WO1995005034A1 - Data-directed scrambler for multi-bit noise-shaping d/a converters - Google Patents

Data-directed scrambler for multi-bit noise-shaping d/a converters Download PDF

Info

Publication number
WO1995005034A1
WO1995005034A1 PCT/US1994/008836 US9408836W WO9505034A1 WO 1995005034 A1 WO1995005034 A1 WO 1995005034A1 US 9408836 W US9408836 W US 9408836W WO 9505034 A1 WO9505034 A1 WO 9505034A1
Authority
WO
WIPO (PCT)
Prior art keywords
swapper
signal
output
input
cell
Prior art date
Application number
PCT/US1994/008836
Other languages
French (fr)
Inventor
Robert W. Adams
Tom W. Kwan
Original Assignee
Analog Devices, Incorporated
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Analog Devices, Incorporated filed Critical Analog Devices, Incorporated
Priority to DE69426266T priority Critical patent/DE69426266T2/en
Priority to JP7506525A priority patent/JPH09501287A/en
Priority to EP94925188A priority patent/EP0712549B1/en
Publication of WO1995005034A1 publication Critical patent/WO1995005034A1/en

Links

Classifications

    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M1/00Analogue/digital conversion; Digital/analogue conversion
    • H03M1/06Continuously compensating for, or preventing, undesired influence of physical parameters
    • H03M1/0617Continuously compensating for, or preventing, undesired influence of physical parameters characterised by the use of methods or means not specific to a particular type of detrimental influence
    • H03M1/0634Continuously compensating for, or preventing, undesired influence of physical parameters characterised by the use of methods or means not specific to a particular type of detrimental influence by averaging out the errors, e.g. using sliding scale
    • H03M1/0656Continuously compensating for, or preventing, undesired influence of physical parameters characterised by the use of methods or means not specific to a particular type of detrimental influence by averaging out the errors, e.g. using sliding scale in the time domain, e.g. using intended jitter as a dither signal
    • H03M1/066Continuously compensating for, or preventing, undesired influence of physical parameters characterised by the use of methods or means not specific to a particular type of detrimental influence by averaging out the errors, e.g. using sliding scale in the time domain, e.g. using intended jitter as a dither signal by continuously permuting the elements used, i.e. dynamic element matching
    • H03M1/0665Continuously compensating for, or preventing, undesired influence of physical parameters characterised by the use of methods or means not specific to a particular type of detrimental influence by averaging out the errors, e.g. using sliding scale in the time domain, e.g. using intended jitter as a dither signal by continuously permuting the elements used, i.e. dynamic element matching using data dependent selection of the elements, e.g. data weighted averaging
    • H03M1/0668Continuously compensating for, or preventing, undesired influence of physical parameters characterised by the use of methods or means not specific to a particular type of detrimental influence by averaging out the errors, e.g. using sliding scale in the time domain, e.g. using intended jitter as a dither signal by continuously permuting the elements used, i.e. dynamic element matching using data dependent selection of the elements, e.g. data weighted averaging the selection being based on the output of noise shaping circuits for each element
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M1/00Analogue/digital conversion; Digital/analogue conversion
    • H03M1/66Digital/analogue converters
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M3/00Conversion of analogue values to or from differential modulation
    • H03M3/30Delta-sigma modulation
    • H03M3/50Digital/analogue converters using delta-sigma modulation as an intermediate step
    • H03M3/502Details of the final digital/analogue conversion following the digital delta-sigma modulation

Definitions

  • This invention relates to D/A converters of the sigma-delta noise-shaping type. More particularly, this invention relates to multi-bit noise-shapers, i.e., where the quantizer truncates the digital word to more than 1-bit.
  • One of the problems with noise-shaping D/A con ⁇ verters is that of removing the large amount of out-of-band noise produced by the digital modulator.
  • One approach to solving this problem is to use switched-capacitor low-pass filters, as illustrated in Figure 1. Such filters are however relatively expensive in silicon area, and often exhibit non-ideal behavior such as increased noise and distortion products.
  • Another approach to solving this problem is to use a multi-bit noise-shaper, where the quantizer truncates the digital word to more than 1 bit.
  • Such a system is illustrated in Figure 2.
  • thermometer code This is a code where all of the bits are equally weighted. For example, a 5-bit binary code converted to a thermometer code would require 31 equally-weighted bits (the "all bits off” state does not require an output bit). If the input number was "11”, then the thermometer code would have the bottom 11 bits set to a "1" and the top 20 bits set to "0".
  • thermometer code In a thermometer code, it is only the number of "on" bits during a clock period that is important. This fact forms the basis for all of the scrambling algorithms described in this application.
  • scrambling works by dynamically assigning a bit at the scrambler out ⁇ put to one of the input bits, and changing this assignment on a dynamic basis (normally every clock cycle) .
  • each input bit may be connected to one and only one output bit.
  • Random scrambling would choose a random configuration of the scrambler on each clock cycle and could be accomplished in hardware by having a 2 N input multiplexer for every output line, where N is the number of binary bits at the input to the thermometer decoder. In this way, every out ⁇ put could independently select an input source, as long as the input source was not already used. However, since this technique requires 2 N switches per output line, it is not practical when N is large.
  • thermometer-decoder output is fed to a DAC consisting of 2 N nominally equally-weighted branches that sum into a summing junction, there will always be some error due to analog matching constraints that cause the weights to be slightly in error. Without any scrambling, this error would result in harmonic distortion, as every particular thermometer-decoded bit is active only during a particular range of input codes to the noise-shaper. With scrambling, there is no longer any correlation between a particular bit coming out of the scrambler and a particular range of input values to the noise-shaper. This causes the error introduced by bit-weight errors in the DAC to be de-correlated from the signal, and hence the effect of a bit-weight error is to cause increased broadband white noise instead of distortion.
  • the Carley article proposes a scrambler that is based on a fast-fourier-transfor -like "Butterfly". This algorithm is carried out by the switch arrangement shown in Figure 3, illustrated for use with an 8-bit input.
  • the switches in each column work in pairs.
  • the first column in Figure 3 contains two switches (SO and S4) that are controlled by logic signal "1A". These two switches use opposite polarities of the control signal 1A as indicated by the logic inversion bubble on one of the two switches.
  • switch "SO” and switch “S4" are both con ⁇ nected to the same two inputs (input 10 and input 14) , it will be seen that these two switches operate as a "swapper”, as illustrated in block format in Figure 4.
  • This block will either pass its two inputs directly through to the outputs, when the control signal (Norm/Swap) is low, or "swap” them when the control signal goes high so as to connect the inputs to the outputs reversely.
  • the Butterfly scrambling algorithm can be re-drawn with all groups of switches that receive the same two inputs shown as "swapper” cells, as shown in Figure 5. For this algo ⁇ rithm each swapper cell would receive a random logic signal (high or low) on each clock cycle.
  • This approach is based on a rotation-type scrambler where a barrel shifter is used as the scrambling block.
  • Such a barrel shifter preserves the number of "1's" at the output.
  • a scram ⁇ bler arrangement similar in architecture to prior art arrangements and comprising a plurality of identical switch ⁇ ing cells, referred to as swapper cells, with each having two input terminals and two output terminals.
  • An activat ⁇ ing signal controls the internal cell switches to connect the input terminals to the output terminals either directly or to connect them reversely, such that the input signals are "swapped" as they pass through the cell.
  • there are three sets of swapper cells one set of which receives the input signal as a number of bits of a thermometer code.
  • the activating signal for each cell is con ⁇ trolled to either high or low ("swap" or "no swap") by logic circuitry which responds to the states of the cell input signal bits and the state of a difference signal developed as the integrated difference of prior swapper output signals.
  • the functioning of the scrambler thus is controlled by the actual received data. This has the effect of shifting the noise due to an error in weighting to higher frequencies, out of the passband.
  • FIGURE 1 is a block diagram illustrating a one-bit noise-shaping D/A converter with a switched capacitor output filter
  • FIGURE 2 is a block diagram illustrating a multi-bit noise-shaping converter with a thermometer-code output
  • FIGURE 3 presents a diagram illustrating a Butterfly scrambling algorithm
  • FIGURE 4 shows a swapper cell in block format
  • FIGURE 5 presents diagram atically a scrambler formed of three sets of interconnected swapper cells
  • FIGURE 6 shows a swapper cell controlled by data- directed logic
  • FIGURE 7 shows the scrambler of Figure 5 rearranged and with output signal legends
  • FIGURE 8 is a diagrammatic showing of a data-directed swapper cell
  • FIGURE 9 is a circuit diagram showing an implementation of logic circuitry for a data-directed swapper cell.
  • the preferred embodiment of the present invention is based on the scrambler architecture employed to carry out the FFT-like "Butterfly" algorithm shown in Figure 5 and discussed above. That prior art architecture makes use of a number of "swapper” cells which are activated to "swap” or “no swap” condition by random logic control sig ⁇ nals (high or low) on each clock signal.
  • the activating con ⁇ trol signal is derived by performing logical operations on (1) the incoming data signals to the swapper cell, and (2) a signal developed as the integrated difference of past output signals of the swapper cell.
  • the swapper cell is identi ⁇ fied at 20, and includes the usual complementary switch pair, not shown in detail but shown diagrammatically at 22 and 24.
  • the upper input signal (IN A) is connectible by switch 22 to either the upper output terminal (OUT A) or the lower output terminal (OUT B) .
  • the other switch 24 operates simultaneously to connect the lower input signal (IN B) to (1) the lower output terminal (OUT B) or (2) the upper output terminal (OUT A) .
  • the input signals A, B are connected either directly to the corresponding output terminals A and B (upper input to upper output and lower input to lower output) , or " reversely to those output terminals, in accordance with the state (low or high) of the control signal applied at the activating signal terminal 26.
  • This logic circuitry receives the two input signals A and B and also a signal developed as the integrated difference between past swapper cell output signals.
  • the swapper cell output terminals OUT A and OUT B may in one apparatus configuration be connected to respective 1's counters functioning as integrators 32 and 34. These counters increment each time a "1" occurs at their input at the clock time.
  • the integrator outputs are directed to a differencing circuit 36 which produces and stores a signal representing the difference between the integrator outputs (COUNT A - COUNT B) .
  • This signal is applied to the logic circuitry 30 which performs a logical operation for each clock pulse at terminal 38, and produces a corresponding "swap, no-swap" signal for the control signal terminal 26 of the swapper cell.
  • the logic circuitry 30 operates on these applied signals (IN A, IN B and COUNT A - COUNT B) in the manner described in the truth table set forth below:
  • the first two columns are the input signals IN A and IN B; the third column is the differ ⁇ ence signal “COUNT A-COUNT B".
  • the fourth column gives the state of the swap control signal. It may be noted that when the inputs are the same, the state of the swap control signal is not specified; that is, it can be either high or low ("don't care") and is shown as "X”. When the inputs are not the same, the difference signal will be changed from its current state, i.e., from "0" to "1", or from "1" to "0".
  • the logic circuitry 30 causes the "new CNTA-CNTB" difference signal to always be either a "0" or a "1", so that the "CNTA-CNTB” result can be stored in only one bit.
  • the hardware realization of this circuitry therefore only requires one bit of "state”.
  • a simple combinational logic circuit that receives the current two input bits and the "CNTA-CNTB" state bit can issue the "swap/no swap" signal as well as update the state of the "CNTA-CNTB” difference signal. Accordingly, it is not necessary to actually have integrating counters such as illustrated at 32 and 34.
  • the difference signal is constrained to shift only between 0 and 1, so the integral always is bounded by being kept close to zero. Thus there will be no net d-c component developed in the output signal.
  • Each individual output bit represents the average (sum) of all of the thermometer-decoded output bits from the noise-shaper, plus a first-order noise-shaping term. This term contains two sources of shaped noise, one from the digital noise-shaper itself, and the other from the data-directed scrambling operation.
  • Various hardware arrangements can be used for carrying out the logical operations associated with the data-directed scrambler cell described above.
  • Logic ' synthesis may be employed for converting the truth table to appropriate logic elements.
  • a design resulting from such synthesis is shown in Figure 8. This design requires only two XOR gates (XI, X2) and a single clocked D-flip/flop (DFF) which holds the difference signal "state” informa ⁇ tion.
  • the XOR gates provide the logic to produce the next "state” bit, as well as the swap/no swap signal at the control signal "select" terminal 26 of the swapper cell.
  • the hardware for carrying out the logic specified by the above truth table need not include the dedicated integrating counters 32 and 34 shown in Figure 6, nor is there any need to provide a connection from the swapper output back to the logic circuitry. This is because the difference between the two swapper outputs is known by only knowing the state of the inputs and the state of the swap/no swap signal. That is, the operation carried out by the logic circuitry of Figure 8 produces the same results as the configuration shown in Figure 6, with the flip/flop DFF storing a state signal (0 or 1) which represents the integrated difference of past swapper output signals.
  • a transistor level implementation of the logic circuitry 30 is shown in Figure 9.
  • the first XOR gate function has been shifted to a position between the master and slave section of the D-flip/flop.
  • the XOR function is provided by selecting either the inverted or the non-inverted output from the master flip/flop section and applying it to the slave section.
  • the integrated difference of past output signals can take only the values of 0 and 1.
  • the difference signal can, to provide additional flexibility, assume more than two states.
  • the system may be arranged such that the difference signal can assume any of three states, such as "1", "0" and "-1".
  • INA and INB are the two swapper inputs, and "RANDOM” is the output of a random number generator that puts out “0” or “1” on every clock cycle. "X” indicates that the output of the random number generator is not used in the logic operation.
  • STATE refers to the stored value of COUNT A - COUNT B, where SO means COUNT B is ahead of COUNT A by 1, SI means COUNT A equals COUNT B, and S2 means COUNT A is ahead of COUNT B by 1.
  • SWAP refers to the swapper control signal, with “1” meaning swap and "0” meaning no swap.
  • “NEW STATE” gives the next value of the "STATE” variable described above.
  • Scrambling provides a way to gain the low out-of- band noise benefits of multi-bit designs while still retain ⁇ ing the excellent distortion characteristics of 1-bit designs.
  • the scrambling algorithm described herein allows for a relaxation of analog matching requirements by as much as 30 dB compared to a random-scrambler approach. This allows very high resolution noise-shaped DACs to be implemented in modest cost CMOS processes.

Abstract

A scrambler for use with thermometer code digital signals and having a number of interconnected identical switching cells in the form of swapper cells (20) with two inputs and two outputs. A control signal (SWAP/NO SWAP) determines whether the inputs are connected directly or reversely to the outputs. The control signal is developed by logic circuitry (30) which receives as inputs the two swapper cell input bits and a state bit representing the integrated difference of past swapper output signals.

Description

DATA-DIRECTED SCRAMBLER FOR MULTI-BIT NOISE-SHAPING D/A CONVERTERS
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to D/A converters of the sigma-delta noise-shaping type. More particularly, this invention relates to multi-bit noise-shapers, i.e., where the quantizer truncates the digital word to more than 1-bit.
2. Description of the Prior Art
One of the problems with noise-shaping D/A con¬ verters is that of removing the large amount of out-of-band noise produced by the digital modulator. One approach to solving this problem is to use switched-capacitor low-pass filters, as illustrated in Figure 1. Such filters are however relatively expensive in silicon area, and often exhibit non-ideal behavior such as increased noise and distortion products. Another approach to solving this problem is to use a multi-bit noise-shaper, where the quantizer truncates the digital word to more than 1 bit. Such a system is illustrated in Figure 2. By using multiple levels and arranging the loop such that the idle-channel limit-cycle noise only spans a few quantization levels, it is possible to reduce the out-of-band noise significantly compared to the 1-bit case. For example, a loop employing a 5-bit quantizer should allow for reduction of out-of-band noise by roughly a factor of 30:1.
The reason this latter approach has not commonly been used is that errors in the digital-to-analog converter used to turn the multi-bit digital signal into an analog voltage will cause linearity errors in the analog output. This is quite different from the single-bit case, where perfect linearity is guaranteed by virtue of the fact that only two levels exist at the analog output.
Proposals have been made of ways to use "scram¬ bling" techniques to minimize the linearity problem with multi-bit noise-shapers. Examples include an article by L. Richard Carley titled "A Noise-Shaping Coder Topology'for 15+ Bit Converters", IEEE J. Solid-state Circuits, Vol. SC-24, pp. 267-273, April 1989, and an article by Bosco H. Leung et al titled "Multibit Sigma-Delta A/D Converter Incorporating A Novel Class of Dynamic Element Matching Techniques", IEEE Transactions on Circuits and Systems - II: Analog and Digital Signal Processing, Vol. 39, No. 1, January 1992.
All of such "scrambling" arrangements require that the multi-bit output of the noise shaper first be converted to a "thermometer code". This is a code where all of the bits are equally weighted. For example, a 5-bit binary code converted to a thermometer code would require 31 equally-weighted bits (the "all bits off" state does not require an output bit). If the input number was "11", then the thermometer code would have the bottom 11 bits set to a "1" and the top 20 bits set to "0".
In a thermometer code, it is only the number of "on" bits during a clock period that is important. This fact forms the basis for all of the scrambling algorithms described in this application. Conceptually, scrambling works by dynamically assigning a bit at the scrambler out¬ put to one of the input bits, and changing this assignment on a dynamic basis (normally every clock cycle) .
It will be evident that these input bit-to-output bit assignments must be mutually exclusive. That is, each input bit may be connected to one and only one output bit. Random scrambling would choose a random configuration of the scrambler on each clock cycle and could be accomplished in hardware by having a 2N input multiplexer for every output line, where N is the number of binary bits at the input to the thermometer decoder. In this way, every out¬ put could independently select an input source, as long as the input source was not already used. However, since this technique requires 2N switches per output line, it is not practical when N is large.
If a thermometer-decoder output is fed to a DAC consisting of 2N nominally equally-weighted branches that sum into a summing junction, there will always be some error due to analog matching constraints that cause the weights to be slightly in error. Without any scrambling, this error would result in harmonic distortion, as every particular thermometer-decoded bit is active only during a particular range of input codes to the noise-shaper. With scrambling, there is no longer any correlation between a particular bit coming out of the scrambler and a particular range of input values to the noise-shaper. This causes the error introduced by bit-weight errors in the DAC to be de-correlated from the signal, and hence the effect of a bit-weight error is to cause increased broadband white noise instead of distortion.
The Carley article, referred to above, proposes a scrambler that is based on a fast-fourier-transfor -like "Butterfly". This algorithm is carried out by the switch arrangement shown in Figure 3, illustrated for use with an 8-bit input. The switches in each column work in pairs. For example, the first column in Figure 3 contains two switches (SO and S4) that are controlled by logic signal "1A". These two switches use opposite polarities of the control signal 1A as indicated by the logic inversion bubble on one of the two switches.
This Butterfly topology does not allow all possible configurations of input to output mappings, but only a limited subset. However, every input line can connect to every output line, and the number of switches need only be K-N (where K = 2N and N = number of binary bits at the input to the thermometer decoder) instead of K2 as before.
Since switch "SO" and switch "S4" are both con¬ nected to the same two inputs (input 10 and input 14) , it will be seen that these two switches operate as a "swapper", as illustrated in block format in Figure 4. This block will either pass its two inputs directly through to the outputs, when the control signal (Norm/Swap) is low, or "swap" them when the control signal goes high so as to connect the inputs to the outputs reversely. Accordingly, the Butterfly scrambling algorithm can be re-drawn with all groups of switches that receive the same two inputs shown as "swapper" cells, as shown in Figure 5. For this algo¬ rithm each swapper cell would receive a random logic signal (high or low) on each clock cycle. It may be noted that in Figure 5, the order of the switches in each column has been altered to provide that the interconnection between stages is the same for each stage. Such modification of the flow graph also has been made for FFT structures, as described by A. V. Oppenheim et al, in "Digital Signal Processing", Prentice-Hall, 1975, pg. 309.
The increase in white noise floor power resulting from random scrambling may be acceptable for low-perform¬ ance designs, but to achieve very high performance (for example, greater than lOOdB) still requires very high accuracy in the DAC thermometer weights. To overcome this problem, a solution has been proposed by Leung et al (see above) called "individual level averaging", which results in the output spectrum of each bit at the scrambler output being noise-shaped. This noise-shaping causes the error produced by a non-ideal DAC weight to occur only at high frequencies that are above the band of interest. This allows larger errors to occur in the DAC without increasing the in-band noise penalty.
This approach is based on a rotation-type scrambler where a barrel shifter is used as the scrambling block. For an N-bit input with K = 2N thermometer-decode inputs to the scrambler, there are K unique rotations possible. Such a barrel shifter preserves the number of "1's" at the output. For each of the possible K = 2 thermometer- decoded scrambler input patterns (where N is the number of bits in the loop quantizer) , a memory stores the state of the barrel shifter that was used to map input bits into output bits. Whenever a particular input code is presented to the barrel shifter, the "memory" for that particular input code is examined, and a barrel-shift control word is selected which has not been used yet in the sequence. For example, in the 3-bit case, there are 8 possible input codes and 8 possible barrel-shift positions. For each input code there is a memory that is 7 locations deep. All 7 locations are examined for a particular input code, and whichever barrel-shift code is not found in the memory is used for the current shift control word. This word is then entered in the memory for that particular input code.
This technique guarantees that over a long time period there is no DC error for every possible input code, as all possible barrel-shift locations are used for that input code in the shortest possible time. However, the amount of time taken for this sequence to complete for a given code is data-dependent because it depends on the frequency of occurrence of that particular input code. It is quite possible to get very long repeat times, especially for noise shapers with a high number of quantization levels. This tends to produce lower-frequency noise, and reduces the benefit of this technique in comparison to random averaging using the Butterfly scrambler described above. A serious drawback is that the spectrum of each individual bit output may contain low-frequency noise that degrades the passband signal-to-noise ratio. SUMMARY OF THE INVENTION
In a preferred embodiment of the invention, to be described hereinafter in detail, there is provided a scram¬ bler arrangement similar in architecture to prior art arrangements and comprising a plurality of identical switch¬ ing cells, referred to as swapper cells, with each having two input terminals and two output terminals. An activat¬ ing signal controls the internal cell switches to connect the input terminals to the output terminals either directly or to connect them reversely, such that the input signals are "swapped" as they pass through the cell. In the described embodiment, there are three sets of swapper cells, one set of which receives the input signal as a number of bits of a thermometer code. The output bits of that set are connected as inputs to the second set, and the outputs of the latter set are connected as inputs to the third set which produces the final scrambled output. In such a circuit configuration, more than 3 sets will be required for N>3 (K>8) . For example, 5 sets are required for K=32 (N=5) .
In accordance with an important aspect of the invention, the activating signal for each cell is con¬ trolled to either high or low ("swap" or "no swap") by logic circuitry which responds to the states of the cell input signal bits and the state of a difference signal developed as the integrated difference of prior swapper output signals. The functioning of the scrambler thus is controlled by the actual received data. This has the effect of shifting the noise due to an error in weighting to higher frequencies, out of the passband. Other objects, advantages and aspects of the invention will in part be pointed out in, and in part apparent from, the following detailed description considered together with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGURE 1 is a block diagram illustrating a one-bit noise-shaping D/A converter with a switched capacitor output filter;
FIGURE 2 is a block diagram illustrating a multi-bit noise-shaping converter with a thermometer-code output;
FIGURE 3 presents a diagram illustrating a Butterfly scrambling algorithm;
FIGURE 4 shows a swapper cell in block format;
FIGURE 5 presents diagram atically a scrambler formed of three sets of interconnected swapper cells;
FIGURE 6 shows a swapper cell controlled by data- directed logic;
FIGURE 7 shows the scrambler of Figure 5 rearranged and with output signal legends;
FIGURE 8 is a diagrammatic showing of a data-directed swapper cell; and
FIGURE 9 is a circuit diagram showing an implementation of logic circuitry for a data-directed swapper cell. DETAILED DESCRIPTION OF PREFERRED EMBODIMENT
The preferred embodiment of the present invention is based on the scrambler architecture employed to carry out the FFT-like "Butterfly" algorithm shown in Figure 5 and discussed above. That prior art architecture makes use of a number of "swapper" cells which are activated to "swap" or "no swap" condition by random logic control sig¬ nals (high or low) on each clock signal. In the inventive embodiment described herein, however, the activating con¬ trol signal is derived by performing logical operations on (1) the incoming data signals to the swapper cell, and (2) a signal developed as the integrated difference of past output signals of the swapper cell.
This new swapper cell arrangement is illustrated in block format in Figure 6. The swapper cell is identi¬ fied at 20, and includes the usual complementary switch pair, not shown in detail but shown diagrammatically at 22 and 24. In such switch configuration, the upper input signal (IN A) is connectible by switch 22 to either the upper output terminal (OUT A) or the lower output terminal (OUT B) . The other switch 24 operates simultaneously to connect the lower input signal (IN B) to (1) the lower output terminal (OUT B) or (2) the upper output terminal (OUT A) . Thus, the input signals A, B are connected either directly to the corresponding output terminals A and B (upper input to upper output and lower input to lower output) , or" reversely to those output terminals, in accordance with the state (low or high) of the control signal applied at the activating signal terminal 26.
The state of this activating control signal is determined by the operation of logic circuitry shown in block format at 30. This logic circuitry receives the two input signals A and B and also a signal developed as the integrated difference between past swapper cell output signals. To this end, the swapper cell output terminals OUT A and OUT B may in one apparatus configuration be connected to respective 1's counters functioning as integrators 32 and 34. These counters increment each time a "1" occurs at their input at the clock time.
The integrator outputs are directed to a differencing circuit 36 which produces and stores a signal representing the difference between the integrator outputs (COUNT A - COUNT B) . This signal is applied to the logic circuitry 30 which performs a logical operation for each clock pulse at terminal 38, and produces a corresponding "swap, no-swap" signal for the control signal terminal 26 of the swapper cell.
The logic circuitry 30 operates on these applied signals (IN A, IN B and COUNT A - COUNT B) in the manner described in the truth table set forth below:
2-State Truth Table INA INB CNTA-CNTB SWAP CTRL NEW CNTA-CNTB
0 0 0 0 0 1 0 1 0 0 1 1 1 0 0 1 0 1 1 1 0 1 1 1
Figure imgf000012_0001
In the above table, the first two columns are the input signals IN A and IN B; the third column is the differ¬ ence signal "COUNT A-COUNT B". The fourth column gives the state of the swap control signal. It may be noted that when the inputs are the same, the state of the swap control signal is not specified; that is, it can be either high or low ("don't care") and is shown as "X". When the inputs are not the same, the difference signal will be changed from its current state, i.e., from "0" to "1", or from "1" to "0".
The logic circuitry 30 causes the "new CNTA-CNTB" difference signal to always be either a "0" or a "1", so that the "CNTA-CNTB" result can be stored in only one bit. The hardware realization of this circuitry therefore only requires one bit of "state". A simple combinational logic circuit that receives the current two input bits and the "CNTA-CNTB" state bit can issue the "swap/no swap" signal as well as update the state of the "CNTA-CNTB" difference signal. Accordingly, it is not necessary to actually have integrating counters such as illustrated at 32 and 34. The difference signal is constrained to shift only between 0 and 1, so the integral always is bounded by being kept close to zero. Thus there will be no net d-c component developed in the output signal.
It can be shown that this scrambling technique results in "noise-shaping" of each individual swapper out¬ put bit. Thus, if a number of such data-directed swapper cells are arranged in interconnected format, for example for an 8-bit thermometer code input as shown in Figure 7, the outputs from each swapper cell will, as indicated by the legends on the drawing, represent an average of the input signals received by that swapper cell, plus a noise-shaping error term (identified as N.S.).
In a multi-bit sigma-delta loop, the sum of all the thermometer-decoded output bits will represent the to¬ tal input signal plus a noise-shaping term. Therefore, the following can be said of the scrambler shown in Figure 7: (1) The sum of all of the scrambled output bits (assuming no errors) is equal to the noise-shaping input plus an Nth-order noise-shaping term, where N is the order of the digital shaper. (N=2 in the noise shaper of Figure 2.)
(2) Each individual output bit represents the average (sum) of all of the thermometer-decoded output bits from the noise-shaper, plus a first-order noise-shaping term. This term contains two sources of shaped noise, one from the digital noise-shaper itself, and the other from the data-directed scrambling operation.
An advantage of using this approach is that if a particular output weight is incorrect, most of the error will occur at higher frequencies, and very little of the error will occur in the band of interest. Since each bit represents the input plus two noise-shaping terms, an error in the output weight will cause no harmonic distortion.
Various hardware arrangements can be used for carrying out the logical operations associated with the data-directed scrambler cell described above. Logic ' synthesis may be employed for converting the truth table to appropriate logic elements. A design resulting from such synthesis is shown in Figure 8. This design requires only two XOR gates (XI, X2) and a single clocked D-flip/flop (DFF) which holds the difference signal "state" informa¬ tion. The XOR gates provide the logic to produce the next "state" bit, as well as the swap/no swap signal at the control signal "select" terminal 26 of the swapper cell.
It will be seen that the hardware for carrying out the logic specified by the above truth table need not include the dedicated integrating counters 32 and 34 shown in Figure 6, nor is there any need to provide a connection from the swapper output back to the logic circuitry. This is because the difference between the two swapper outputs is known by only knowing the state of the inputs and the state of the swap/no swap signal. That is, the operation carried out by the logic circuitry of Figure 8 produces the same results as the configuration shown in Figure 6, with the flip/flop DFF storing a state signal (0 or 1) which represents the integrated difference of past swapper output signals.
A transistor level implementation of the logic circuitry 30 is shown in Figure 9. In this circuitry, the first XOR gate function has been shifted to a position between the master and slave section of the D-flip/flop. The XOR function is provided by selecting either the inverted or the non-inverted output from the master flip/flop section and applying it to the slave section.
In the embodiment described above, the integrated difference of past output signals can take only the values of 0 and 1. This is not a necessary limitation, and the difference signal can, to provide additional flexibility, assume more than two states. For example, the system may be arranged such that the difference signal can assume any of three states, such as "1", "0" and "-1". In practical terms, and referring to Figure 6, this would mean that the count in the upper integrator 32 can be (1) one count greater than that of the lower integrator 34, (2) equal to that of the lower integrator, or (3) one count behind the lower integrator. It is often desirable to insure that no "patterns" occur in the data at the scrambler output, which might cause discrete frequencies at each scrambler output bit. By allowing 3 possible values of integrated differ¬ ence (-1, 0, +1) , it is possible under certain conditions to randomly select between swap and no-swap. For example, when the difference is 0 and the inputs are 01 or 10, a random selection may be made. This random selection helps to prevent pattern noise. A logical truth table showing how such an arrangement would operate is set forth below:
3-State Truth Table INA INB RA STATE
Figure imgf000016_0001
Figure imgf000016_0002
In this truth table, INA and INB are the two swapper inputs, and "RANDOM" is the output of a random number generator that puts out "0" or "1" on every clock cycle. "X" indicates that the output of the random number generator is not used in the logic operation. "STATE" refers to the stored value of COUNT A - COUNT B, where SO means COUNT B is ahead of COUNT A by 1, SI means COUNT A equals COUNT B, and S2 means COUNT A is ahead of COUNT B by 1. "SWAP" refers to the swapper control signal, with "1" meaning swap and "0" meaning no swap. "NEW STATE" gives the next value of the "STATE" variable described above. The only time "RANDOM" is used is when the state is "SI" (COUNT A = COUNT B) and the inputs are different. Under those conditions, the choice of "swap/no swap" may be random, i.e., either integrator may be pushed ahead by one count. As in the case of the 2-state truth table described previously, the hardware for carrying out the 3-state algorithm need not include physical integrating counters. The development of hardware suitable for this algorithm can be derived from the 3-state truth table by means of logic synthesizing computer programs known in the art.
Scrambling provides a way to gain the low out-of- band noise benefits of multi-bit designs while still retain¬ ing the excellent distortion characteristics of 1-bit designs. The scrambling algorithm described herein allows for a relaxation of analog matching requirements by as much as 30 dB compared to a random-scrambler approach. This allows very high resolution noise-shaped DACs to be implemented in modest cost CMOS processes.
Although a preferred embodiment of the invention has been disclosed herein in detail, it is to be understood that this is for the purpose of illustrating the invention, and should not be construed as necessarily limiting the scope of the invention since it is apparent that many changes can be made by those skilled in the art while still practicing the invention claimed herein.

Claims

What is Claimed is:
1. For use in apparatus for converting digital signals to analog signals, wherein a multi-bit digital signal is transformed to a thermometer code which passes through a scrambler to a D/A converter to produce an analog output signal; an improved scrambler comprising: input means and output means; a plurality of switchable scrambler devices configured as swapper cells interconnected to provide a number of mutually-exclusive selectable signal paths from said input means to said output means; each of said swapper cells including: two input terminals to receive respective signal bits and two output terminals to produce corresponding output bits; means within each swapper cell for controllably connecting the two input terminals to the two output terminals either directly or reversely; select terminal means for receiving an activating signal designating whether the swapper cell is to effect said direct connection or said reverse connection; logic means for producing said activating signal; said logic means including means for storing the integrated difference of past swapper output signals; said logic means further including means for computing the new value of the activating signal based on the current two input values to said swapper cell and said stored integrated difference signal and for updating the value of the stored difference signal.
2. Apparatus as claimed in Claim 1, wherein said means for updating the value of the stored difference signal comprises means responsive to the present two input signals and the new value of the activating signal.
3. Apparatus as claimed in Claim 1, wherein said swapper cells are arranged in a plurality of equal-numbered sets; the input terminals of first set receiving the individual bits of said thermometer code; and the input terminals of each succeeding set receiving the output signals of the preceding set.
4. Apparatus as claimed in Claim 1, wherein said logic means comprises a clocked flip-flop; a first XOR device having one input driven by the output of said flip-flop; a second XOR device having one input driven by the output of said first XOR device; means connecting the output of said second XOR device to the input of said flip-flop; and means connecting the two swapper cell input signals to the other inputs of said first and second XOR devices respectively; the output of said first XOR device being connected to said swapper cell as said activating signal.
5. In a digital-to-analog conversion system wherein a digital thermometer code signal is passed through scrambler apparatus of the type comprising means definable as a plurality of swapper cells each having two inputs and two outputs and including means operable by a control signal for establishing either a direct connection between each input and a corresponding output or a reverse connection between said inputs and outputs, said plurality of swapper cells being interconnected to provide a number of mutually-exclusive signal paths selectable in accordance with the states of the control signals for the individual swapper cells; the method of developing the control signal for each swapper cell which comprises: storing a signal representing the integrated difference of past output signals of the swapper cell; and operating logically on the state of the two input signals for the swapper cell and said integrated difference to produce said control signal for the swapper cell.
6. The method of Claim 5, wherein the logical operation includes a comparison of the two input signals to determine whether a change should be made to the integrated difference signal.
7. The method of Claim 6, wherein the logical operation changes said integrated difference signal if the two inputs are different.
8. The method of Claim 5, wherein said integrated difference of past swapper cell outputs establishes whether the difference between the two outputs is one or the other of two possible states.
9. The method of Claim 5, wherein said integrated difference of past swapper cell outputs establishes whether the difference between the two cell outputs is one of three possible states.
10. In a digital-to-analog conversion system wherein a digital thermometer code signal is passed through scram¬ bler apparatus of the type comprising means definable as a plurality of swapper cells each having two inputs and two outputs and including means operable by a control signal for establishing either a direct connection between each input and a corresponding output or a reverse connection between said inputs and outputs, said plurality of swapper cells being interconnected to provide a number of mutually- exclusive signal paths selectable in accordance with the states of the control signals for the individual swapper cells; the method of developing the control signal for each swapper cell which comprises: operating logically on the states of the two swapper input signals for the respective swapper cell; including in said logical operation the states of the output signals of the swapper; and producing said control signal from the results of said logical operation.
11. The method of Claim 10, wherein said logical operation carries out the following truth table:
Figure imgf000021_0001
In this truth table, the first two columns are the swapper cell inputs; the third column is the integrated difference of past swapper outputs; the fourth column is the state of the swap control signal; the fifth column is the new state of the difference signal.
12. The method of Claim 10, wherein said logical operation carries out the following truth table:
INA INB RANDOM STATE SWAP NEW STATE
(swap=l)
Figure imgf000021_0002
Figure imgf000021_0003
INA and INB are the two swapper inputs, and "RANDOM" is the output of a random number generator that puts out "0" or "1" on every clock cycle. "X" indicates that the output of the random number generator is not used in the logic operation. "STATE" refers to the stored value of COUNT A - COUNT B, where SO means COUNT B is ahead of COUNT A by 1, SI means COUNT A equals COUNT B, and S2 means COUNT A is ahead of COUNT B by 1. "SWAP" refers to the swapper control signal, with "1" meaning swap and "0" meaning no swap. "NEW STATE" gives the next value of the "STATE" variable described above.
PCT/US1994/008836 1993-08-05 1994-08-05 Data-directed scrambler for multi-bit noise-shaping d/a converters WO1995005034A1 (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
DE69426266T DE69426266T2 (en) 1993-08-05 1994-08-05 DATA-CONTROLLED ENCODER FOR DA-CONVERTERS WITH MULTIBIT NOISE SIGNALING
JP7506525A JPH09501287A (en) 1993-08-05 1994-08-05 Data directed scrambler for multi-bit noise to shaping D / A converter
EP94925188A EP0712549B1 (en) 1993-08-05 1994-08-05 Data-directed scrambler for multi-bit noise-shaping d/a converters

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US08/102,574 US5404142A (en) 1993-08-05 1993-08-05 Data-directed scrambler for multi-bit noise shaping D/A converters
US08/102,574 1993-08-05

Publications (1)

Publication Number Publication Date
WO1995005034A1 true WO1995005034A1 (en) 1995-02-16

Family

ID=22290538

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1994/008836 WO1995005034A1 (en) 1993-08-05 1994-08-05 Data-directed scrambler for multi-bit noise-shaping d/a converters

Country Status (5)

Country Link
US (1) US5404142A (en)
EP (1) EP0712549B1 (en)
JP (1) JPH09501287A (en)
DE (1) DE69426266T2 (en)
WO (1) WO1995005034A1 (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2350950A (en) * 1999-06-07 2000-12-13 Nippon Precision Circuits Delta sigma D/A converter
US6797491B2 (en) 2000-06-26 2004-09-28 Stressgen Biotechnologies Corporation Human papilloma virus treatment
US8423165B2 (en) 2006-05-21 2013-04-16 Trigence Semiconductor, Inc. Digital/analogue conversion apparatus
US9219960B2 (en) 2009-12-16 2015-12-22 Trigence Semiconductor Inc. Acoustic playback system
US9226053B2 (en) 2008-06-16 2015-12-29 Trigence Semiconductor, Inc. Digital speaker driving apparatus
US9300310B2 (en) 2009-12-09 2016-03-29 Trigence Semiconductor, Inc. Selection device

Families Citing this family (84)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5887145A (en) * 1993-09-01 1999-03-23 Sandisk Corporation Removable mother/daughter peripheral card
US7137011B1 (en) * 1993-09-01 2006-11-14 Sandisk Corporation Removable mother/daughter peripheral card
JP3469326B2 (en) * 1994-08-16 2003-11-25 バー−ブラウン・コーポレーション Digital to analog converter
US5684482A (en) * 1996-03-06 1997-11-04 Ian A. Galton Spectral shaping of circuit errors in digital-to-analog converters
WO1998048515A1 (en) * 1997-04-18 1998-10-29 Steensgaard Madsen Jesper Oversampled digital-to-analog converter based on nonlinear separation and linear recombination
US5936562A (en) * 1997-06-06 1999-08-10 Analog Devices, Inc. High-speed sigma-delta ADCs
US5982313A (en) * 1997-06-06 1999-11-09 Analog Devices, Inc. High speed sigma-delta analog-to-digital converter system
US6124813A (en) * 1997-06-06 2000-09-26 Analog Devices, Inc. Self-linearizing multi-bit DACs
US5977899A (en) * 1997-09-25 1999-11-02 Analog Devices, Inc. Digital-to-analog converter using noise-shaped segmentation
EP1093687B1 (en) * 1998-07-07 2002-03-13 Infineon Technologies AG Linearized multi-bit digital/analog converter and the use thereof in a multi-bit delta-sigma analog/digital converter
US6300892B2 (en) 1998-07-07 2001-10-09 Infineon Technologies Ag Linearized multibit digital/analog converter and its use in a multibit delta-sigma analog/digital converter
WO2000008765A2 (en) 1998-08-06 2000-02-17 Steensgaard Madsen Jesper Delta-sigma a/d converter
US6218977B1 (en) 1998-09-25 2001-04-17 Conexant Systems, Inc. Methods and apparatus for distributing mismatched error associated with data converter elements
US6348884B1 (en) 1999-01-06 2002-02-19 Jesper Steensgaard-Madsen Idle-tone-free mismatch-shaping encoders
US6154162A (en) * 1999-01-06 2000-11-28 Centillium Communications, Inc. Dual-stage switched-capacitor DAC with scrambled MSB's
JP3232457B2 (en) * 1999-02-05 2001-11-26 日本プレシジョン・サーキッツ株式会社 Delta-sigma D / A converter
US6816100B1 (en) 1999-03-12 2004-11-09 The Regents Of The University Of California Analog-to-digital converters with common-mode rejection dynamic element matching, including as used in delta-sigma modulators
US6420991B1 (en) 1999-09-08 2002-07-16 Texas Instruments Incorporated Dynamic element matching for converting element mismatch into white noise for a pipelined analog to digital converter
US6211805B1 (en) 1999-09-08 2001-04-03 Texas Instruments Incorporated Noise shaping dynamic element mismatch in analog to digital converters
US6411232B1 (en) 1999-09-30 2002-06-25 Motorola, Inc. Method and system for determining an element conversion characteristic contemporaneous with converting and input signal in a signal converter
US6466147B1 (en) 1999-10-25 2002-10-15 Hrl Laboratories, Llc Method and apparatus for randomized dynamic element matching DAC
DE60030599T2 (en) * 1999-11-10 2006-12-21 Fujitsu Ltd., Kawasaki Noise shaping in segmented mixed signal circuits
US6433711B1 (en) * 1999-12-14 2002-08-13 Texas Instruments Incorporated System and method for offset error compensation in comparators
US6373424B1 (en) 1999-12-21 2002-04-16 Texas Instruments Incorporated Method and apparatus for obtaining linearity in a pipelined analog-to-digital converter
US6466153B1 (en) 1999-12-23 2002-10-15 Texas Instruments Incorporated Highspeed, high spurious-free dynamic range pipelined analog to digital converter
US6545623B1 (en) 1999-12-23 2003-04-08 Texas Instruments Incorporated High speed analog-domain shuffler for analog to digital converter
JP2001251190A (en) * 2000-03-08 2001-09-14 Nippon Precision Circuits Inc Delta/sigma d/a converter
JP3725001B2 (en) 2000-03-28 2005-12-07 株式会社東芝 Selection circuit, D / A converter and A / D converter
US6917321B1 (en) 2000-05-21 2005-07-12 Analog Devices, Inc. Method and apparatus for use in switched capacitor systems
US7199740B1 (en) 2000-05-21 2007-04-03 Analog Devices, Inc. Method and apparatus for use in switched capacitor systems
SE522416C2 (en) * 2000-05-23 2004-02-10 Ericsson Telefon Ab L M Procedure for scrambling data words and scrambler
DE60015958T2 (en) * 2000-08-10 2005-12-01 Stmicroelectronics S.R.L., Agrate Brianza Digital-to-analog converter circuit
US6441759B1 (en) 2000-08-30 2002-08-27 Hrl Laboratories, Llc Multi-bit ΔΣ modulator having linear output
US6919832B2 (en) 2000-09-11 2005-07-19 Broadcom Corporation Methods and systems for high speed quantizers
WO2002023728A2 (en) * 2000-09-11 2002-03-21 Broadcom Corporation Method and apparatus for mismatched shaping of an oversampled converter
WO2002023731A2 (en) 2000-09-11 2002-03-21 Broadcom Corporation Methods and systems for digital dither
WO2002023732A2 (en) 2000-09-11 2002-03-21 Broadcom Corporation Methods and systems for high speed quantizers
WO2002023733A2 (en) 2000-09-11 2002-03-21 Broadcom Corporation Sigma-delta digital-to-analog converter
IT1320694B1 (en) * 2000-10-06 2003-12-10 St Microelectronics Srl DYNAMIC EQUALIZATION METHOD OF THE ELEMENTS OF A DIGITAL / ANALOG MULTIBIT CONVERTER INTEGRATED WITH BALANCED OUTPUT FOR
US7068788B2 (en) * 2001-01-04 2006-06-27 Maxim Integrated Products, Inc. Data encryption for suppression of data-related in-band harmonics in digital to analog converters
US6522277B2 (en) * 2001-02-05 2003-02-18 Asahi Kasei Microsystems, Inc. Circuit, system and method for performing dynamic element matching using bi-directional rotation within a data converter
SE522569C2 (en) * 2001-02-27 2004-02-17 Ericsson Telefon Ab L M Dynamic element metering in a / d converters
US6975682B2 (en) * 2001-06-12 2005-12-13 Raytheon Company Multi-bit delta-sigma analog-to-digital converter with error shaping
US6518899B2 (en) * 2001-06-13 2003-02-11 Texas Instruments Incorporated Method and apparatus for spectral shaping of non-linearity in data converters
US6697004B1 (en) 2001-10-01 2004-02-24 Silicon Wave, Inc. Partial mismatch-shaping digital-to-analog converter
US6570521B1 (en) 2001-12-27 2003-05-27 Analog Devices, Inc. Multistage scrambler for a digital to analog converter
US6762702B2 (en) * 2002-01-24 2004-07-13 Broadcom Corporation Shuffler apparatus and related dynamic element matching technique for linearization of unit-element digital-to-analog converters
US6795003B2 (en) * 2003-01-30 2004-09-21 Broadcom Corporation Hardware-efficient implementation of dynamic element matching in sigma-delta DAC's
US6614377B1 (en) 2002-02-08 2003-09-02 Analog Devices, Inc. Data-directed scrambler for noise-shaping mixed-signal converters with an arbitrary number of quantization levels
US6617990B1 (en) * 2002-03-06 2003-09-09 Meshnetworks Digital-to-analog converter using pseudo-random sequences and a method for using the same
JP4237448B2 (en) * 2002-05-22 2009-03-11 株式会社ルネサステクノロジ Manufacturing method of semiconductor device
KR20030086896A (en) * 2002-05-03 2003-11-12 톰슨 라이센싱 소시에떼 아노님 Thermometer code digital to audio converter
EP1359671A1 (en) * 2002-05-03 2003-11-05 Thomson Licensing S.A. Thermometer code digital to analog converter for audio signals
US6819276B1 (en) 2003-05-13 2004-11-16 Analog Devices, Inc. Noise-shaper system and method
US6867721B1 (en) 2003-09-23 2005-03-15 Realtek Semiconductor Corp Spectral shaping dynamic encoder for a DAC
US7606321B2 (en) * 2004-01-22 2009-10-20 Broadcom Corporation System and method for simplifying analog processing in a transmitter incorporating a randomization circuit
US7193548B2 (en) * 2004-01-30 2007-03-20 Hrl Laboratories, Llc Switching arrangement and DAC mismatch shaper using the same
US7095345B2 (en) * 2004-06-29 2006-08-22 Analog Devices, Inc. Hybrid tuning circuit for continuous-time sigma-delta analog-to-digital converter
WO2006034177A1 (en) * 2004-09-17 2006-03-30 Analog Devices, Inc. Multi-bit continuous-time front-end sigma-delta adc using chopper stabilization
US7079063B1 (en) * 2005-04-18 2006-07-18 Analog Devices, Inc. System and method for tri-level logic data shuffling for oversampling data conversion
DE102005035225B4 (en) * 2005-07-25 2007-06-14 Micronas Gmbh Linearization Circuitry and Digital Element Adaptation Linearization Method for Digital to Analog Converter
JP2007060160A (en) * 2005-08-23 2007-03-08 Fujitsu Ltd Semiconductor integrated circuit
US20070126616A1 (en) * 2005-12-07 2007-06-07 Min Hyung Cho Dynamically linearized digital-to-analog converter
US7425910B1 (en) * 2006-02-27 2008-09-16 Marvell International Ltd. Transmitter digital-to-analog converter with noise shaping
US8502557B2 (en) 2006-06-05 2013-08-06 Analog Devices, Inc. Apparatus and methods for forming electrical networks that approximate desired performance characteristics
US7492297B2 (en) * 2006-08-11 2009-02-17 Realtek Semiconductor Corp. Digital-to-analog converter and method thereof
US7446687B2 (en) * 2006-10-27 2008-11-04 Realtek Semiconductor Corp. Method and apparatus to reduce internal circuit errors in a multi-bit delta-sigma modulator
US7420494B1 (en) 2007-04-30 2008-09-02 Analog Devices, Inc. Mismatch shaping Δ-Σ analog to digital converter system
US7446688B1 (en) * 2007-05-09 2008-11-04 Windond Electronics Corporation Sequence generation for mismatch-shaping circuits
US9356568B2 (en) 2007-06-05 2016-05-31 Analog Devices, Inc. Apparatus and methods for chopper amplifiers
US7545295B2 (en) * 2007-09-14 2009-06-09 Realtek Semiconductor Corp. Self-calibrating digital-to-analog converter and method thereof
US7679539B2 (en) * 2008-03-25 2010-03-16 Megawin Technology Co., Ltd. Randomized thermometer-coding digital-to-analog converter and method therefor
US7663523B1 (en) 2008-04-04 2010-02-16 On Semiconductor System unit element selection
US7777658B2 (en) * 2008-12-12 2010-08-17 Analog Devices, Inc. System and method for area-efficient three-level dynamic element matching
US8085177B2 (en) * 2009-09-22 2011-12-27 Mediatek Singapore Pte. Ltd. Digital to analog converter system and method with multi-level scrambling
US7916058B1 (en) * 2009-10-05 2011-03-29 Texas Instruments Incorporated Digital-to-analog converter (DAC) with reference-rotated DAC elements
US8253612B2 (en) * 2009-10-16 2012-08-28 Realtek Semiconductor Corp. Self-calibrating R-2R ladder and method thereof
US10720919B2 (en) 2011-11-16 2020-07-21 Analog Devices, Inc. Apparatus and methods for reducing charge injection mismatch in electronic circuits
US8810443B2 (en) 2012-04-20 2014-08-19 Linear Technology Corporation Analog-to-digital converter system and method
US9124287B1 (en) 2014-12-22 2015-09-01 Pmc-Sierra Us, Inc. Scrambler with built in test capabilities for unary DAC
US9735799B1 (en) 2016-07-29 2017-08-15 Analog Devices, Inc. Envelope-dependent noise-shaped segmentation in oversampling digital-to-analog converters
US9768800B1 (en) 2016-09-13 2017-09-19 Analog Devices, Inc. Envelope dependent output stage scalability
US11683624B1 (en) 2020-05-12 2023-06-20 Qualcomm Technologies, Inc. Transducer with analog and digital modulators
US11757466B2 (en) 2020-08-10 2023-09-12 Analog Devices, Inc. System and method for dynamic element matching for delta sigma converters

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4703310A (en) * 1980-07-09 1987-10-27 U.S. Philips Corporation Digital/analog converter with capacitor-free elimination of a.c. components
US5221926A (en) * 1992-07-01 1993-06-22 Motorola, Inc. Circuit and method for cancelling nonlinearity error associated with component value mismatches in a data converter
US5305004A (en) * 1992-09-29 1994-04-19 Texas Instruments Incorporated Digital to analog converter for sigma delta modulator

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0253950B1 (en) * 1986-07-21 1991-07-17 Deutsche ITT Industries GmbH Monolithic integratable digital-analog converter
US4851841A (en) * 1987-10-02 1989-07-25 Crystal Semiconductor Corporation Gain scaling of oversampled analog-to-digital converters
US5128317A (en) * 1990-09-27 1992-07-07 International Flavors And Fragrances Inc. Camphonyl spirocyclooxaoctane-containing compositions, organoleptic uses thereof and process for preparing same

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4703310A (en) * 1980-07-09 1987-10-27 U.S. Philips Corporation Digital/analog converter with capacitor-free elimination of a.c. components
US5221926A (en) * 1992-07-01 1993-06-22 Motorola, Inc. Circuit and method for cancelling nonlinearity error associated with component value mismatches in a data converter
US5305004A (en) * 1992-09-29 1994-04-19 Texas Instruments Incorporated Digital to analog converter for sigma delta modulator

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP0712549A4 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
SG83798A1 (en) * 1999-06-07 2001-10-16 Nippon Precision Circuits Delta sigma d/a converter
GB2350950B (en) * 1999-06-07 2003-07-23 Nippon Precision Circuits Delta sigma D/A converter
GB2350950A (en) * 1999-06-07 2000-12-13 Nippon Precision Circuits Delta sigma D/A converter
US6797491B2 (en) 2000-06-26 2004-09-28 Stressgen Biotechnologies Corporation Human papilloma virus treatment
US9276540B2 (en) 2006-05-21 2016-03-01 Trigence Semiconductors, Inc. Digital/analogue conversion apparatus
US8423165B2 (en) 2006-05-21 2013-04-16 Trigence Semiconductor, Inc. Digital/analogue conversion apparatus
US9681231B2 (en) 2006-05-21 2017-06-13 Trigence Semiconductor, Inc. Digital/analog conversion apparatus
US9226053B2 (en) 2008-06-16 2015-12-29 Trigence Semiconductor, Inc. Digital speaker driving apparatus
US9693136B2 (en) 2008-06-16 2017-06-27 Trigence Semiconductor Inc. Digital speaker driving apparatus
US9300310B2 (en) 2009-12-09 2016-03-29 Trigence Semiconductor, Inc. Selection device
US9735796B2 (en) 2009-12-09 2017-08-15 Trigence Semiconductor, Inc. Selection device
US9544691B2 (en) 2009-12-16 2017-01-10 Trigence Semiconductor, Inc. Acoustic playback system
US9219960B2 (en) 2009-12-16 2015-12-22 Trigence Semiconductor Inc. Acoustic playback system

Also Published As

Publication number Publication date
JPH09501287A (en) 1997-02-04
DE69426266T2 (en) 2001-05-31
DE69426266D1 (en) 2000-12-14
US5404142A (en) 1995-04-04
EP0712549A1 (en) 1996-05-22
EP0712549B1 (en) 2000-11-08
EP0712549A4 (en) 1998-04-15

Similar Documents

Publication Publication Date Title
US5404142A (en) Data-directed scrambler for multi-bit noise shaping D/A converters
US5684482A (en) Spectral shaping of circuit errors in digital-to-analog converters
Galton Spectral shaping of circuit errors in digital-to-analog converters
US6522277B2 (en) Circuit, system and method for performing dynamic element matching using bi-directional rotation within a data converter
CN101160723B (en) System and method for tri-level logic data shuffling for oversampling data conversion
US7561088B1 (en) Multi-loop data weighted averaging in a delta-sigma DAC
US6842130B2 (en) Hardware-efficient implementation of dynamic element matching in sigma-delta DAC's
KR100911427B1 (en) The method for matching dynamic element and multi-bit digital-analog converter using the same, delta-sigma modulator and delta-sigma digital-analog converter including the multi-bit digital-analog converter
US7030792B2 (en) Suppressing digital-to-analog converter (DAC) error
US20050146454A1 (en) State-delayed technique and system to remove tones of dynamic element matching
US5349353A (en) Method and apparatus for mixed analog and digital processing of delta modulated pulse streams including digital-to-analog conversion of a digital input signal
US7205913B2 (en) Efficient data-directed scrambler for noise-shaping mixed-signal converters
Rakuljic et al. Tree-structured DEM DACs with arbitrary numbers of levels
US6867721B1 (en) Spectral shaping dynamic encoder for a DAC
KR100398013B1 (en) Selection circuit, d/a converter and a/d converter
US6753799B2 (en) Randomizer for sigma-delta-type converter
De Maeyer et al. Addressing static and dynamic errors in unit element multibit DACs
Zhang et al. A segmented data-weighted-averaging technique
JP2001077692A (en) D/a converting circuit
KR20040065290A (en) Digital to analogue converter
JPH08102670A (en) Oversampling d/a converter
KR20060009317A (en) Recursive bit-stream converter and method for recursive bit-stream conversion
KR950002302B1 (en) A/d converter
CN113078903A (en) System and method for improving matching in a signal converter
CN110912560A (en) Multimode oversampling analog-to-digital converter with reconfigurable data weighted average

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): JP

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 1994925188

Country of ref document: EP

WWP Wipo information: published in national office

Ref document number: 1994925188

Country of ref document: EP

WWG Wipo information: grant in national office

Ref document number: 1994925188

Country of ref document: EP