The present Application is based on International Application No. PCT/EP2004/052518, filed on Oct. 13, 2004, which in turn corresponds to French Application No 0312182, filed on Oct. 17, 2003, and priority is hereby claimed under 35 USC §119 based on these applications. Each of these applications are hereby incorporated by reference in their entirety into the present application.
The invention pertains to a comparison circuit for an analog/digital converter. The comparison circuit comprises a network of comparators each comparing an analog voltage to be converted with a reference voltage. The analog voltage to be converted generally arises from a sample-and-hold module allowing the whole assembly of comparators of the network to receive the same analog voltage at the moment at which they perform the comparison with the reference voltage.
The reference voltages received by the comparators are distributed over a range in which the analog voltage can vary. The distribution is generally uniform over the range and it is for example obtained by means of a network of resistors, all of like value and linked in series between the terminals of a source of supply voltage of the comparator. There are substantially as many resistors as comparators. The reference voltages are then tapped off at the various inter-resistor junction points.
Each comparator comprises two outputs, one direct and the other inverse. The voltages present on its outputs are dependent on the potential difference between the analog voltage and the reference voltage received by the comparator concerned.
For a given comparator, for example the comparator of rank n, if its response were perfect, the voltage On present on its direct output ought to be zero when the analog voltage V is equal to the reference voltage Vn. However, the response of the comparators is not perfect and a voltage mismatch, termed the offset voltage, is noted between the reference voltage Vref n and the analog voltage V causing a zero voltage On on the direct output of the comparator of rank n. In practice it is noted that each comparator C has its own offset voltage independent of that of the other comparators. In
Additionally, the resolution LSB of an analog digital converter may be expressed by the mismatch in the analog voltage modifying the value of a low-order bit at the output of the converter. The LSB resolution is expressed as follows:
where Vpeak/peak represents the maximum amplitude of the analog voltage that the converter can convert, and where n is the number of comparators in the network. If the resolution LSB is less than three times the offset voltage, there is a loss of linearity of the converter and the low-order bit is no longer meaningful.
The aim of the invention is to reduce the effects of these offset voltages by averaging them over neighboring converters. This reduction makes it possible to improve the resolution of the converter.
Accordingly, the subject of the invention is a comparison circuit for an analog/digital converter comprising a network of comparators each comparing an analog voltage to be converted with a reference voltage, the reference voltages being distributed over a range in which the analog voltage can vary, each comparator comprising a direct output and an inverse output, characterized in that each output, direct or inverse, is linked to the input of a voltage follower, the outputs of each voltage follower being connected either to inputs of a first network of resistors delivering at its outputs, mean voltages that are the average of those present on direct outputs of the comparators receiving reference voltages similar in their distribution over the range, or to inputs of a second network of resistors delivering at its outputs, mean voltages that are the averages of those present on inverse outputs of comparators receiving reference voltages similar in their distribution over the range.
The invention will be better understood and other advantages will become apparent on reading the detailed description of an embodiment given by way of example and illustrated by the appended drawing in which
Each output, direct On−1, On or On+1 or inverse
The outputs of each voltage follower A are connected either to an input of a first network 2 of resistors delivering at its outputs O′n−1, O′n and O′n+1, mean voltages that are the average of those present on the direct outputs of the comparators Cn−1, Cn and Cn+1, or to an input of a second network of resistors delivering at its outputs
Advantageously, each network of resistors comprises a first series assembly of two identical pairs of two identical resistors in series, R1, R2, on the one hand, R3, R4 on the other hand, and a second series assembly of two identical pairs of two identical resistors in series R5, R6 on the one hand, R7, R8 on the other hand. The inputs of the network of resistors are constituted by the ends and the midpoint of the first series assembly, and the outputs of the network of resistors are constituted by the ends and the midpoint of the second series assembly, the midpoint of the first pair and of the second pair of resistors of the first assembly are connected respectively to the midpoint of the first pair and of the second pair of the second assembly. This structure of network of resistors is repeated so as to be able to link up to the outputs of all the comparators C and thus provide as many outputs O′ of the network of resistors as outputs O of the comparators C.
The transfer function of the output O′n of the first network 2 can then be expressed in the following manner:
The first two networks of resistors make it possible to reduce the statistical error due to the various offset voltages of the comparators. More precisely, it is possible to determine the standard deviation a of the offset voltages of the assembly of comparators C of the network. It is possible, with the aid of the transfer function of the first resistor network to determine an equivalent standard deviation σ′ of the comparators as seen from the outputs of the first network 2 of resistors. The equivalent standard deviation σ′ may be expressed in the following manner:
σ′=σ√{square root over (⅜)}≈0.6σ
This reduction in the effect of the offset voltage of the comparators makes it possible practically to improve the resolution by a low-order bit.
The combination of the voltage followers A with the network of resistors makes it possible not to lose gain at the output of the network of resistors with respect to the output of the network of comparators. In the absence of a voltage follower A, the reduction in the effect of the offset voltage of the comparators would be lower.
Advantageously the outputs O′n−1, O′n and O′n+1 of the first network 2 of resistors are connected, by way of voltage followers A, to inputs of a third network 3 of resistors delivering to its outputs O″n−1, O″n and O″n+1 mean voltages that are the average of those present on neighboring inputs of the third network of resistors. Likewise, the outputs
As before, an equivalent standard deviation σ″ can be expressed in the following manner:
σ″=σ′√{square root over (⅜)}=σ√{square root over (⅜)}×√{square root over (⅜)}≈0.36σ
An appreciable decrease in the effect of the offset voltage of the comparators C can be seen here, this decrease being obtained with the aid of the second stage of network of resistors. The voltage followers A connected between the two networks of resistors avoid any loss of gain. The invention could be generalized by chaining together other networks of resistors, decorrelated from the previous networks by means of voltage followers, downstream of the two described here. Nevertheless, this chaining appreciably increases the number of components present on a substrate on which the analog digital converter is made.
The invention can be implemented in respect of a comparison circuit architecture comprising comparators all working in parallel. This architecture is well known in the literature by the name “flash”. The invention may also be implemented in respect of a so-called “folding” comparison circuit architecture comprising a smaller number of comparators working in parallel. These comparators are then used several times over the range. This architecture is well known in the literature.
It will be readily seen by one of ordinary skill in the art that embodiments according to the present invention fulfill many of the advantages set forth above. After reading the foregoing specification, one of ordinary skill will be able to affect various changes, substitutions of equivalents and various other aspects of the invention as broadly disclosed herein. It is therefore intended that the protection granted hereon be limited only by the definition contained in the appended claims and equivalents thereof.
Number | Date | Country | Kind |
---|---|---|---|
03 12182 | Oct 2003 | FR | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/EP2004/052518 | 10/13/2004 | WO | 00 | 4/11/2006 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2005/036754 | 4/21/2005 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
5517134 | Yaklin | May 1996 | A |
6169510 | Bult et al. | Jan 2001 | B1 |
Number | Date | Country | |
---|---|---|---|
20070241953 A1 | Oct 2007 | US |