CA2004503A1 - Dynamic equalization for partial response channels - Google Patents
Dynamic equalization for partial response channelsInfo
- Publication number
- CA2004503A1 CA2004503A1 CA 2004503 CA2004503A CA2004503A1 CA 2004503 A1 CA2004503 A1 CA 2004503A1 CA 2004503 CA2004503 CA 2004503 CA 2004503 A CA2004503 A CA 2004503A CA 2004503 A1 CA2004503 A1 CA 2004503A1
- Authority
- CA
- Canada
- Prior art keywords
- signal
- response
- output signal
- filter
- filtered
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Landscapes
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Abstract
ABSTRACT
In a data recovery system comprising a receiver filter for match filtering and equalizing a signal received from a data transmission or storage channel, and a combined dicode and partial response filter for automatic gain controlling and subsequent detection of the matched filtered equalized data signal and in response generating an output signal, the improvement comprising a combination of peak/power detector for detecting amplitude of the output signal and in response generating an amplitude/power output signal, a lowpass filter for filtering the amplitude/power output signal and in response generating a filtered output signal, an averager for averaging the amplitude/power output signal and in response generating an averaged amplitude/power output signal, a differential amplifier for subtracting the averaged power output signal from the filtered output signal and in response generating a difference signal, and an equalizer connected to an output of the receiver filter and an input of the combined dicode and partial response filter for receiving the difference signal and in response dynamically equalizing the equalized and matched filtered data signal in accordance with dynamic channel response variations in the combined dicode and partial response filter.
In a data recovery system comprising a receiver filter for match filtering and equalizing a signal received from a data transmission or storage channel, and a combined dicode and partial response filter for automatic gain controlling and subsequent detection of the matched filtered equalized data signal and in response generating an output signal, the improvement comprising a combination of peak/power detector for detecting amplitude of the output signal and in response generating an amplitude/power output signal, a lowpass filter for filtering the amplitude/power output signal and in response generating a filtered output signal, an averager for averaging the amplitude/power output signal and in response generating an averaged amplitude/power output signal, a differential amplifier for subtracting the averaged power output signal from the filtered output signal and in response generating a difference signal, and an equalizer connected to an output of the receiver filter and an input of the combined dicode and partial response filter for receiving the difference signal and in response dynamically equalizing the equalized and matched filtered data signal in accordance with dynamic channel response variations in the combined dicode and partial response filter.
Description
20(~450;~
DYNAMIC EOUALIZATION FOR PARTIAL RESPONSE CHANNELS
Field of the Invention This invention relates in general to data recovery systems, and more particularly to dynamic equalization for partial response channels within a data recovery system.
Backaround of the In~ention Optimum performanoe of data recovery hardware in high efficiency data transmission channels or high - 10 density data recording channels requires that non-ideal channel behavior (e.g. linear and non-linear distortion, time base errors, etc) be identified and corrected. For example, automatic gain control (AGC) can be used to compensate for non-dispersive channel fading, and adaptive egualization is commonly used in the various channels to correct stationary equalization errors.
Brief Description of the Drawinas -The present invention will be described in greater detail below, along with a discussion of the relevant prior art, in conjunction with the following drawings in which:
Figure 1 is a block schematic diagram of a general partial response channel in accordance with the prior art;
Figure 2 is a block schematic diagram of a dicode~
partial response filter in accordance with the prior art;
Figure 3 is a block diagram of a combined dicode and partial response detector in accordance with the prior art:
Figure 4 is block diagram of a dynamic equalization system for use with combined dicode and partial response detector in accordance with the preferred embodiment of the present invention;
Figure 5 i5 a blocX diagram of a dynamic equalization system in accordance with an alternative embodiment of the present invention; and Figures 6a and 6b are photomicrographs of the experimental data showing output signals of a combined dicode and partial response detector with and without dynamic equalization.
Description of the Prior Art Compensation for non-ideal channel characteristics is particularly important when partial response detection is employed. A general partial response channel is shown in Figure 1 comprising a transmission or recording channel 1 for receiving a source data signal and a receiver filter 2 for processing the signal of the channel output. Typically, the receiver filter 2 comprises a matched filter and equalizer for minimizing noise and removing linear distortion from the signal. A
sampler 3 is provided for sampling the signal output from receiver filter 2 at a sampling rate fb = 1/T (i.e.
using the recovered data clock signal).
Since equalization of the channel 1 via receiver filter 2 may enhance additive noise within the received signal, the signal is then passed through a partial response filter shown generally as 4 which distorts the equalized signal in a known fashion in order to minimize the additive noise. The partial response filter 4 is implemented as a series of parallel connected signal multipliers 6, 8...10 having respective outputs connected to a summer 16, and being separated by respective unit delay circuits 12...14. The output of summer 16 is connected to a decoder ~not shown) for decoding of the filtered signal with the known distortion accommodated for, in a well known manner.
The partial r~esponse filter 4 may be characterized by the equation ~ ~ D' , where D
~-~
represents the delay operator for unit delay circuits 12...14.
Turning to Figure 2, a dicode partial response filter 18 is shown comprising a pair of parallel connected multipliers 20 and 22 separated by a unit delay circ~it 24. In operation, the matched filtered, equalized signal from filter 2 is multipli d by a gain of 1 in multiplier 20 and the single clock period delayed signal output from unit delay circuit 24 is multiplied by a gain of -1 (e.g. inverted) via multiplier 22 and summed with the original signal in summer 26. In other words, the delayed signal is subtracted from the original signal. Hence, the dicode filter 18 may be characterized by the discrete time equation F(D~ D, or in the frequency domain as T(w) = 2j sin(wT/2), where w = radian frequency.
Thus, whenever there is a transition in the source data signal, a pulse appears at the output of summer 26 in the dicode filter 18. The polarity of the pulse indicates the direction of the transition.
The original data signal can then be decoded by sampling the signal at the output of summer 26 with the - data clock (fb = l/T) and decoding a data transition when the signal exceeds a predetermined detection threshold. This entire operation may be referred to as threshold detection, and is represented by the threshold decoder 28 within Figure 2.
An advantage of using dicode partial response is that no dc response is required in the transmission or recording channel 1. However, in order to ensure accurate data recovery, the gain of the channel 1 must be controlled to ensure that the detection threshold is accurately set within threshold decoder 28. Typically, the channel gain is controlled to make the peaks of the positive and negative pulses approximately constant in amplitude.
A disadvantage of dicode partial response is that it requires a high signal to noise ratio (SNR) near the Nyquist freguency (i.e. one half of sampling frequency fb). In channels where the high frequency SNR is poor, higher order partial response filters are often used.
For example, if the dicode (1-D) filter 18 is followed by a duobinary (l+D) filter, a bandpass characteristic is created, in which frequencies near Nyquist as well as DC are attenuated. ~he combined (l-D) and (~+D) filters yields an overall (1-D2) characteristic known as Class IV partial response. Again, gain control is required in such a~configuration for detection of the data signals since the signal may be positive, negative or zero.
However, since the signal peaks output from the duobinary filter are much less uniform in amplitude as compared wi~h the dicode signal, depending on the source data, gain control aimed at controlling the average value of the peaks of the signal has been found to be typically of low accuracy. In general, the higher the order of the partial response filter polynomial, the more the signal characteristics depend on the source data, making signal estimation prior to detection more difficult.
The advantages of dicode filtering for automatic gain control (AGC) and Class IV partial response filtering for detection, can be combined as shown in Figure 3. The dicode signal output from filter 18 is applied to an automatic gain control circuit 30 and the output of the duobinary or other partial response filter 32 then enters the decoder for actual detection of the filtered source data signal. In this case, the partial response filter may be characterized by the following frequency domain equation:
P(w) = 2T Cos(wT/2) Thus, the frequency domain representation for the combined dicode and partial response detector of Figure 3 may be characterized as follows:
T(w) = j2T sin(wT/2)P(w) = j2T sin(wT) 5The above equations are valid for Class IV partial response filters, although it also contemplated that other forms of partial response filter may be used.
The system of Figure 3 is adequate for time-stationary channels; however, the system has been found to be inadequate when the response of the channel 1 changes dynamically. For example, in a recording channel, an increase in the effective distance between the recording head and the recording medium causes much greater attenuation at higher than lower frequencies.
As a result, loss of signal amplitude in the dicode signal will be greater than in the Class IV signal if dynamic increases in head-to-medium spacing occur.
Correction of channel gain using the dicode signal will then cause an increase in signal level as observed at the output of the partial response filter 32.
Assuming perfect AGC in the dicode channel, variations in signal amplitude measured at the output of the second partial response filter 32 can be attributed to source data or to variations in the response of the channel.
Summary of the Invention In the latter case, the variations in signal amplitude measured at the output of the partial response ~ilter can be used to correct the dynamic response changes by controlling the amount of response correction of an appropriate equalizer.
Thus, in accordance with an aspect of the present invention, there is provided in a data recovery system comprising a receiver for match filtering and equalizing a signal received from a data transmission or storage channel, and a combined dicode and partial response 2004~i03 filter for automatic gain controlling and filtering for detection of said equalized and matched filtered data signal and in response generating an output signal, the improvement comprising:
a) means for detecting amplitude/power of said output signal and in response generating an amplitude/power output signal;
b) means for lowpass filtering said amplitude/power output signal and in response generating a filtered output signal:
c) means for averaging said amplitude/power output signal and in response generating an averaged amplitude/power output signal;
d) means for subtracting said averaged amplitude/power output signal from said filtered output signal and in response generating an amplified difference signal; and e) equalization means connected to an output of said receiver filter and an input of said combined dicode and partial response filter for receiving said difference signal and in response dynamically equalizing said equalized and matched filtered data signal in accordance with dynamic channel response variations in said combined dicode and partial response filter.
In accordance with a further aspect of the present invention, there is provided in a data recovery system comprising a receiver for match filtering and equalizing an input signal, a combined dicode and partial response filter for automatic gain controlling and filtering for detection of said equalized and matched filtered data signal and in response generating an output signal, and a decoder for receiving and decoding said output signal and in response generating a decoded data signal, the improvement comprising:
a) a second combined dicode partial response filter for receiving and filtering said decoded data signal and in response generating a filtered decoded data signal;
b) first power detector means for detecting power of said output signal and in response generating a first power output signal;
c) second power detector means for detecting power of said filtered decoded data signal and in response generating a second power output signal;
d) first lowpass filter means for lowpass filtering said first power output signal and in response generating a first filtered signal;
e) second lowpass filter means for lowpass filtering said second power output signal and in response generating a second filtered signal;
f) delay circuit means for delaying said first filtered signal so as to be substantially in phase with said second filtered signal, thereby : 20 compensating for time delays within said decoder;
g) first subtraction means for subtracting said second filtered signal from said first delayed filtered signal and in response generating a first amplified difference signal;
h) third lowpass filter means for lowpass filtering said first difference signal and in response generating a third filtered signal;
i) means for averaging said first difference signal and in response generating an averaged signal;
j) second subtracting means for subtracting said averaged signal from said third filtered signal and in response generating a second amplified difference signal; and k) egualizer means connected to an output of said recei~er filter and an input of said combined dicode and partial response filter for recei~ing ;~00~503 said second difference signal and in response dynamically equalizing said equalized and makched filtered data signal an accordance with variations in the statistics of said input data signal.
Detailed Description of the Invention Turning to Figure 4, a dynamic equalizer 3~ is shown for simultaneously correcting variations in signal amplitude and dispersion within the transmission or recording channel. In particular, the power or peak amplitude of partial response filter 32 is detected within peak/power detector 36. The signal output from detector 36 is then lowpass fi~tered via filter 38 and averaged via amplitude averager 4~. The averaged signal output from circuit 40 is then subtracted from the lowpass filtered signal output from filter 38 via a differential amplifier 42, and the output signal VC is then applied to dynamic equalizer 34 as a voltage control feedback signal to correct for variations in channel response.
The equalizer shown by way of example 34 may be characterized ~y the frequency domain equation ek(Vc w) where k is a constant proportional to the recording channel loss constant known as "spacing loss".
Provided that equalizer 34 exhibits frequency characteristics appropriate to the cause of the dynamic frequency variations within the channel, the change in response can be adequately corrected by the embodiment of Figure 4. Thus, the embodiment of Figure 4 may be used to compensate for variations in spacing loss in a 0 magnetic recording channel, as discussed above.
FurthermDre, provided that the lowpass filter 38 in the feedback loop has a sufficiently narrow bandwidth, variations of signal level measured at the output of the second partial response filter (i.e.
filter 32) attributable to variations in the statistics ~:00450~
of the source data should be small, assuming the source data is approximately random.
However, if a wider filter is needed to compensate for rapid variations in frequency characteristics, the changing signal statistics attributable to changing data source statistics can present a problem.
Thus, in accordance with the altexnative embodiment of Figure 5, a compensation circuit is provided using feedback from the decoded data output from a decoder 43 which can be a threshold or other decoder.
More particularly, the decoded data signal is applied to a further combined dicode and second partial response filter 44 which is equivalent to the combined dicode and partial response filter 19. The power of the signal output from filter 44 is detected via power detector 46 similar to the peak/power detector 36 depicted in Figure 4. The power output signals from detectors 36 and 46 are filtered via respective lowpass filters 48 and 50, and the signal output from lowpass filter 48 is delayed via delay circuit 52 which compensates for delays inherent within decoder 43.
The filtered power signal output from lowpass filter 50 is subtracted from the delayed filtered power signal output from delay circuit 52 via a further differential amplifier 54, and the voltage control output signal vc' from amplifier 54 is applied to low pass filter 38 and averager 40 as discussed above in connection with the embodiment of Figure 4. Similarly, the lowpass filter 38 and averager 40 are connected to non-inverting and inverting inputs respectively of differential amplifier 42 which, in turn, is connected to a control voltage input of dynamic equalizer 34.
The power of the signal output from the second combined dicode and partial response filter 44 should be the same as the power output from partial response filter 32 under ideal conditions, except for the delay 20~)4503 through the decoder 43. Thus, in accordance with the embodiment of Figure 5, only the statistics of the source data signal will be detected and differentiated within amplifier ~4, and thereby reduced.
Finally, turning to Figures 6a and 6b, experimental data is shown from tests conducted on a rotary magnetic recorder. At the end of the video scan, there is a loss of signal which is believed to be attributable to decreasing head to tape contact as the head leaves the trailing edge of the tape. The signal in the upper trace of Figure 6a shows the effect of such signal loss after gain control has been applied using the dicode signal. Figure 6b shows the Class IV signal after dynamic equalization has been applied in accordance with the present invention. The envelope of the signal in Figure 6b is seen to be approximately flat, indicating that the equalization has effectively eliminated the error resulting from the aforementioned increased spacing loss.
Additional modifications and variations of the present invention are possible. All such modifications and embodiments are believed to be within the sphere and scope of the present invention as defined by the claims appended hereto.
DYNAMIC EOUALIZATION FOR PARTIAL RESPONSE CHANNELS
Field of the Invention This invention relates in general to data recovery systems, and more particularly to dynamic equalization for partial response channels within a data recovery system.
Backaround of the In~ention Optimum performanoe of data recovery hardware in high efficiency data transmission channels or high - 10 density data recording channels requires that non-ideal channel behavior (e.g. linear and non-linear distortion, time base errors, etc) be identified and corrected. For example, automatic gain control (AGC) can be used to compensate for non-dispersive channel fading, and adaptive egualization is commonly used in the various channels to correct stationary equalization errors.
Brief Description of the Drawinas -The present invention will be described in greater detail below, along with a discussion of the relevant prior art, in conjunction with the following drawings in which:
Figure 1 is a block schematic diagram of a general partial response channel in accordance with the prior art;
Figure 2 is a block schematic diagram of a dicode~
partial response filter in accordance with the prior art;
Figure 3 is a block diagram of a combined dicode and partial response detector in accordance with the prior art:
Figure 4 is block diagram of a dynamic equalization system for use with combined dicode and partial response detector in accordance with the preferred embodiment of the present invention;
Figure 5 i5 a blocX diagram of a dynamic equalization system in accordance with an alternative embodiment of the present invention; and Figures 6a and 6b are photomicrographs of the experimental data showing output signals of a combined dicode and partial response detector with and without dynamic equalization.
Description of the Prior Art Compensation for non-ideal channel characteristics is particularly important when partial response detection is employed. A general partial response channel is shown in Figure 1 comprising a transmission or recording channel 1 for receiving a source data signal and a receiver filter 2 for processing the signal of the channel output. Typically, the receiver filter 2 comprises a matched filter and equalizer for minimizing noise and removing linear distortion from the signal. A
sampler 3 is provided for sampling the signal output from receiver filter 2 at a sampling rate fb = 1/T (i.e.
using the recovered data clock signal).
Since equalization of the channel 1 via receiver filter 2 may enhance additive noise within the received signal, the signal is then passed through a partial response filter shown generally as 4 which distorts the equalized signal in a known fashion in order to minimize the additive noise. The partial response filter 4 is implemented as a series of parallel connected signal multipliers 6, 8...10 having respective outputs connected to a summer 16, and being separated by respective unit delay circuits 12...14. The output of summer 16 is connected to a decoder ~not shown) for decoding of the filtered signal with the known distortion accommodated for, in a well known manner.
The partial r~esponse filter 4 may be characterized by the equation ~ ~ D' , where D
~-~
represents the delay operator for unit delay circuits 12...14.
Turning to Figure 2, a dicode partial response filter 18 is shown comprising a pair of parallel connected multipliers 20 and 22 separated by a unit delay circ~it 24. In operation, the matched filtered, equalized signal from filter 2 is multipli d by a gain of 1 in multiplier 20 and the single clock period delayed signal output from unit delay circuit 24 is multiplied by a gain of -1 (e.g. inverted) via multiplier 22 and summed with the original signal in summer 26. In other words, the delayed signal is subtracted from the original signal. Hence, the dicode filter 18 may be characterized by the discrete time equation F(D~ D, or in the frequency domain as T(w) = 2j sin(wT/2), where w = radian frequency.
Thus, whenever there is a transition in the source data signal, a pulse appears at the output of summer 26 in the dicode filter 18. The polarity of the pulse indicates the direction of the transition.
The original data signal can then be decoded by sampling the signal at the output of summer 26 with the - data clock (fb = l/T) and decoding a data transition when the signal exceeds a predetermined detection threshold. This entire operation may be referred to as threshold detection, and is represented by the threshold decoder 28 within Figure 2.
An advantage of using dicode partial response is that no dc response is required in the transmission or recording channel 1. However, in order to ensure accurate data recovery, the gain of the channel 1 must be controlled to ensure that the detection threshold is accurately set within threshold decoder 28. Typically, the channel gain is controlled to make the peaks of the positive and negative pulses approximately constant in amplitude.
A disadvantage of dicode partial response is that it requires a high signal to noise ratio (SNR) near the Nyquist freguency (i.e. one half of sampling frequency fb). In channels where the high frequency SNR is poor, higher order partial response filters are often used.
For example, if the dicode (1-D) filter 18 is followed by a duobinary (l+D) filter, a bandpass characteristic is created, in which frequencies near Nyquist as well as DC are attenuated. ~he combined (l-D) and (~+D) filters yields an overall (1-D2) characteristic known as Class IV partial response. Again, gain control is required in such a~configuration for detection of the data signals since the signal may be positive, negative or zero.
However, since the signal peaks output from the duobinary filter are much less uniform in amplitude as compared wi~h the dicode signal, depending on the source data, gain control aimed at controlling the average value of the peaks of the signal has been found to be typically of low accuracy. In general, the higher the order of the partial response filter polynomial, the more the signal characteristics depend on the source data, making signal estimation prior to detection more difficult.
The advantages of dicode filtering for automatic gain control (AGC) and Class IV partial response filtering for detection, can be combined as shown in Figure 3. The dicode signal output from filter 18 is applied to an automatic gain control circuit 30 and the output of the duobinary or other partial response filter 32 then enters the decoder for actual detection of the filtered source data signal. In this case, the partial response filter may be characterized by the following frequency domain equation:
P(w) = 2T Cos(wT/2) Thus, the frequency domain representation for the combined dicode and partial response detector of Figure 3 may be characterized as follows:
T(w) = j2T sin(wT/2)P(w) = j2T sin(wT) 5The above equations are valid for Class IV partial response filters, although it also contemplated that other forms of partial response filter may be used.
The system of Figure 3 is adequate for time-stationary channels; however, the system has been found to be inadequate when the response of the channel 1 changes dynamically. For example, in a recording channel, an increase in the effective distance between the recording head and the recording medium causes much greater attenuation at higher than lower frequencies.
As a result, loss of signal amplitude in the dicode signal will be greater than in the Class IV signal if dynamic increases in head-to-medium spacing occur.
Correction of channel gain using the dicode signal will then cause an increase in signal level as observed at the output of the partial response filter 32.
Assuming perfect AGC in the dicode channel, variations in signal amplitude measured at the output of the second partial response filter 32 can be attributed to source data or to variations in the response of the channel.
Summary of the Invention In the latter case, the variations in signal amplitude measured at the output of the partial response ~ilter can be used to correct the dynamic response changes by controlling the amount of response correction of an appropriate equalizer.
Thus, in accordance with an aspect of the present invention, there is provided in a data recovery system comprising a receiver for match filtering and equalizing a signal received from a data transmission or storage channel, and a combined dicode and partial response 2004~i03 filter for automatic gain controlling and filtering for detection of said equalized and matched filtered data signal and in response generating an output signal, the improvement comprising:
a) means for detecting amplitude/power of said output signal and in response generating an amplitude/power output signal;
b) means for lowpass filtering said amplitude/power output signal and in response generating a filtered output signal:
c) means for averaging said amplitude/power output signal and in response generating an averaged amplitude/power output signal;
d) means for subtracting said averaged amplitude/power output signal from said filtered output signal and in response generating an amplified difference signal; and e) equalization means connected to an output of said receiver filter and an input of said combined dicode and partial response filter for receiving said difference signal and in response dynamically equalizing said equalized and matched filtered data signal in accordance with dynamic channel response variations in said combined dicode and partial response filter.
In accordance with a further aspect of the present invention, there is provided in a data recovery system comprising a receiver for match filtering and equalizing an input signal, a combined dicode and partial response filter for automatic gain controlling and filtering for detection of said equalized and matched filtered data signal and in response generating an output signal, and a decoder for receiving and decoding said output signal and in response generating a decoded data signal, the improvement comprising:
a) a second combined dicode partial response filter for receiving and filtering said decoded data signal and in response generating a filtered decoded data signal;
b) first power detector means for detecting power of said output signal and in response generating a first power output signal;
c) second power detector means for detecting power of said filtered decoded data signal and in response generating a second power output signal;
d) first lowpass filter means for lowpass filtering said first power output signal and in response generating a first filtered signal;
e) second lowpass filter means for lowpass filtering said second power output signal and in response generating a second filtered signal;
f) delay circuit means for delaying said first filtered signal so as to be substantially in phase with said second filtered signal, thereby : 20 compensating for time delays within said decoder;
g) first subtraction means for subtracting said second filtered signal from said first delayed filtered signal and in response generating a first amplified difference signal;
h) third lowpass filter means for lowpass filtering said first difference signal and in response generating a third filtered signal;
i) means for averaging said first difference signal and in response generating an averaged signal;
j) second subtracting means for subtracting said averaged signal from said third filtered signal and in response generating a second amplified difference signal; and k) egualizer means connected to an output of said recei~er filter and an input of said combined dicode and partial response filter for recei~ing ;~00~503 said second difference signal and in response dynamically equalizing said equalized and makched filtered data signal an accordance with variations in the statistics of said input data signal.
Detailed Description of the Invention Turning to Figure 4, a dynamic equalizer 3~ is shown for simultaneously correcting variations in signal amplitude and dispersion within the transmission or recording channel. In particular, the power or peak amplitude of partial response filter 32 is detected within peak/power detector 36. The signal output from detector 36 is then lowpass fi~tered via filter 38 and averaged via amplitude averager 4~. The averaged signal output from circuit 40 is then subtracted from the lowpass filtered signal output from filter 38 via a differential amplifier 42, and the output signal VC is then applied to dynamic equalizer 34 as a voltage control feedback signal to correct for variations in channel response.
The equalizer shown by way of example 34 may be characterized ~y the frequency domain equation ek(Vc w) where k is a constant proportional to the recording channel loss constant known as "spacing loss".
Provided that equalizer 34 exhibits frequency characteristics appropriate to the cause of the dynamic frequency variations within the channel, the change in response can be adequately corrected by the embodiment of Figure 4. Thus, the embodiment of Figure 4 may be used to compensate for variations in spacing loss in a 0 magnetic recording channel, as discussed above.
FurthermDre, provided that the lowpass filter 38 in the feedback loop has a sufficiently narrow bandwidth, variations of signal level measured at the output of the second partial response filter (i.e.
filter 32) attributable to variations in the statistics ~:00450~
of the source data should be small, assuming the source data is approximately random.
However, if a wider filter is needed to compensate for rapid variations in frequency characteristics, the changing signal statistics attributable to changing data source statistics can present a problem.
Thus, in accordance with the altexnative embodiment of Figure 5, a compensation circuit is provided using feedback from the decoded data output from a decoder 43 which can be a threshold or other decoder.
More particularly, the decoded data signal is applied to a further combined dicode and second partial response filter 44 which is equivalent to the combined dicode and partial response filter 19. The power of the signal output from filter 44 is detected via power detector 46 similar to the peak/power detector 36 depicted in Figure 4. The power output signals from detectors 36 and 46 are filtered via respective lowpass filters 48 and 50, and the signal output from lowpass filter 48 is delayed via delay circuit 52 which compensates for delays inherent within decoder 43.
The filtered power signal output from lowpass filter 50 is subtracted from the delayed filtered power signal output from delay circuit 52 via a further differential amplifier 54, and the voltage control output signal vc' from amplifier 54 is applied to low pass filter 38 and averager 40 as discussed above in connection with the embodiment of Figure 4. Similarly, the lowpass filter 38 and averager 40 are connected to non-inverting and inverting inputs respectively of differential amplifier 42 which, in turn, is connected to a control voltage input of dynamic equalizer 34.
The power of the signal output from the second combined dicode and partial response filter 44 should be the same as the power output from partial response filter 32 under ideal conditions, except for the delay 20~)4503 through the decoder 43. Thus, in accordance with the embodiment of Figure 5, only the statistics of the source data signal will be detected and differentiated within amplifier ~4, and thereby reduced.
Finally, turning to Figures 6a and 6b, experimental data is shown from tests conducted on a rotary magnetic recorder. At the end of the video scan, there is a loss of signal which is believed to be attributable to decreasing head to tape contact as the head leaves the trailing edge of the tape. The signal in the upper trace of Figure 6a shows the effect of such signal loss after gain control has been applied using the dicode signal. Figure 6b shows the Class IV signal after dynamic equalization has been applied in accordance with the present invention. The envelope of the signal in Figure 6b is seen to be approximately flat, indicating that the equalization has effectively eliminated the error resulting from the aforementioned increased spacing loss.
Additional modifications and variations of the present invention are possible. All such modifications and embodiments are believed to be within the sphere and scope of the present invention as defined by the claims appended hereto.
Claims (3)
1. In a data recovery system comprising a receiver for match filtering and equalizing a signal received from a data transmission or storage channel, and a combined dicode and partial response filter for automatic gain controlling and filtering for detection of said equalized and matched filtered data signal and in response generating an output signal, the improvement comprising:
a) means for detecting amplitude/power of said output signal and in response generating an amplitude/power output signal;
b) means for lowpass filtering said amplitude/power output signal and in response generating a filtered output signal;
c) means for averaging said amplitude/power output signal and in response generating an averaged amplitude/power output signal;
d) means for subtracting said averaged amplitude/power output signal from said filtered output signal and in response generating an amplified difference signal: and e) equalization means connected to an output of said receiver filter and an input of said combined dicode and partial response filter for receiving said difference signal and in response dynamically equalizing said equalized and matched filtered data signal in accordance with dynamic channel response variations in said combined dicode and partial response filter.
a) means for detecting amplitude/power of said output signal and in response generating an amplitude/power output signal;
b) means for lowpass filtering said amplitude/power output signal and in response generating a filtered output signal;
c) means for averaging said amplitude/power output signal and in response generating an averaged amplitude/power output signal;
d) means for subtracting said averaged amplitude/power output signal from said filtered output signal and in response generating an amplified difference signal: and e) equalization means connected to an output of said receiver filter and an input of said combined dicode and partial response filter for receiving said difference signal and in response dynamically equalizing said equalized and matched filtered data signal in accordance with dynamic channel response variations in said combined dicode and partial response filter.
2. The improvement of claim 1 wherein said means for subtracting further comprises a differential amplifier having a non-inverting input connected to said means for lowpass filtering, an inverting input connected to said means for averaging, and an output connected to said equalization means.
3. In a data recovery system comprising a receiver for match filtering and equalizing an input signal, a combined dicode and partial response filter for automatic gain controlling and filtering for detection of said equalized and matched filtered data signal and in response generating an output signal, and a decoder for receiving and decoding said output signal and in response generating a decoded data signal, the improvement comprising:
a) a second combined dicode partial response filter for receiving and filtering said decoded data signal and in response generating a filtered decoded data signal;
b) first power detector means for detecting power of said output signal and in response generating a first power output signal;
c) second power detector means for detecting power of said filtered decoded data signal and in response generating a second power output signal;
d) first lowpass filter means for lowpass filtering said first power output signal and in response generating a first filtered signal;
e) second lowpass filter means for lowpass filtering said second power output signal and in response generating a second filtered signal;
f) delay circuit means for delaying said first filtered signal so as to be substantially in phase with said second filtered signal, thereby compensating for time delays within said decoder;
g) first subtraction means for subtracting said second filtered signal from said first delayed filtered signal and in response generating a first amplified difference signal;
h) third lowpass filter means for lowpass filtering said first difference signal and in response generating a third filtered signal;
i) means for averaging said first difference signal and in response generating an averaged signal;
j) second subtracting means for subtracting said averaged signal from said third filtered signal and in response generating a second amplified difference signal; and k) equalizer means connected to an output of said receiver filter and an input of said combined dicode and partial response filter for receiving said second difference signal and in response dynamically equalizing said equalized and matched filtered data signal an accordance with variations in the statistics of said input data signal.
5. The improvement of claim 4 wherein said first subtraction means comprises a differential amplifier having a non-inverting input connected to the said delay circuit means, an inverting input connected to said second lowpass filter means, and an output connected to said third lowpass filter means and said means for averaging.
6. The improvement of claim 4 wherein said second subtraction means comprises a differential amplifier having a non-inverting input connected to said third lowpass filter means, an inverting input connected to said means for averaging, and an output connected to said equalizer means.
a) a second combined dicode partial response filter for receiving and filtering said decoded data signal and in response generating a filtered decoded data signal;
b) first power detector means for detecting power of said output signal and in response generating a first power output signal;
c) second power detector means for detecting power of said filtered decoded data signal and in response generating a second power output signal;
d) first lowpass filter means for lowpass filtering said first power output signal and in response generating a first filtered signal;
e) second lowpass filter means for lowpass filtering said second power output signal and in response generating a second filtered signal;
f) delay circuit means for delaying said first filtered signal so as to be substantially in phase with said second filtered signal, thereby compensating for time delays within said decoder;
g) first subtraction means for subtracting said second filtered signal from said first delayed filtered signal and in response generating a first amplified difference signal;
h) third lowpass filter means for lowpass filtering said first difference signal and in response generating a third filtered signal;
i) means for averaging said first difference signal and in response generating an averaged signal;
j) second subtracting means for subtracting said averaged signal from said third filtered signal and in response generating a second amplified difference signal; and k) equalizer means connected to an output of said receiver filter and an input of said combined dicode and partial response filter for receiving said second difference signal and in response dynamically equalizing said equalized and matched filtered data signal an accordance with variations in the statistics of said input data signal.
5. The improvement of claim 4 wherein said first subtraction means comprises a differential amplifier having a non-inverting input connected to the said delay circuit means, an inverting input connected to said second lowpass filter means, and an output connected to said third lowpass filter means and said means for averaging.
6. The improvement of claim 4 wherein said second subtraction means comprises a differential amplifier having a non-inverting input connected to said third lowpass filter means, an inverting input connected to said means for averaging, and an output connected to said equalizer means.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CA 2004503 CA2004503A1 (en) | 1989-12-04 | 1989-12-04 | Dynamic equalization for partial response channels |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CA 2004503 CA2004503A1 (en) | 1989-12-04 | 1989-12-04 | Dynamic equalization for partial response channels |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA2004503A1 true CA2004503A1 (en) | 1991-06-04 |
Family
ID=4143697
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA 2004503 Abandoned CA2004503A1 (en) | 1989-12-04 | 1989-12-04 | Dynamic equalization for partial response channels |
Country Status (1)
| Country | Link |
|---|---|
| CA (1) | CA2004503A1 (en) |
-
1989
- 1989-12-04 CA CA 2004503 patent/CA2004503A1/en not_active Abandoned
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6188738B1 (en) | Clock extraction circuit | |
| KR100257962B1 (en) | Receiving circuit and method comprising adaptive equalizer having characteristics determined by signal envelope measurement | |
| US5483552A (en) | Adaptive equalizing apparatus for controlling the input signal level of quantized feedback | |
| US5265125A (en) | Signal detection apparatus for detecting digital information from PCM signal | |
| US20040190661A1 (en) | Method and system for equalizing communication signals | |
| KR100480317B1 (en) | Waveform equalizer | |
| GB2320866A (en) | An equalization arrangement in which initial values which determine tap coefficients are adaptively chosen | |
| US6173019B1 (en) | Control loop for data signal baseline correction | |
| JP3428376B2 (en) | Automatic equalization system | |
| US6363111B1 (en) | Control loop for adaptive multilevel detection of a data signal | |
| KR0152023B1 (en) | Playback signal equalization method and apparatus | |
| US6845134B2 (en) | Arrangement for receiving a digital signal from a transmission medium | |
| CA2004503A1 (en) | Dynamic equalization for partial response channels | |
| KR100213032B1 (en) | Digital signal detection device in magnetic recording and playback device | |
| JP3428339B2 (en) | Phase synchronization control circuit | |
| JP3519110B2 (en) | Signal regenerator for binary signal | |
| JPS6333234B2 (en) | ||
| KR20040019377A (en) | Apparatus and method for detecting ripples caused by multipath propagation and controlling receiving antenna and tuner | |
| US4207528A (en) | Timing-phase control system for duobinary transmission | |
| KR0144817B1 (en) | Digital playback signal detection device and method | |
| BE1007528A3 (en) | TRANSMISSION WITH IMPROVED equalizer. | |
| JP3428359B2 (en) | Waveform equalization circuit | |
| JPS593046B2 (en) | Data transmission automatic equalization method | |
| KR0148184B1 (en) | Method and apparatus of equalizing reproduced signal of recording and reproducing system for improving asymmetry improving asymmetry | |
| KR100234240B1 (en) | A reproduction equalization circuit in the magneto-optical recording / reproducing apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FZDE | Dead |