Method and device for digital compensation of dynamic distortion in high-speed transmitters

Abstract

A transmitter includes a shift register, a lookup table, and a digital to analog converter. The shift register is configured to receive an input signal and to output delayed copies of the input signal. The lookup table is configured to store compensation values estimated based on the input signal and the delayed copies of the input signal. The digital to analog converter is configured to output a transmit signal based on the input signal and the compensation values. The compensation values are designed to mitigate distortion of the transmit signal from conversion of the input signal to a digital signal.

Claims

1. A transmitter comprising: a shift register configured to receive an input signal and to output delayed copies of the input signal; a lookup table configured to store compensation values estimated based on the input signal and the delayed copies of the input signal; and a digital to analog converter configured to output a transmit signal based on the input signal and the compensation values, the compensation values designed to mitigate distortion of the transmit signal from conversion of the input signal to a digital signal.

2. The transmitter of claim 1 wherein the digital to analog converter is configured to scale an amplitude of the transmit signal based on the compensation values.

3. The transmitter of claim 1 wherein the digital to analog converter is configured to output the transmit signal having a frequency response that is generated based on the compensation values.

4. The transmitter of claim 1 wherein the input signal includes symbols for transmission, and wherein the digital to analog converter is configured to output the transmit signal including the symbols having symbol levels generated based on the compensation values.

5. The transmitter of claim 1 wherein the digital to analog converter is configured to generate the transmit signal having an improved signal to noise and distortion ratio by mitigating the distortion in the transmit signal based on the compensation values.

6. The transmitter of claim 1 further comprising a digital signal processor configured to estimate the compensation values based on the input signal and the delayed copies of the input signal.

7. The transmitter of claim 1 further comprising a digital signal processor configured to: populate the lookup table with initial values to generate an output at the digital to analog converter, the output spanning a full range of the digital to analog converter output; determine an average value of the output of the digital to analog converter based on samples taken from the output; determine desired values for the output of the digital to analog converter, the desired values representing desired signal characteristics for the transmit signal; and estimate the compensation values based on the average values and the desired values to mitigate the distortion in the transmit signal.

8. The transmitter of claim 7 wherein the digital signal processor is configured to calculate the desired values using one or more of linear equalization, modulation, and amplitude scaling of the output of the digital to analog converter to prevent clipping of the transmit signal.

9. The transmitter of claim 1 further comprising a digital signal processor configured to iteratively estimate the compensation values based on the input signal and the delayed copies of the input signal and to periodically update the lookup table in response to iteratively estimated compensation values.

10. A method comprising: receiving an input signal; generating delayed copies of the input signal; storing in a lookup table compensation values estimated based on the input signal and the delayed copies of the input signal; and generating a transmit signal based on the input signal and the compensation values, the compensation values designed to mitigate distortion of the transmit signal from conversion of the input signal to a digital signal.

11. The method of claim 10 further comprising scaling an amplitude of the transmit signal based on the compensation values.

12. The method of claim 10 further comprising generating the transmit signal having a frequency response that is generated based on the compensation values.

13. The method of claim 10 wherein the input signal includes symbols for transmission, the method further comprising generating the transmit signal including the symbols having symbol levels generated based on the compensation values.

14. The method of claim 10 further comprising generating the transmit signal having an improved signal to noise and distortion ratio by mitigating the distortion in the transmit signal based on the compensation values.

15. The method of claim 10 further comprising estimating the compensation values based on the input signal and the delayed copies of the input signal.

16. The method of claim 10 further comprising: populating the lookup table with initial values to generate an output at a digital to analog converter, the output spanning a full range of the digital to analog converter output; determining an average value of the output of the digital to analog converter based on samples taken from the output; determining desired values for the output of the digital to analog converter, the desired values representing desired signal characteristics for the transmit signal; and estimating the compensation values based on the average values and the desired values to mitigate the distortion in the transmit signal.

17. The method of claim 16 further comprising calculating the desired values using one or more of linear equalization, modulation, and amplitude scaling of the output of the digital to analog converter to prevent clipping of the transmit signal.

18. The method of claim 10 further comprising: iteratively estimating the compensation values based on the input signal and the delayed copies of the input signal; and periodically updating the lookup table in response to iteratively estimated compensation values.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The following diagrams are merely examples, which should not unduly limit the scope of the claims herein. One of ordinary skill in the art would recognize many other variations, modifications, and alternatives. It is also understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this process and scope of the appended claims.

(2) FIG. 1 is an example 4-level pulse amplitude modulation (PAM4) eye diagram showing the effects dynamic distortion.

(3) FIG. 2 is an example PAM4 eye diagram showing the effects of dynamic distortion compensation according to an example of the present invention.

(4) FIG. 3A is a simplified block diagram illustrating a conventional transmitter device implemented as a digital-to-analog converter (DAC).

(5) FIG. 3B is an example eye diagram illustrating corresponding to the transmitter device shown in FIG. 3A.

(6) FIG. 4 is simplified block diagram illustrating a transmitter device according to an example of the present invention.

(7) FIGS. 5A and 5B are simplified flow diagrams illustrating a method for digital compensation of dynamic distortion according to an example of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

(8) The present invention generally relates to communication systems and integrated circuit (IC) devices. More specifically, the present invention provides for a device and method of operation for digital compensation of dynamic distortion in high-speed transmitters. Merely by way of example, the present invention is applied to a 4-level pulse amplitude modulation (PAM) transmitter. However, the present invention has a much broader range of applicability, such as other PAM-type transmitters, quadrature amplitude modulation (QAM) transmitters, amplitude modulation (AM) type transmitters, AM transceivers, AM communication systems, and the like.

(9) As explained above, high-speed transmitters can be affected by numerous impairments such as distortion, limited bandwidth, insertion losses, reflections, and noise. In some cases, such as in large-signal single-ended transmitters, the distortion depends not only on the symbol being transmitted, but also on previously and subsequently transmitted symbols (i.e., precursors and postcursors, respectively). We call this type of distortion dynamic distortion, and we note that it is different from static distortion, which depends only on the symbol being transmitted. Dynamic distortion is also different from linear impairments, which may introduce linear dependence on precursors and postcursors but will not introduce distortion. As a result, dynamic distortion cannot be addressed with established techniques, such as static predistortion or equalization.

(10) The following description is presented to enable one of ordinary skill in the art to make and use the invention and to incorporate it in the context of particular applications. Various modifications, as well as a variety of uses in different applications will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to a wide range of embodiments. Thus, the present invention is not intended to be limited to the embodiments presented but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

(11) In the following detailed description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. However, it will be apparent to one skilled in the art that the present invention may be practiced without necessarily being limited to these specific details. In other instances, well-known structures and devices are shown in block diagram form, rather than in detail, in order to avoid obscuring the present invention.

(12) The reader's attention is directed to all papers and documents which are filed concurrently with this specification and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference. All the features disclosed in this specification, (including any accompanying claims, abstract, and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features.

(13) Furthermore, any element in a claim that does not explicitly state means for performing a specified function, or step for performing a specific function, is not to be interpreted as a means or step clause as specified in 35 U.S.C. Section 112, Paragraph 6. In particular, the use of step of or act of in the Claims herein is not intended to invoke the provisions of 35 U.S.C. 112, Paragraph 6.

(14) Please note, if used, the labels left, right, front, back, top, bottom, forward, reverse, clockwise and counter clockwise have been used for convenience purposes only and are not intended to imply any particular fixed direction. Instead, they are used to reflect relative locations and/or directions between various portions of an object.

(15) FIG. 1 is an example 4-level pulse amplitude modulation (PAM4) eye diagram showing the effects dynamic distortion. An extreme case for a PAM4 transmitter is shown in eye diagram 100 where the falling edge is much slower than the rising edge, which results in the bottom PAM4 eye being closed.

(16) FIG. 2 is an example PAM4 eye diagram showing the effects of dynamic distortion compensation according to an example of the present invention. For comparison purposes, the transmit eye in eye diagram 200 has much less dynamic distortion and is more likely to provide good performance for a data link.

(17) As shown by the comparisons of FIGS. 1 and 2, it is highly desirable to implement compensation techniques to address dynamic distortion in transmitter devices. This distortion can sometimes be addressed with analog techniques at the cost of increased power consumption, but often such analog techniques are not effective at very high frequencies. Thus, the present invention uses digital compensation techniques that will be effective for high-speed transmitters operating at very high frequencies.

(18) In an example, the present invention provides a digital compensation method to remove dynamic distortion from the transmitter output. This method can be further enhanced to introduce certain desirable characteristics into the output signal, such as precursor or postcursor equalization, eye modulation, amplitude scaling, or the like. Merely by way of example, this method assumes a PAM4 modulation scheme. However, this method can be applied to transmitters using any type of amplitude modulation (AM) scheme, such as other types of PAM schemes, quadrature amplitude modulation (QAM), and the like. An example derivation of the digital compensation method is discussed in further detail below.

(19) FIG. 3A is a simplified block diagram illustrating a conventional transmitter device 301. Without loss of generality, let us assume that the transmitter 301 is implemented as a digital-to-analog converter (DAC) 310, which turns input digital codes x.sub.n into an output waveform y(t). FIG. 3B is an eye diagram 302 corresponding to the output of transmitter 301. This output waveform y(t) can then be sampled at times to in the middle of the eye to yield samples y.sub.n (as shown in FIG. 3B). Also, without loss of generality, let us assume that the transmitter 301 is sending PAM4 signals.

(20) Ideally, the output voltage y.sub.n at time to should only be a function of the input code x.sub.n:
y.sub.n=y(x.sub.n)

(21) In practice, the output voltage depends on multiple precursors and postcursors:
y.sub.n=y[x.sub.n?1,x.sub.n?l+1, . . . ,x.sub.n, . . . ,x.sub.n+k?1,x.sub.n+k?]+?.sub.n

(22) Here, x.sub.n?1, x.sub.n?l+1, . . . x.sub.n?1 are precursors and x.sub.n+1, . . . , x.sub.n+k?1, x.sub.n+k are postcursors, and ?.sub.n is the residual error, which can be due to noise or other impairments not correlated with the signal. Without loss of generality, let us assume that the output voltage is a function of a single precursor and a single postcursor:
y.sub.n=y[x.sub.n?1,x.sub.n,x.sub.n+1]+?.sub.n

(23) To compensate for dynamic distortion, we need a method that removes the part of the distortion that is dependent on x.sub.t?1 and x.sub.n+1 (i.e., the precursor and postcursor components, respectively). According to an example of the present invention, the method can involve calculating the average value of each possible sequence in the transmission signal.

(24) For each possible sequence of three symbols at the input (X.sub.?1, X.sub.0, X.sub.+1), let us define y[X.sub.?1,X.sub.0,X.sub.+1] as the average value of y.sub.n when x.sub.n?1=X.sub.?1, x.sub.n=X.sub.0, and x.sub.n+1=X.sub.+1. For example, suppose we have a long continuous-time output y(t) that we sample in the middle of the eye to obtain samples y.sub.n (as shown in FIG. 3B). From the sampled output y.sub.n, we can select just instances where three consecutive samples correspond to input values (X.sub.?1, X.sub.0, X.sub.+1) and for these instances we can average the cursor output y.sub.n to calculate y[X.sub.?1,X.sub.0,X.sub.+1]. As an example, in the case of PAM4 transmission with no equalization, each of X.sub.?1, X.sub.0, and X.sub.+1 can take one of four possible values. Hence, there are 4.sup.3=64 possible sequences, and we can define an average value y[X.sub.?1,X.sub.0,X.sub.+1] for each of these 4.sup.3 sequences.

(25) Using these average values, we need to change the output signal to remove undesirable characteristics (i.e., dynamic distortion) and introduce other desirable characteristics (precursor or postcursor equalization, eye modulation, or amplitude scaling). Let us define the desired output voltage values that have said desirable characteristics as {tilde over (y)}[X.sub.?1,X.sub.0,X.sub.+1]. Preferably, we should generate {tilde over (y)}[X.sub.?1,X.sub.0,X.sub.+1] without modifying the analog circuitry since that would entail increased power and design complexity and the desired performance may not even be achievable for very high data rates. Instead, we can adjust the input values to the DAC in such a way as to induce the required changes on the transmit signal at the DAC output. This can be done with a shift register and a look-up table (LUT), as shown in FIG. 4.

(26) FIG. 4 is a simplified block diagram illustrating a transmitter device according to an example of the present invention. As shown, device 400 includes a DAC 410, an LUT 420, and shift registers 431, 432. The LUT 420 is coupled to the DAC 410 and configured to receive an input signal along with delayed input signals from the shift registers 431, 432.

(27) In an example, the input is passed through the shift registers 431, 432 to generate delayed copies of the signal (x.sub.n?1, x.sub.n, and x.sub.n+1), which are then used to index the LUT 420. Through this indexing, the LUT 420 is configured with values or codes that can compensate for dynamic distortion and introduce said desirable characteristics. In a specific example, for each possible input sequence (X.sub.?1, X.sub.0, X.sub.+1), the LUT stores codes that cancel the effects of dynamic distortion in the DAC 410.

(28) In other examples, the transmitter device 400 can include additional circuit components configured before or after the LUT-DAC configuration. Additional circuit components may be connected between the LUT and the DAC as well. The circuit components can include various filters, amplifiers, mixers, and the like and combinations thereof. In such cases, the functionality of the LUT-DAC configuration remains the same. These examples of the transmitter device 400 can also be configured within a transceiver device or in a broader communication system or the like. Those of ordinary skill in the art will recognize variations, modifications, and alternatives to the configuration shown in FIG. 4.

(29) Although LUTs are used to apply equalization, eye modulation or amplitude scaling, examples of the present invention use the LUT to perform more powerful signal processing, including dynamic distortion compensation.

(30) For instance, linear equalization can be implemented as:
z[X.sub.?1,X.sub.0,X.sub.+1]=a.sub.?1.Math.X.sub.?1+a.sub.0.Math.X.sub.0+a.sub.1.Math.X.sub.+1
and eye modulation and amplitude scaling can be implemented as:
z[X.sub.?1,X.sub.0,X.sub.+1]=c(X.sub.0)
These two equations can be combined into the following equation:
z[X.sub.?1,X.sub.0,X.sub.+1]=a.sub.?1.Math.c(X.sub.?1)+a.sub.0.Math.c(X.sub.0)+a.sub.1.Math.c(X.sub.+1)

(31) From this equation, we can see that linear equalization, eye modulation, and amplitude scaling introduce certain relationships between the elements of the LUT. For example, for eye modulation all z[X.sub.?1,X.sub.0,X.sub.+1] entries for a given X.sub.0 value are identical. On the other hand, for dynamic distortion compensation, it is important that each LUT entry is independently set.

(32) Let us first calculate the desired output voltage {tilde over (y)}[X.sub.?1,X.sub.0,X.sub.+1] including the required equalization, eye modulation, and amplitude scaling, using an equation similar to the one above:
{tilde over (y)}[X.sub.?1,X.sub.0,X.sub.+1]=a.sub.?1.Math.c(X.sub.?1)+a.sub.0.Math.c(X.sub.0)+a.sub.1.Math.c(X.sub.+1)

(33) The desired output must be shifted and scaled to make use of the available range of the DAC without clipping. For example, assuming a 7-bit DAC, no codes can be lower than 0 and no codes can be higher than 127. Therefore, we need to identify the sequences most likely to clip. Assuming that the output can have both positive and negative values, clipping at the upper end is most likely to occur for the sequence that has the main cursor (X.sub.0) set to the top PAM4 level. In addition, this sequence has the maximum ratio {tilde over (y)}[X.sub.?1,X.sub.0,X.sub.+1]/f[X.sub.?1,X.sub.0,X.sub.+1]. Let us define y.sub.max=y[X.sub.?1,X.sub.0,X.sub.+1] for this sequence most likely to clip at the upper end.

(34) Similarly, clipping at the lower end is most likely to occur for the sequence that has the main cursor (X.sub.0) set to the bottom PAM4 level. This sequence also has the maximum ratio {tilde over (y)}[X.sub.?1,X.sub.0,X.sub.+1]/y[X.sub.?1,X.sub.0,X.sub.+1]. Let us define y.sub.min=y[X.sub.?1,X.sub.0,X.sub.+1] for this sequence most likely to clip at the lower end.

(35) Once y.sub.max and y.sub.min are identified, the signal can be easily shifted and scaled to fit the available range without clipping as follows:

(36) $\hat{y}\left[X_{-1},X_0,X_{+1}\right]=\frac{S.Math.\tilde{y}\left[X_{-1},X_0,X_{+1}\right]}{\max \left(\tilde{y}\left[X_{-1},X_0,X_{+1}\right]\right)-\min \left(\tilde{y}\left[X_{-1},X_0,X_{+1}\right]\right)}.Math.\left(y_{\max }-y_{\min }\right)+\frac{1}{2}.Math.\left(y_{\max }+y_{\min }\right)$
where S is the desired scaling factor and ?[X.sub.?1,X.sub.0,X.sub.+1] is the desired output voltage for each sequence, after scaling and shifting.

(37) The next step is to estimate the required LUT entries z[X.sub.?1,X.sub.0,X.sub.+1] that take the output from the initial y[X.sub.?1,X.sub.0,X.sub.+1] values to the desired values ?[X.sub.?1,X.sub.0,X.sub.+1]. In an example, this can be done using the LMS algorithm as follows:
z.sub.k+1[X.sub.?1,X.sub.0,X.sub.+1]=z.sub.k[X.sub.?1,X.sub.0,X.sub.+1]+?.Math.(y[X.sub.?1,X.sub.0,X.sub.+1]?(y[X.sub.?1,X.sub.0,X.sub.+1])
where ? is the convergence coefficient.

(38) FIGS. 5A and 5B are simplified flow diagrams illustrating a method for digital compensation of dynamic distortion according to an example of the present invention. As shown, flow diagram 501 of FIG. 5A connects to flow diagram 502 of FIG. 5B by the connector A. From the steps described above, we can now define the full procedure 10 according to an example of the present invention as follows: 1. Fill the LUT with a set of starting values configured to produce valid waveforms at the output of the DAC (flow diagram step 511). Generally, it is best to exercise the whole range of the DAC at this step. For example, in the case of using a PAM4 coding scheme with a 7-bit DAC, we can fill the LUT with z[*, X.sub.0,*]=0/43/85/127. 2. Send an input signal pattern x.sub.n through the DAC (flow diagram step 512). This signal pattern can be any random signal pattern (e.g., higher-order PRBS for PAM4). 3. Capture the output waveform with a recording device (flow diagram step 513). This recording device can be an on-chip receiver or test equipment, such as a digital scope, or the like. 4. Determine the average output value y[X.sub.?1,X.sub.0,X.sub.+1] for each possible sequence of three symbols at the input (flow diagram step 514). These average output values can be sampled from the middle of the eye of the output waveform. 5. Determine the desired output Y[X.sub.?1,X.sub.0,X.sub.+1] for each of these possible sequences (flow diagram step 515). The desired output calculation can include techniques for linear equalization, eye modulation, amplitude scaling, and the like and combinations thereof. 6. Adjust the desired output to ?[X.sub.?1,X.sub.0,X.sub.+1] for all possible sequences to eliminate the risk of clipping (flow diagram step 516). This adjustment includes determining the sequences most likely to clip at the upper end and the lower end. Afterwards, the adjusted desired output can be determined by the calculation discussed previously. 7. Estimate the next iteration of LUT entries z.sub.k+1[X.sub.?1,X.sub.0,X.sub.+1] that will take the output from the initial average output to the adjusted desired output (flow diagram step 517). As discussed above, the LUT entries can be estimated using an LMS algorithm. However, this step may use other algorithms, such as recursive least squares (RLS), Affine Projection (AP), and the like. 8. Send the same input signal pattern x.sub.n as before through the DAC, as described in step 2 (flow diagram step 518). 9. Capture the output waveform using the recording device, as described in step 3 (flow diagram step 519). 10. Check whether the output waveform meets one or more check conditions (flow diagram step 520). These check conditions can include whether the output waveform has low enough distortion (e.g., by measuring signal to noise and distortion ratio (SNDR)) and additionally meets the requirements of equalization, eye modulation and scaling set out in the beginning. Other check conditions can include eye height (i.e., eye opening), bit error rate (BER), level separation mismatch ratio (RLM), transmission dispersion and eye closure quaternary (TDECQ), and the like. If the desired check conditions are not satisfied, the process can be repeated starting from step 4 (shown by flow diagram steps 521, 522, 523, and 524). In practice, iterations may be required because dynamic distortion changes as the LUT is adapted. 11. Perform other steps, as desired (flow diagram step 530).

(39) The above sequence of steps is used to perform a method for digital compensation of dynamic distortion according to one or more embodiments of the present invention. Depending upon the embodiment, one or more of these steps can be combined, or removed, or other steps may be added without departing from the scope of the claims herein. One of ordinary skill in the art will recognize variations, modifications, and alternatives.

(40) In an example, the method steps described above can be performed via hardware by a digital signal processor (DSP) electrically coupled to the LUT, the DAC, and the recording device. The DSP can also be programmed to perform these method steps via firmware stored in a non-volatile memory (NVM) device (e.g., read-only memory (ROM), erasable read-only memory (EPROM), flash memory, etc.) configured within the transmitter device or encompassing communication device. The DSP can also be implemented directly in the recording device (e.g., in the receiver) and the LUT updates can be communicated back to the transmitter device by means of a back channel, where available. Or, these methods steps can also be performed by software, such as by a computing system with test equipment (e.g., digital scope) connected to the transmitter device to implement the steps described previously. Of course, there can be other variations, modifications, and alternatives.

(41) While the above is a full description of the specific embodiments, various modifications, alternative constructions and equivalents may be used. Therefore, the above description and illustrations should not be taken as limiting the scope of the present invention which is defined by the appended claims.