second accumulator, the input of described second accumulator is used to receive modulation signal;

Described accumulator amplifier, described accumulator amplifier input terminal is coupled to the output of described second accumulator, the output of described accumulator amplifier is coupled to the first adder of described digital macroblock, and described accumulator Amplifier Gain is identical with described ∑ Delta modulator compensating module Amplifier Gain; And

Described modulator amplifier, the input of described modulator amplifier is used to receive described modulation signal, and the output of described modulator amplifier is coupled to the second adder of described digital macroblock.

Images