Author Topic: Input signal appearing on the +Vpin of an opamp  (Read 2070 times)

0 Members and 1 Guest are viewing this topic.

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Input signal appearing on the +Vpin of an opamp
« on: September 13, 2020, 09:18:23 pm »
A weird question I know, but some years back at college I remember while we were working on some op amp circuitry we detected our sine wave input appear on the +V pin of the op amp. Can't remember what we made of it back then.
This can't be normal right? I think I am having a similar issue on a project I am working on right now. Unfortunately I haven't got a scope to test it out at the moment.
Anyone had a similar experience? Is it a common failure mode for op amps?
Thanks
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Input signal appearing on the +Vpin of an opamp
« Reply #1 on: September 13, 2020, 09:28:04 pm »
What configuration are you talking about?

Are you talking about the positive power supply rail? It sounds like the isn't enough supply decoupling, or the impedance of the power supply is too high.
 

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Re: Input signal appearing on the +Vpin of an opamp
« Reply #2 on: September 13, 2020, 09:41:38 pm »
What configuration are you talking about?

Are you talking about the positive power supply rail? It sounds like the isn't enough supply decoupling, or the impedance of the power supply is too high.
Yep on the positive supply. I am working on a guitar practice amp at the moment and this is on the preamp stage. It's in an inverting amplifier configuration.
Could I please ask you to explain what you meant by "the impedance of the power supply is too high".
Thanks a million
 

Offline niconiconi

  • Frequent Contributor
  • **
  • Posts: 366
  • Country: cn
Re: Input signal appearing on the +Vpin of an opamp
« Reply #3 on: September 14, 2020, 09:00:30 am »
What configuration are you talking about?

Are you talking about the positive power supply rail? It sounds like the isn't enough supply decoupling, or the impedance of the power supply is too high.
Could I please ask you to explain what you meant by "the impedance of the power supply is too high".

What Zero999 said is that, when the opamp tries to amplify a signal, say, a sine wave, its current consumption increases. If the power supply impedance is too high (high resistance connection, no bypass/decoupling, or if the circuit is somehow broken and draws more current than expected), the supply voltage at +V drops according to Ohm's Law. And after the sine wave peaks, voltage starts to decrease, and the current consumption of the opamp decreases again, voltage drop at +V decreases. The final result is that you see the output signal of the opamp has "leaked" into the +V rail. Another related problem is common-impedance coupling - if many chips are connected to the same power supply without any bypass/decoupling capacitance, the behavior of one chip can cause the power rail voltage to drop, and this drop affects other chips as well.

Not sure if it's related to your problem, just to explain "the impedance of the power supply is too high".
« Last Edit: September 14, 2020, 09:02:16 am by niconiconi »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Input signal appearing on the +Vpin of an opamp
« Reply #4 on: September 14, 2020, 11:06:53 am »
See the simulation of an inverting amplifier, with a power supply with a resistance of 100R. The red trace shows the output signal superimposed on the supply voltage.
 

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Re: Input signal appearing on the +Vpin of an opamp
« Reply #5 on: September 14, 2020, 11:14:29 am »
Thanks @niconiconi and @Zero999 for the detailed explanation. Sorry to sound like a complete noob but here goes - This would be considered as a fault or a bad design right? It's not supposed to happen?
In a case like this would it indicate that the op amp is faulty or needs to be changed or the decoupling or the bypass capacitors needs to be changed?
Thanks for the help really do appreciate it.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Input signal appearing on the +Vpin of an opamp
« Reply #6 on: September 15, 2020, 09:59:03 am »
Thanks @niconiconi and @Zero999 for the detailed explanation. Sorry to sound like a complete noob but here goes - This would be considered as a fault or a bad design right? It's not supposed to happen?
In a case like this would it indicate that the op amp is faulty or needs to be changed or the decoupling or the bypass capacitors needs to be changed?
Thanks for the help really do appreciate it.
No, the op-amp is fine. It will be due to a bad connection/power supply, or poor decoupling.
 
The following users thanked this post: Yamin

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: Input signal appearing on the +Vpin of an opamp
« Reply #7 on: September 16, 2020, 03:23:18 pm »
If one op-amp impresses that much input signal on a power-supply line, then you will see cross-talk (unwanted coupling) to any other amplifiers sharing that power-supply line.
 
The following users thanked this post: Yamin

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Re: Input signal appearing on the +Vpin of an opamp
« Reply #8 on: September 17, 2020, 09:50:58 am »
Umm..I did come across some material which points out that this could be done on purpose by the designer (as in, its designed so that the VS pin follows the output  :-//). That can't be right??

So to troubleshoot the unit I am working should I have a look at the decoupling caps on the power rail? Should I add a cap on the power rails to see if I could rectify the problem I am having? If so what would be a good value to check?

Thanks so much 
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Input signal appearing on the +Vpin of an opamp
« Reply #9 on: September 17, 2020, 12:25:28 pm »
If one op-amp impresses that much input signal on a power-supply line, then you will see cross-talk (unwanted coupling) to any other amplifiers sharing that power-supply line.
It depends on who sensitive the circuit is. Normally the power supply rejection of the op-amp will handle variations in power supply voltage.

Umm..I did come across some material which points out that this could be done on purpose by the designer (as in, its designed so that the VS pin follows the output  :-//). That can't be right??
Yes, it's posible to design an op-amp circuit like that, so the output and power supply rail share the same connection.

Quote
So to troubleshoot the unit I am working should I have a look at the decoupling caps on the power rail? Should I add a cap on the power rails to see if I could rectify the problem I am having? If so what would be a good value to check?

Thanks so much
Check the connections to the power supply first. What sort of power supply are you using?
 

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Re: Input signal appearing on the +Vpin of an opamp
« Reply #10 on: September 17, 2020, 02:40:06 pm »
Quote
So to troubleshoot the unit I am working should I have a look at the decoupling caps on the power rail? Should I add a cap on the power rails to see if I could rectify the problem I am having? If so what would be a good value to check?
Thanks so much
Check the connections to the power supply first. What sort of power supply are you using?

Well I am working on a practice guitar amp, there is tons of noise on the final output. Not even possible to hear the input signal. I'm troubleshooting around the pre amp section. There is a DSP section inbetween the final output and the preamp. The power supply is a 19V power brick (I have tried different power supplies). There are some LDO regulators and the 12V is derived from there.

Yes, it's posible to design an op-amp circuit like that, so the output and power supply rail share the same connection.

Could you please explain to me why that would be the case (sorry to sound like a noob).

Thanks again.
« Last Edit: September 17, 2020, 02:51:09 pm by Yamin »
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: Input signal appearing on the +Vpin of an opamp
« Reply #11 on: September 17, 2020, 02:54:12 pm »
If one op-amp impresses that much input signal on a power-supply line, then you will see cross-talk (unwanted coupling) to any other amplifiers sharing that power-supply line.
It depends on who sensitive the circuit is. Normally the power supply rejection of the op-amp will handle variations in power supply voltage.
Yes, a good op-amp has a decent PSRR, but now that single-supply op-amps are prevalent, the required bias divider (as in the OP's circuit) can easily couple more voltage to the non-inverting input than the equivalent offset from PSRR.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Input signal appearing on the +Vpin of an opamp
« Reply #12 on: September 18, 2020, 10:01:10 pm »
If one op-amp impresses that much input signal on a power-supply line, then you will see cross-talk (unwanted coupling) to any other amplifiers sharing that power-supply line.
It depends on who sensitive the circuit is. Normally the power supply rejection of the op-amp will handle variations in power supply voltage.
Yes, a good op-amp has a decent PSRR, but now that single-supply op-amps are prevalent, the required bias divider (as in the OP's circuit) can easily couple more voltage to the non-inverting input than the equivalent offset from PSRR.
The original poster hasn't posted a schematic. Are you referring to the one I posted? If so then yes, it doesn't have great power supply rejection, but that's easy to resolve by increasing C1 and/or adding another RC filter before the non-inverting input.

Quote
So to troubleshoot the unit I am working should I have a look at the decoupling caps on the power rail? Should I add a cap on the power rails to see if I could rectify the problem I am having? If so what would be a good value to check?
Thanks so much
Check the connections to the power supply first. What sort of power supply are you using?

Well I am working on a practice guitar amp, there is tons of noise on the final output. Not even possible to hear the input signal. I'm troubleshooting around the pre amp section. There is a DSP section inbetween the final output and the preamp. The power supply is a 19V power brick (I have tried different power supplies). There are some LDO regulators and the 12V is derived from there.
What sort of noise is it? A hum, buzzing noise or hiss?

Quote
Yes, it's posible to design an op-amp circuit like that, so the output and power supply rail share the same connection.

Could you please explain to me why that would be the case (sorry to sound like a noob).

Thanks again.
The main reason would be to simplify wiring, as only two conductors are needed, or a single wire if you can use earth return, but it makes the op-amp circuit more complex. Here's an example. The output of the op-amp is connected to its positive supply, which is connected to a power supply, via a resistor. The power supply and resistor could be at one end of a piece of cable, with the amplifier and traducer at the other. The amplifier has two different gains: DC and AC or signal frequency. At DC the op-amp has a gain of 1+R2/R1, which amplifies the voltage on D1 by 11, causing the output to sit at just over 6V. At AC, R3 is in parallel with R2, making 100k, so the amplifier has a gain of 2. For higher AC gains than 11, put R3 and C3 across R1.

« Last Edit: September 18, 2020, 10:09:05 pm by Zero999 »
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: Input signal appearing on the +Vpin of an opamp
« Reply #13 on: September 19, 2020, 01:49:34 pm »
Yes, I was referring to your posted schematic, which is typical for single-supply operation of op-amps.  (I was looking for a typical circuit and saw yours, but hastily referred to it as the OP's.)   PSRR of single-supply circuits, including bias chains, is an important design consideration.  I was involved in a re-design of a system with over a thousand dual-rail op-amps at the end of long power distribution wiring, and we had to add point-of-load regulators to avoid cross-talk from one end of the chain to the other due to power supply impedance.  (This was due to the PSRR of the op-amps themselves, and the class-B output stages pulling relatively large currents from the supply at high output voltage, thus affecting the voltage at other amplifiers far away, resulting in non-local crosstalk.)
« Last Edit: September 19, 2020, 05:48:53 pm by TimFox »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf