INPUT STAGE FOR A SIGNAL PROCESSING CIRCUIT
20250247063 · 2025-07-31
Assignee
Inventors
Cpc classification
H03F2200/261
ELECTRICITY
International classification
Abstract
An input stage for a signal processing circuit, the input comprising: a first differential pair; and a second differential pair; wherein the first differential pair and the second differential pairs have opposing skews which controls a voltage offset of the input stage.
Claims
1. An input stage for a signal processing circuit, the input stage comprising: a first differential pair; and a second differential pair; wherein the first differential pair and the second differential pair have opposing skews which controls a voltage offset of the input stage.
2. The input stage of claim 1, wherein the signal processing circuit is one of an amplifier or a comparator.
3. The input stage of claim 2, wherein the amplifier is an operational transconductance amplifier.
4. The input stage of claim 1, wherein the first differential pair comprises a first transistor and a second transistor such that the first transistor and second transistor have a skew in a first direction.
5. The input stage of claim 4, wherein the second differential pair comprises a third transistor and a fourth transistor such that the third transistor and fourth transistor have a skew in a second direction, where the second direction is opposite to the first direction.
6. The input stage of claim 5, wherein the first transistor, second transistor, third transistor and fourth transistor are one of a metal oxide semiconductor or a bipolar transistor.
7. The input stage of claim 5, wherein the first transistor, second transistor, third transistor and fourth transistor each comprise an output terminal such that the output terminal of the first transistor and the third transistor are coupled together to form a positive node and the output terminal of the second and the fourth transistor are coupled together to form a negative node.
8. The input stage of claim 7, wherein the positive node is configured to carry a first current and the negative node is configured to carry a second current.
9. The input stage of claim 8, wherein the first transistor and the third transistor each comprise a control terminal, the control terminal being configured to receive a positive input, such that the positive inputs of the control terminals of the first transistor and third transistor are coupled together.
10. The input stage of claim 9, wherein the second transistor and the fourth transistor each comprise a control terminal, the control terminal being configured to receive a negative input, such that the negative inputs of the control terminals of the second transistor and fourth transistor are coupled together.
11. The input stage of claim 10, wherein the first transistor, second transistor, third transistor and fourth transistor each further comprise an input terminal such that the input terminals of the first transistor and the second transistor are together coupled to form a first tail node and the input terminal of the third transistor and the fourth transistor are coupled together to form a second tail node.
12. The input stage of claim 10, wherein the first tail node and the second tail node are configured to receive a bias current.
13. The input stage of claim 12, wherein the first current and the second bias currents are equal and the bias currents are not adjusted.
14. The input stage of claim 12, wherein the first current and the second current are not equal and either: the bias current at the first tail node is adjusted; or the bias current at the second tail node is adjusted; or the bias current at the first tail node and the second tail node are adjusted.
15. The input stage of claim 14, wherein the first current and the second current are equal after the bias current is adjusted.
16. The input stage of claim 15, wherein the bias current is adjusted by a trimming current.
17. The input stage of claim 16, wherein: when the bias current is adjusted by the trimming current, the voltage offset is controlled such that the voltage offset of the input stage is minimised.
18. The input stage of claim 16, wherein: when the bias current is adjusted by the trimming current, the voltage offset is controlled such that the voltage offset of the input stage is set to a pre-determined value.
19. The input stage of claim 16, wherein either: the trimming current is added to or taken away from the bias current at the first tail node; or the trimming current is added to or taken away from the bias current at the second tail node; or the trimming current is added to the bias current at the first tail node and the trimming current is taken away from the bias current at the second tail node; or the trimming current is taken away from the bias current at the first tail node and the trimming current is added to the bias current at the second tail node.
20. The input stage of claim 19, wherein the trimming current has an absolute value dependent on the skew of the first differential pair, the skew of the second differential pair and the voltage offset of the input stage.
21. An apparatus comprising: a signal processing circuit, an input stage for the signal processing circuit, the input stage comprising a first differential pair and a second differential pair wherein the first differential pair and the second differential pair have opposing skews which controls a voltage offset of the input stage.
22. A method for controlling a voltage offset of an input stage for a signal processing circuit, the method comprising: providing the input stage with a first differential pair and a second differential pair, wherein the first differential pair and the second differential pair have opposing skews.
23. The method of claim 22, wherein: the first differential pair and the second differential pair are configured to receive a bias current; and the bias current of either the first differential pair, or the second differential pair, or both the first and the second differential pair is adjusted by a trimming current such that the voltage offset of the input stage is controlled.
Description
DESCRIPTION OF THE DRAWINGS
[0042] The disclosure is described in further detail below by way of example and with reference to the accompanying drawings, in which:
[0043]
[0044]
[0045]
[0046]
[0047]
[0048]
DESCRIPTION
[0049]
[0050] After the input stage 110, the OTA is formed of a current mirror comprising transistors P1, P2, P3, P4, N1 and N2. The current mirrors take the amplified V.sub.diff from the differential input pair and replicates the output current to pass onto the next stage. This next stage represents any circuit that an amplifier or OTA can be part of. The components N0p, Non, P1, P2, P3, P4, N1 and N2 all form part of the signal path for the OTA 100 as they are used to transmit the signal to the next stage of the circuit.
[0051] Ideally, the two transistors N0p and N0n are of the same size, in other words they have a 1:1 ratio. Therefore, in ideal operation if V.sub.pos=V.sub.neg then the OTA 100 would be balanced and there would be no current present at OUT. However, POS and NEG are not identical due to a mismatch in tolerance during manufacturing of the transistors. This results in an inbuilt voltage offset into the OTA focussed at the input stage 110. There are a number of known ways in the prior art to account for this voltage offset. A selection are shown in
[0052] Technique 1 involves adding a digitally controlled resistor or trimmer into the sources of the input differential pair in order to digitally trim and change the current in either the positive branch or the negative branch such that the current mirror is balanced. However, this technique increases the complexity of the design of the OTA 100 and results in a decreased gain of the input stage 110 and can potentially degrade power supply rejection ratio by coupling noise into this sensitive nets. Technique 2 involves trimming of the voltage offset occurs in the current mirror by adding a bank of variable switches to form a digitally controlled current mirror. However, this increases the complexity of the design of the OTA 100 and also increases the parasitic capacitance which has a negative impact on the overall speed of the OTA. In technique 3, the trimming occurs before the OUT stage by either adding or removing a trimmed current value. However, this technique adds parasitic capacitance to the high impedance node, increases the complexity of design of OTA 100 and also slows the response of the OTA 100.
[0053] There are other techniques, however all the techniques shown in
[0054]
[0055] The first differential pair 210 comprises a first transistor N0a and a second transistor N0b. The transistors N0a and N0b have a skew 212 in a first direction N:1. The second differential pair 220 comprises a third transistor N0c and a fourth transistor N0d. The transistors N0c and N0d have a skew 222 in a second direction 1:N. The skew 222 of the second differential pair 220 is in an opposite direction to the skew 212 of the first differential pair 210. The first and second skews each have a skewing factor N. The skewing factor N is set by designer with respect to the maximum possible mismatch of the input pair within some margin. It indicates the strength of one transistor in the differential pair relative to the other. For example, for the first differential pair 210, the skew is in the direction of N0a. This means that the first transistor N0a is stronger, and hence, produces a larger output current than the second transistor N0b. For the second differential pair 210, the skew is in the direction of N0d, such that the fourth transistor N0d produces a larger output than the third transistor N0c. Both N0a and N0d have the same skewing factor, therefore the skew of the first differential pair 210 and the second differential pair 220 is equal and opposite. The transistors N0a, N0b, N0c and N0d can be, for example, a metal oxide semiconductor or a bipolar transistor. Each of the first N0a, second N0b, third N0c and fourth N0d transistors comprise an output terminal o. The output terminal o of the first transistor N0a and the third transistor N0c are coupled together to form a positive node 230 which is configured to carry a first current I.sub.POS. The output terminal o of the second transistor N0b and the fourth transistor N0d are coupled together to form a negative node 240 which is configured to carry a second current I.sub.NEG. The first current I.sub.POS and the second current I.sub.NEG then passed onto the next stage. The next stage refers to the signal processing circuit that the input stage 200 is part of. This could be, for example, an amplifier or a comparator.
[0056] The first transistor N0a and the third transistor N0c each comprise a control terminal c. The control terminals c are configured to receive a positive input POS. The positive inputs POS for the control terminals c of the first and third transistors N0a and N0c are coupled together. The second transistor N0b and the fourth transistor N0d each comprise a control terminal c. The control terminals c are configured to receive a negative input NEG. The negative inputs NEG for the control terminals c of the second and fourth transistors N0b and N0d are coupled together.
[0057] Each of the first N0a, second N0b, third N0c and fourth N0d transistors further comprise an input terminal i. The input terminal i of the first transistor N0a and the second transistor N0b are connected together to form a first tail node T1. The input terminal i of the third transistor N0c and the fourth transistor N0d are connected together to form a second tail node T2. Both the first tail node T1 and the second tail node T2 are configured to receive a bias current I.sub.BIAS.
[0058] The first, second, third and fourth transistors N0a, N0b, N0c and N0d could be, for example, a metal oxide semiconductors. For example, therefore, the control terminal c could be a gate terminal, the input terminal i could be a source terminal and the output terminal o could be a drain terminal.
[0059] In the embodiment of
[0060]
[0061] The input stage 200 is the theoretical ideal embodiment of the present disclosure, however, either the first differential pair 210 or the second differential pair 220 will be produce a larger current than the other due to manufacturing tolerances or other unpredictable random process variations. This mismatch exists as the manufacturing process is unable to create two completely identical transistors. Whilst they can be manufactured to be similar, they will never be exactly the same. The skewing N of the first differential pair and the second differential pair is chosen to be at least equal to or greater than the maximum expected manufacturing mismatch. However, as the mismatch can be greater than expected the skewing N cannot compensate for it alone in all cases. Therefore, the first current I.sub.POS is not equal to I.sub.NEG which is undesirable. Therefore, the bias current I.sub.BIAS is adjusted by a trimming current I.sub.TRIM. The trimming current I.sub.TRIM is pre-set adjustment before implementing the input stage 300A in an amplifier or a comparator. The trimming current I.sub.TRIM is set by shorting the positive input POS and the negative input NEG and changing the value of I.sub.TRIM until the signal processing circuit trips (in the case of a comparator) or the output of the signal processing circuit goes from one rail to the other (in the case of the amplifier). The bias current can be adjusted at the second tail node T2 as shown in the input stage 300A by feature 310. The bias current I.sub.BIAS is adjusted by either adding or removing a trimming current I.sub.TRIM. The value of the trimming current I.sub.TRIM could be either positive or negative. The adjustment to the bias current I.sub.BIAS does not need to occur at the second tail node T2 but could happen instead at the first tail node T1.
[0062] In the embodiment of
[0063]
[0064] The input stage 300A applies one-sided trimming. Applying an adjustment to the bias current I.sub.BIAS at only either the first tail node T1 or the second tail node T2 can change the DC operation point in the output of the signal processing circuit. For example, in an ideal circuit the expectation is that I.sub.POS=I.sub.NEG=I.sub.BIAS when the input differential voltage is zero. However, for the embodiment 300A this is not true as the one-sided trimming changes the tail current such that I.sub.POS=I.sub.NEG but these currents are not equal to I.sub.BIAS. This is due to the addition or subtraction of I.sub.TRIM to one side only. Therefore, the steady-state biasing of the circuit (DC) is not kept the same and has been changed. To overcome this, the embodiment shown in
[0065]
[0066]
[0067] For all embodiments of the present disclosure, the value of I.sub.TRIM and the skewing factor are related and are chosen depending on the purpose of the signal processing circuit that the input stage is to be used for or the size of the circuit. Smaller input stages will require a higher skewing factor, as smaller devices have a larger percentual error in, for example, size and doping. The range for the trimming current I.sub.TRIM must be set accordingly. The skewing factor N and the range of I.sub.TRIM must be large enough to compensate for the fabrication mismatch. If the skewing factor is large, the range of I.sub.TRIM does not need to be so large. On the other hand, if the skewing factor is marginal then the I.sub.TRIM range will be equal to I.sub.BIAS.
[0068] The example embodiments of
[0069] Various improvements and modifications may be made without departing from the scope of the disclosure.
[0070] A skilled person will therefore appreciate that variations of the disclosed arrangements are possible without departing from the disclosure. Accordingly, the above description of the specific embodiments is made by way of example only and not for the purposes of limitation. It will be clear to the skilled person that minor modifications may be made without significant changes to the operation described.