0

Here is my basic understanding of common mode range for diff input amplifiers:

As far as I understand the common mode voltage is nothing but the offset at which the diff signal is travelling above a common reference.

And if this CM voltage exceeds the amplifier rail voltages due to capacitive coupling or EMI, this is a problem. So as a remedy, common mode chokes(for high freq.) or filters(for lower freq.) are employed to attenuate this CM voltage so that the amplifier can operate well.

It seems like chokes and filters are frequency dependent and they attenuate CM voltages if the CM voltages are alternating/varying in time.

But can there be an interference or noise source which causes increase in CM voltage as DC? In other words is there a phenomenon called DC common mode interference in practice? If so, what are the ways to eliminate it?

floppy380
  • 1,845
  • 7
  • 38
  • 74

1 Answers1

1

In other words is there a phenomenon called DC common mode interference in practice?

Yes and it's numerically embodied for any decent op-amp under the heading CMRR (common-mode rejection ratio). A good op-amp will usually have a graph indicating something like this: -

enter image description here

And that good op-amp will have a number quoted like this (made-up example): -

CMRR = 95 dB, DC to 1 kHz.

In addition there will be a clause to that number that says something like: -

For input voltages between -10 volts and +10 volts. Vsup = +/- 15 volts

In other words if you raised or lowered the common voltage by 10 volts the CMRR would remain better than 95 dB. Clearly this gets worse at higher frequencies due to the inability to control parasites and the falling open-loop gain.

For a differentail amplifier made from 4 resistors and an op-amp it will be a lot worse because the resistor tolerances will cause offsets.

can there be an interference or noise source which causes increase in CM voltage as DC?

Anything that transmits data and uses phantom powering of a remote device is liable to this problem.

Andy aka
  • 456,226
  • 28
  • 367
  • 807
  • All electromagnetic issues due to interference is AC in nature. Can you give me an example of an "external noise source" which would increase the "DC" CM voltage of diff inputs? Normally they use chokes and filters for AC interference. What can cause an increase in DC common mode voltage as an external interference? – floppy380 Jun 03 '18 at 15:28
  • Even in your graph min freq. is 10Hz not DC – floppy380 Jun 03 '18 at 15:29
  • Can you see that the X axis is logarithmic therefore they will never show DC on a graph like that but I'm sure anyone is capable of predicting that the graph implies 95 dB at DC (as the text would state in the data sheet. – Andy aka Jun 03 '18 at 15:32
  • I can't think of an example of an external noise source that would increase the measurable DC level but certainly EMI can reduce the CMRR of an amplifier due to HF rectification. – Andy aka Jun 03 '18 at 15:34
  • I see so there can also be DC common mode issues due to conductive interference. But my question was what component or method is used to suppress "DC common mode voltages". For AC interference chokes and filter but what about for DC? – floppy380 Jun 03 '18 at 15:35
  • Do you agree that chokes and capacitor filters are useless for DC common mode suppression? – floppy380 Jun 03 '18 at 15:36
  • Imagine 1kV DC common mode voltage and amplifier rails are 10V. How can this DC CM voltage suppressed to levels acceptable by the amplifier. If it were AC they would use filter and chokes – floppy380 Jun 03 '18 at 15:37
  • If you have a DC rail of 1 kV superimposed on a differential signal (that contains useful 0 Hz information) then you have to attenuate to bring each line within the common mode input range of the differential receiver. If the signal contains no useful 0 Hz information then get rid of the 1 kV DC using a series capacitor in each line (and protection too). – Andy aka Jun 03 '18 at 15:40
  • There is an excellent app note on how resistors affect CMRR: http://www.analog.com/media/en/reference-design-documentation/design-notes/dn502f.pdf – Peter Smith Jun 03 '18 at 15:49
  • @Andyaka Yes it was my question indeed. A pure 1kV DC common mode interference "somehow" suddenly superimposed in a system where the diff amp rails are only +-10V. If 1kV were CM AC interference they could attenuate it down to 5V by chokes or filters(imao). So they would have a precaution for a possible AC interference(high freq. at least). Your last comment is a very good solution. How about an isolation amplifier before the diff amp? – floppy380 Jun 03 '18 at 16:29
  • An isolation amplifier could also work but the devil is in the detail. – Andy aka Jun 03 '18 at 16:44