EP1859560A1

EP1859560A1 - Pulse width modulation data transmission method and transmitter and receiver therefor

Info

Publication number: EP1859560A1
Application number: EP06724909A
Authority: EP
Inventors: Dieter Thoss; Jens Boettcher; Andreas Kneer; Michael Besemer
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2005-03-03
Filing date: 2006-03-02
Publication date: 2007-11-28
Also published as: CN101133589A; US20080212668A1; WO2006092424A1; DE102005009735A1; JP2008532399A

Abstract

The invention relates to a method for transmitting serial data from a transmitter to a receiver consisting a) in fractionating a transmissible data value into several data words, b) in setting a pulse duration proportional to a non-negative whole number represented by a respective data word, including a constant positive supplement and producing a corresponding impulse, c) in transmitting an impulse (IH, IM, IL) having a respective set impulse duration from the transmitter to the receiver, and d) in detecting respective impulse duration on the receiver and in restoring a respective data word therefrom and, subsequently, the data value.

Description

ROBERT BOSCH GMBH, 70442 STUTTGART

Data transmission method, transmitter and receiver for this

State of the art

The present invention relates to a method for the serial data transmission from a transmitter to a receiver as well as transmitters and receivers which are suitable for carrying out the method.

In modern motor vehicles, more and more digital data processing circuits are being used, in particular for detecting and controlling or regulating operating parameters of the internal combustion engine of a motor vehicle. Digital data lines running in the vicinity of the internal combustion engine are subject to strong electromagnetic interference from the operation of the latter, in particular by the ignition sparks. This problem is particularly serious when a parameter such as the combustion chamber pressure is to be detected directly at the high-bandwidth combustion chamber. While standards for digital data transmission in automobiles such as CAN are known, their implementation is still quite costly. It would be conceivable, a measured quantity to be detected in However, such an approach is inherently inaccurate and allows higher transmission rates only when short measurement times of the transmitted signal are present analog signal level and thus in turn increased susceptibility to be accepted.

Another known method of transmitting data that is inexpensive to implement and less susceptible to electromagnetic interference is so-called pulse width modulation. In this method, a pulse duration proportional to the binary numerical value of the data word is determined for a digital data word to be transmitted, a pulse with the defined duration is transmitted from a transmitter to a receiver, and the data word is reconstructed on the basis of the pulse duration detected at the receiver.

This known method has two major disadvantages. On the one hand, it is not possible to transmit a pulse representing the numerical value 0, since its duration would be 0. On the other hand, the time required to transmit a data value increases exponentially with the resolution (ie, the number of bits) of the data value. The duration of the shortest pulse that can be transmitted from the transmitter to the receiver is determined by the structure of the transmitter, the receiver and a transmission medium connecting both, so that a magnification tion of accuracy by one bit in each case brings a doubling of the pulse duration with it.

Advantages of the invention

The present invention as defined in claims 1 and 12, a method for data transmission by pulse width modulation and a suitable transmitter are provided, which also allow the transmission of the value 0 by the pulse duration in each case proportional to a through the data word to be transmitted nonnegative integer plus a constant positive markup. Thus, the duration of the transmitted pulse when transmitting the value 0 corresponds only to the numerical value of the surcharge, and a detectable pulse of non-disappearing duration arrives at the receiver.

Preferably, the supplement contains a first term corresponding to a number between 0 and 1, preferably 0.5. If this term is neglected when reconstructing the data word at the receiver, the data word will nevertheless be correctly returned.

Preferably, the pulse to be transmitted is obtained by a digital-time conversion of an extended data word whose bits on the one hand comprises the original data word to be transmitted and at least one bit representative of the first term whose significance is lower than that of the bits of the specified data word is. Alternatively or additionally, the addition preferably also contains a term whose numerical value corresponds to the largest integer plus 1 represented by the original data word. This ensures that the minimum duration of each transmitted pulse is at least equal to this largest representable integer, and extremely short pulses, the transmission of which requires a high bandwidth of the transmission medium, are avoided.

Such a pulse is easily obtainable by digital-to-digital conversion of an extended data word whose bits comprise, on the one hand, the original data word to be transmitted and, in addition, one bit of value 1 whose significance is higher than that of the bits of the original data word ,

Very fast transmission of data values with very high resolution can be achieved by first dividing the bits of the digital data value to be transmitted into several data words and carrying out the transmission by pulse width modulation as described above one after the other for each of the data words thus obtained. This is only possible because the inventive method also allows the transmission of a data word with the value 0. A receiver suitable for this purpose is the subject of claim 16.

In order to save time in the transmission of the multiple data words one after the other, it is expedient to transmit these immediately in succession with alternating levels.

In this case, a time window of predefined length can be provided for each data value in which the data value, divided into an odd number of data words, is transmitted. As a result, at the end of each time window, a fill signal Ü is transmitted whose level is equal to the pulse used to transmit the second data word, so that the same level can be used for the first data words of different data values.

When a data value divided into an even number of data words is transmitted in a time window of predefined length, the first pulses of successive data values, which respectively correspond to their first data words, are transmitted at alternating levels.

The use of a fixed-duration time window for each data value results in the starting times of the first pulses of each data value having a fixed time interval corresponding to the duration of the time window. A receiver is therefore able to recognize them and thus to synchronize themselves on the basis of the periodic return of the beginning of the first pulses.

Description of the embodiments Further features and advantages of the invention will become apparent from the following description of embodiments with reference to the accompanying figures. Show it:

Fig. 1 is a block diagram of a transmitter according to the invention;

2 shows an exemplary course of a transmission signal of the transmitter from FIG. 1,

Fig. 3 is a block diagram of a receiver complementary to the transmitter of Fig. 1;

4 is a block diagram of a transmitter according to a second embodiment;

5 shows an exemplary course of a transmission signal according to a second embodiment of the method;

Fig. 6 is a block diagram of a receiver complementary to the transmitter of Fig. 4;

7 shows an exemplary course of a transmission signal according to a modification of the method;

8 shows a block diagram of a transmitter according to a third embodiment of the invention;

FIG. 9 shows an exemplary course of a transmission signal of the transmitter of FIG. 8; FIG. and 10 shows an exemplary course of a transmission signal according to a further modification of the method.

The transmitter shown schematically in FIG. 1 comprises a sensor, in the present case a pressure sensor 1, which is attached to a combustion chamber of a motor vehicle internal combustion engine to detect its internal pressure, and an analog-to-digital converter 2 which generates an output signal of the pressure sensor 1 converts into a digital data value with a resolution of 12 bits. The bits are labeled in order of increasing significance with a ₀ , ai, ..., an. A multiplexer 3 has three inputs 3-H, 3-M, 3-L and an output 3-0 of 7-bit width. The four most significant bits of each of the three inputs 3-H, 3-M, 3-L are connected to bits on, aio to a ₈ , a ₇ to a ₄ and a ₃ to a ₀ from the converter 2. The bit of next low significance at the three inputs is in each case permanently connected to the level Logic 1, the two bits of lowest significance to Logical 0. The wiring of the individual bits, designated i ₀ to iε in the order of increasing significance, of the three inputs is summarized shown in Table 1 below.

Table 1

Assuming the numerical value of the least significant input bit of the converter 4 connected to one bit from the converter 2 to be 1, the three bits I ₂ , ii and i _{0 represent} a penalty of 0.5.

A 2-bit control signal present at a control input 3-C of the multiplexer 3 determines which one of its inputs 3-H, 3-M, 3-L and 3-N is switched through to the output 3-0. Connected to the output 3-0 is a digital-to-time converter 4 which serves to provide output pulses whose duration is linearly proportional to the numerical value represented by the 7-bit word output by the multiplexer 3, respectively. For this purpose, the digital-time converter 4 comprises a 7-bit counter which counts clocks of a clock generator 5 and a comparator which compares the contents of the counter with the data value input to the converter 4 and a positive one Output level on a transmission line 6, as long as the count value is smaller than the input data value and goes to 0, when the count value is larger. The clock generator 5 also supplies the 2-bit control signal to the control input 3-C which changes its value every 128 clock cycles. The change in the value of the control signal is in each case synchronized with the overflow of the counter of the converter 4 recurring every 128 cycles.

FIG. 2 shows by way of example the transmission signal output at the output of the digital-time converter 4 in the course of time t. Time windows FH, FM, FL of 128 clock periods duration are each symbolized by dashed rectangles. Each of the timeline The FH, FM, FL contains a pulse IH, IM, IL whose duration is equal to the clock period multiplied by the 7-bit binary number applied to the input 3-H, 3-M or 3-L of the multiplexer 3 , For example, assuming one period of the clock generator 5 of 50 ns (corresponding to a clock frequency of 20 MHz), three times 128 x 50 ns = 19.2 μs are required to obtain a 12-bit output value of the analog-digital converter. Transducer 2 to send. For conventional pulse width modulation transmission of the same data value, 2 ¹² × 50 ns = 4096 × 50 ns = 204.8 μs, ie approximately 30 times, would be required.

A receiver for the signal shown in FIG. 2 is shown in FIG. It comprises a clock generator 11 whose clock frequency must match that of the clock generator 5 with sufficient accuracy to avoid errors in restoring the output of the analog-to-digital converter 2 on the receiver side, although the requirements for the accuracy of the period match are not all that high are high, as will become clear in the following. A time-to-digital converter 12 receives on the one hand the pulses supplied by the transducer 4 of the transmitter and on the other hand, the clock signal from the clock generator 11 and counts each during the continuation of a pulse, the clock periods. The count result is a binary value of 7 bits wide. The four most significant bits of the count result are recorded in a shift register 13 having three memory locations 13H, 13M, 13L of 4 bits each. The 3 least significant bits of the count result are discarded. In these 3 bits, the counting result of the converter 12 must therefore be from the corresponding input value of the converter 4 without transmission errors occur, so that deviations between the clock periods of the clock generators 5 and 11, which lead to deviations only in these three bits, can be allowed.

In each case after the reception of 3 pulses by the time-to-digital converter 12, the memory locations 13H, 13M, 13L of the shift register 13 contain the bits at to a ₈ , a ₇ to a ₅ and a ₃ to a _{0 of the} same output value of the analogue output. Digital converter 2. By simultaneously reading out all 3 memory locations, this 12-bit data value is restored on the receiver side.

In order to enable the receiver to distinguish between the differently significant pulses of the same data value, it can be provided that an empty time window is transmitted between time slots FL and FH associated with successive data values, the receiver of which reacts by deleting the shift register 13 , This ensures that the content of the shift register 13 read after the reception of each free pulse belongs to the same data value. The empty time window can be transmitted after every single data value or only after a larger number of data values.

4 shows a second embodiment of a transmitter according to the invention. It differs from the embodiment of FIG. 1 in that the 2 supplied to the control input 3-C of the multiplexer 3 Bit control signal is generated by the digital-to-time converter 4 and its value changes whenever the output of a pulse is completed, and that the pulses of a data value are generated directly in succession with alternating signal levels.

4 shows during the transmission of an output data value of the analog-to-digital converter 2. The transmission of the most significant 4 bits to a _{8 of} this output data corresponding pulse IH lasts from time origin 0 to ti; the pulse IH has the level logic 1. At time ti, the control signal at the input 3-C changes its value, so that the multiplexer 3 the 7 bits a ₇ , a ₆ , a ₅ , a ₄ , 1, 0, 0 outputs and the converter 4 begins to implement this value in the pulse IM, now with level logic 0. The pulse IM ends at the time t ₂ , the control signal at the input 3-C changes again, and the bits (a ₃ , a ₂ , ai, a ₀ , 1, 0, 0) from the converter 4 in the pulse IL with the level logical 1 implemented. With the end of this pulse at the time t ₃ , the transmission of the entire output data of the analog-to-digital converter 2 is completed, and during the remaining duration of the time window F provided for the transmission of this value, the output signal of the converter 4 remains at the value 0.

The duration of the time window F need not be greater than 3 x 2 ⁷ = 384 clock periods of the clock generator 5, since the maximum possible length of the pulses IH, IM, IL each binary "1111100" = 124 Taktperio- is and therefore at the end of the time window F at least 12 clock periods remain with signal level logic 0, which separate the pulse IL from the pulse IH ^{s of} the subsequent time window.

In order to be able to correctly evaluate the pulses at the receiver, the latter must be able to identify the respectively first pulse from a group, which together encode a data value, in this case the pulse IH. For this purpose, the receiver shown schematically in Fig. 6 is additionally provided with a counter 14 which receives the input signal from the transmission line 6 and the clock signal from the clock generator 1 and triggered by a rising edge of the input signal, the number of clocks to the next but one rising edge of the input signal counts. To the output of the counter 14, a register 15 is connected, each of which stores the count 14 output by the counter 14 after completion of a counting cycle. Two inputs of a comparator 16 are connected to the output of the counter 14 and the register 15, respectively, so that the comparator 16 compares respectively the last and the penultimate count. Since the rising edges at the beginning of the pulses IH of successive time windows have a fixed time interval of 3 x 128 clock periods, the counts applied to the inputs of the comparator 16 are always identical when the rising edge triggering the counter 14 is that of the pulse IH is. If the counter has been triggered by the rising edge of the pulse IL, coincidence may possibly occur at the comparator 16; however, once a non- is detected, which must therefore belong to a pulse IL as the last received rising edge, the comparator 16 provides an output signal to the counter 14 and the Schieberegis- ter 13, which resets the counter 14 and clears the contents of the shift register 13. When the counter 14 is triggered again by the subsequent rising edge of the input signal, it can only be the rising edge of a pulse IH and synchronization is established.

In the same way, synchronization can also be achieved if, as assumed above, three consecutive pulses do not encode a data value, but rather a larger odd number of pulses; in the latter case, only several reset operations may be required before synchronization is established.

The achievable with this embodiment, the data rate is higher than in the explained with reference to FIG. 2 transmission method. Although the required length of the time window F corresponds to the sum of the lengths of the time windows FH, FM, FL from FIG. 2, no empty time window is required for synchronization here.

Instead of splitting a data value to be transmitted into three data words to be converted in each case into one pulse, a division into an even number of data words is also possible. FIG. 7 shows the time profile of a transmission signal resulting from such a decomposition in the case of a Decomposition into two data words, each of which is implemented in Impulse IH or IL. Each of the three time windows Fl, F2, F3 shown contains a pulse IH whose length represents the more significant of the two data words, a pulse IL whose level is different from that of the pulse IH and which represents the less significant data word, and a fill pulse R, which fills the remaining time of the window and has the same level as the pulse IH. In order to be able to distinguish the filling pulse R of a time window from the pulse IH of the following one, the levels of identical pulses in successive time windows are different in each case.

Fig. 8 is a block diagram of a third embodiment of a transmitter according to the invention. Sensor 1, analog-to-digital converter 2 and clock generator 5 are the same as described with reference to FIG. 1 and will not be explained again. The multiplexer 3 in Fig. 8 differs from that of Fig. 1 in that its inputs 3-H, 3-M, 3-L are each 8 bits wide, the assignment of the bits iθ to i _{7 of} the three inputs in summarized in Table 2 below.

Table 2:

The digital-time converter 4 thus receives from the multiplexer 3 data words of 8-bit width, their value between binary 10000100 = 132 and binary 11111100 = 252, so that the length of the output from the transducer 4 pulses is between 132 and 252 clock periods. The generation of pulses with a few clock periods duration, which may be strongly attenuated by the transmission line 6 between the transmitter and receiver and therefore may be incorrectly evaluated at the receiver, is avoided in this way.

Fig. 9 illustrates the resulting waveform of the transmit signal: each block IH, IM, IL has a fixed length portion of 128 periods and a variable portion whose length represents the actual information to be transmitted.

A receiver suitable for processing the transmission signal of Fig. 9 has substantially the structure shown in Fig. 6; only the width of the output of the time-to-digital converter 12 is increased from 7 to 8, wherein the most significant bit as well as the three least significant bits of this output is not connected to the shift register 13, because the memory cells here only four Bits are wide.

In all embodiments considered so far, the width-modulated pulses IH, IM, IL were transmitted in a time window F of fixed duration. An increase in the data rate can be achieved by using time windows without a fixed duration, as shown in FIG. 10. This is followed by the three pulses IH, IM, IL of a time window F which code an output value of the analog-to-digital converter 2 a filling pulse R with a fixed length which is greater than the maximum possible length of each of the pulses IH, IM, IL, that is, for example, 128 clock periods in the present case. Due to its large length, the filling pulse can easily be recognized as such on the receiver side, so that a synchronization of the transmitter to the first pulse ICH of each window F does not cause any difficulties. Assuming that the pulses IH, IM, IL take every possible length with equal probability, in the present case the durations can be between four and 124 clock periods, an expected value for their length of 64 clock periods and thus an expected value for the length of the entire window F is 3 * 64 + 128 = 320 clock periods, as opposed to a minimum length of the window F of 384 clock periods in the case of FIG. 5.

Claims

- IT patent claims

1. A process for serial data transmission from a transmitter to a receiver comprising the steps of: a) determining at least one of a representative data to be transmitted value digital data word ((a _0, a i, a _2, a _3), (a _4, .. , a ₇ ); (a ₈ , ..., an)); b) determining a pulse duration which is proportional to a nonnegative integer represented by the data word plus a constant positive penalty; c) sending a pulse (IH, IM, IL) with the fixed pulse duration from the transmitter to the receiver; and d) detecting the pulse duration at the receiver and reconstructing the data word.

2. The method according to claim 1, characterized in that the supplement contains a first term corresponding to a number between 0 and 1, preferably 0.5.

3. Method according to claim 2, characterized in that the pulse (IH, IM, IL) is obtained by a digital-time conversion of an extended data word (i ₀ , ... iε) _ι that in step a) fixed data word ((a ₀ , ai, a ₂ , a ₃ ); (a ₄ , ..., a ₇ ); (a ₈ , •••, an)) and at least one representative of the first term tives bit (i2) as a bit of lower significance than the bits of the specified data word.

4. The method according to any one of the preceding claims, characterized in that the supplement contains a term which corresponds to the largest represented by the set in step a) data word integer plus 1.

5. The method according to claim 4, characterized in that the pulse is obtained by a digital-time conversion of an extended data word (i ₀ , ... i ₇ ), the specified in step a) data word ((a ₀ , ai , a ₂ , a ₃ ), (a ₄ , ..., a ₇ ), (a ₈ , ..., an)) and a bit (i ₇ ) of value 1 and higher significance than the bits of the specified data word includes.

6. The method according to any one of the preceding claims, characterized in that in step a) the bits (a ₀ , β-i, a ₂ , ..., an) of the digital data to be transmitted in a plurality of data words ((a ₀ , β -i, a ₂ , a ₃ ), (a ₄ , ..., SL ₁ ), (a ₈ , ..., an)) and steps b) to d) are carried out for each of the data words thus obtained become.

7. The method according to claim 6, characterized in that the the plurality of data words ((ao, 3-ir a ₂ , SL ₃ ); (a ₄ , ..., a ₇ ); (a ₈ , ..., an )) corresponding pulses (IH, IM, IL) directly be transferred consecutively with alternating levels.

8. The method according to claim 6, 1 or 8, characterized in that each data value in a time window (F) of predetermined length, divided into an odd number of data words, is transmitted.

9. The method of claim 6, 7 or 8, characterized in that each data value in a time window (Fl, F2, F3) of predetermined length, divided into an even number of data words, is transmitted, and that the first pulses (IH ) of consecutive data values are each transmitted at alternating levels.

10. The method according to any one of claims 7 to 9, characterized in that the first pulse

(IH) each data value is started to transmit at the beginning of the time window (F) corresponding to the data value, and that the receiver recognizes the first pulse (IH) of successively transmitted data values at the periodic return of its beginning.

11. The method according to any one of the preceding claims, characterized in that the Ü transmitted data values are pressure measured values, in particular measured values of the combustion chamber pressure in an internal combustion engine.

12. Transmitter for the method according to one of the preceding claims, comprising a digital-time converter (4) for converting a digital data word ((a ₀ , SL ₁ , SL ₂ , SL ₃ ); (SL ₄ ,. SL ₁ ); (SL ₈ , •• -, SL ₁₁ )) into a pulse (IH, IM, IL) whose duration is representative of the numerical value of the data word, characterized in that the duration of the pulse corresponding to the numerical value 0 is equal to, greater than 0.

13. A transmitter according to claim 12, characterized in that the digital-time converter (4) processes an input bit number which is greater than the bit number of the digital data word, wherein input bits (I ₁ , i ₂ , ii, io ) which are not connected to bits of the data word ((a ₀ , SL ₁ , SL ₂ , SL ₃ ), (a ₄ , ..., SL ₁ ), (a ₈ , ..., an)), with are connected to a constant value which is 1 for at least one of these bits.

14. A transmitter according to claim 13, characterized in that the input bits (i ₇ , i ₂ , ii, io) of the digital time converter, which are not connected to the bits of the data word, either sig- nificant or less significant than all Bits of the data word are.

15. Transmitter according to one of claims 12 to 14, characterized in that it comprises a circuit (3) for decomposing a digital data value (a ₀ , SL ₁ , SL ₂ , ..., an) into a plurality of bits of different significance data words containing (a ₀ , SL ₁ , a.2, a ₃ ); (a ₄ , ..., SL ₁ ); (a ₈ , ..., an)).

16. A receiver for the method according to one of the preceding claims, comprising a time-to-digital converter (11) for converting a received pulse (IH, IM, IL) into a digital data word representing a data value proportional to the duration of the pulse CHARACTERIZED IN THAT it comprises a concatenation circuit (13) for concatenating predetermined bits successively from the time-to-digital converter (11) of output data words to a digital data value, wherein at least one least significant bit of the output data words is output from the concatenation circuit (13 ) is neglected.

17. A receiver according to claim 17, characterized in that further at least one bit of highest significance of the output data words of the concatenation circuit (13) is neglected.