0
\$\begingroup\$

I want to include to my design an 'variable offset adjustment' circuit using an operational amplifier. This circuit would allow me to ajust the offset of my design. In my circuit, I already has a buffer (non-inverting unity-gain, voltage follower) in my input, so, I think maybe I can use this op-amp with a potentiometer to ajust the voltage offset level. The thing is that I have heard about this 'variable offset adjustment' circuits, but I don't have any idea of how they look. If I look it on the internet I just found op-amp in which both inputs are connected to GND and that does not make sense to me (and would be against my idea of using the existing buffer). How can I do this? I don't think if it is better to do it in the voltage followers or in the differential amplifier. I don't know if I have this opportunity in the "offnull" pins, the datasheets does not give any clue (maybe this means, that the package has no offset pins?). Would a better solution be to change the part numbers for zero-drift (zero-offset) amplifiers? Would I need offset adjustment in both parts? I need a correction of milivolts.

enter image description here

EDITED: Why this design would not work properly according to an answer to this question?

enter image description here

EDITED 2: In order to understand the discussion and to save a final solution:

enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ Post a schematic of your circuit, or at least the part of it that already has a buffer. Include information on the range of signals that you need to pass, the range of offsets that you need to apply, the power rails, the operating speed. Bear in mind that some solutions are better than others for added noise, or resistance to drift, so some sort of hint as to the sensitivity of the application to those would be useful. Only vague generalities are possible without this level of detail. \$\endgroup\$
    – Neil_UK
    May 24 at 5:45
  • \$\begingroup\$ @Neil_UK Edited. \$\endgroup\$
    – Developing Electronics
    May 24 at 6:05
  • 1
    \$\begingroup\$ Mandatory read on 741: electronics.stackexchange.com/questions/304521/… \$\endgroup\$
    – winny
    May 24 at 8:03

2 Answers 2

Reset to default
1
\$\begingroup\$

The 741 is not a very good op-amp. Also you should not use the offset adjust pins on an op-amp to adjust anything other than the offset of that particular op-amp since you'll otherwise likely be introducing temperature drift.

The first thing is to identify where the zero offset that you want to trim out is coming from. If it's from the amplifiers, you can use amplifiers with lower offset voltage or so-called "zero-drift" types. If it's coming from elsewhere you should estimate the maximum possible offset and figure out how stable that offset adjustment has to be. The more stable it has to be, the more expensive it will be. For example, you might have to create two stable precision references (say +/-2.5V) for the trimpot and buffer the wiper voltage with another op-amp.

In the case shown, if you replaced the three op-amps with a good instrumentation amplifier such as INA849 you could could simplify it a bit. Maximum room-temperature input-referred offset is 35uV so with a gain of 100 you'd have +/-3.5mV at the output with no adjustment. This one is not the "zero-drift" type so it lacks the weirdness at the inputs (transient current spikes).

\$\endgroup\$
7
  • \$\begingroup\$ Thanks for your really good answer. I have two doubts here: 1. I did not understand the last sentences. Does it just mean that it has not low offset voltage? 2. You said I would need an stable reference and another op-amp to read it. Why the new image I have added to the original question would no be a good design? Without reference voltage in IC or another op-amp to read the offset adjustment. \$\endgroup\$
    – Developing Electronics
    May 24 at 9:25
  • \$\begingroup\$ It has 35uV input referred offset voltage, maximum. That is low in many, perhaps most, contexts. You didn't specify if your offset was input or output offset (100:1 difference between them). Your edit is not a good circuit- most importantly the added parts affect the gain by a large amount (and it changes with the offset adjust), the offset adjust range is huge, and it depends on the power supply stability. \$\endgroup\$
    – Spehro Pefhany
    May 24 at 9:30
  • \$\begingroup\$ My offset comes from the inputs. They come from a sensor. Ok, I get it. The huge range could be solved by swithing from +-15V to, for example, 0-5V in the adjustment, I guess. The stability of the PS could be a problem. And the gain change... I did not know that. Thank you. All clear. Do you know where can I find more information/resources on this topic and design tips? \$\endgroup\$
    – Developing Electronics
    May 24 at 9:36
  • \$\begingroup\$ The problem with the gain is that 'looking into' the top of R3 there needs to be 100kΩ to ground or some other 'stiff' voltage source. So here is where you might have an op-amp output on the other side of a 100kΩ (not 50k) R3. In that case the op-amp output would have a gain of +1 so +/-200mV would change the output by +/-200mV and would correspond to +/-2mV offset at the inputs. Lots of design tips in TI, Analog Devices, etc. application notes and looking at (professional) designs and figuring out why they did things is always good. \$\endgroup\$
    – Spehro Pefhany
    May 24 at 9:51
  • 1
    \$\begingroup\$ Yes, that's okay, but I would add a voltage divider after the wiper. Eg. 100k/2k to reduce it to about +/-300mV Otherwise the trimpot will be too touchy. \$\endgroup\$
    – Spehro Pefhany
    May 24 at 11:09
1
\$\begingroup\$

Taking your schematic at face value, the simplest way to add an offset would be to take R3 to a variable voltage (edit) as your edit has just done! (/edit). Not all of those resistors are needed, and it isn't centred at the 5 V offset your first circuit has, but it would do the job of basically making a third voltage term affect the output, as well as the two input voltages.

The problem with this approach is drift. Ideally an offset will be cured at source. If you have an offset from somewhere, and compensate with an offset derived from a different place, then they both have the potential to drift in a different way, especially with temperature. You have shown your adjustment circuit coming from the the +15 V and -15 V rails. How stable are these? Do changes in them track the changes in the offset you are trying to compensate for?

\$\endgroup\$
3
  • \$\begingroup\$ Why not all the resistors are needed? In the first design it was a fixed +5V offset adjustment. In the second one, I can change the value that it is going to be added to the output. My offset comes from a sensor. My +-15V signals come from a good DC/DC. \$\endgroup\$
    – Developing Electronics
    May 24 at 9:41
  • 1
    \$\begingroup\$ @DevelopingElectronics If you're not too worried about exact gain or CMRR, then use resistors, but R4 and R5 are superfluous. If you do want good gain and CMRR, then you need to use the circuit of your second edit. The adjustment is very coarse though. Depending on how much offset you want to include, you could reduce the sensitivity of your potentiometer by putting a fixed resistor in series with either or both ends, to reduce the range of voltages it can select. \$\endgroup\$
    – Neil_UK
    May 24 at 10:52
  • \$\begingroup\$ Super! Thanks a lot \$\endgroup\$
    – Developing Electronics
    May 24 at 10:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.