Offset voltage, ground reference, dual supply

[Eddododo]

Trying to make sure I understand Op amp offset voltage, among other things..

first, am I correct in my understanding that offset voltage occurs whether it's a single-supply with a virtual ground, or a dual +/- supply with a ground reference? In that case, does that also mean that you don't gain an 'advantage' with dual supply /ground ref (DS/Gr) in terms of eliminating coupling capacitors, save maybe the input cap?

When there is significant offset in a DS/Gr circuit, is this affecting ground itself, or is it just that the signal 'hot' gains a DC potential over ground.

In this same case, what is happening to the subsequent stages as they are ALSO independently biased to ground? IE the first stage has .5V offset on the output, the second stage (no coupling cap) is inverted, and the + input is tied to ground.. wha' happing?

I appreciate any insight shared!

[merlinb]

TIn that case, does that also mean that you don't gain an 'advantage' with dual supply /ground ref (DS/Gr) in terms of eliminating coupling capacitors, save maybe the input cap?

Correct.

is it just that the signal 'hot' gains a DC potential over ground.

Yes, the output of the opamp simply idles at a small DC offset voltage (it may be positive or negative depending on circuit details), and the signal will ride on top of the DC.

In this same case, what is happening to the subsequent stages as they are ALSO independently biased to ground?

If there are no coupling caps, the DC offset at the output of the first opamp will be amplified by the second opamp just as if it was a signal voltage (the second opamp will also add its own offset), and so on.

[R.G.]

As Merlin says:

If there are no coupling caps, the DC offset at the output of the first opamp will be amplified by the second opamp just as if it was a signal voltage (the second opamp will also add its own offset), and so on.

Just to beat a dead horse a different way: people occasionally ask why opamps have DC-breaking caps between them if they all have the same reference voltage. This discussion is the answer to that question.

[antonis]

In this same case, what is happening to the subsequent stages as they are ALSO independently biased to ground?

If there are no coupling caps, the DC offset at the output of the first opamp will be amplified by the second opamp just as if it was a signal voltage (the second opamp will also add its own offset), and so on.

Assuming the dead horse isn't flogged enough, DC offset (in case of absent coupling cap) should be amplified only in case of succeeding op-amp being inverting..
In case of non-inverting, and considering DC gain rolled-off to unity, you can let DC offsets to accumulate and place a coupling cap in the spot where total DC offset is getting threatening to signal swing headroom..

[ElectricDruid]

I used to be into sado-necro-bestiality, but then I realised I was just flogging a dead horse.

 

[bluebunny]

I used to be into sado-necro-bestiality, but then I realised I was just flogging a dead horse.

We like to set up these gags for you.  Nicely done, Tom. 

[R.G.]

This discussion sparked a dim memory. A leftover from long ago was still in my archives. In today's refined and hair-triggered times, it's probably offensive to someone, but at the time it was intended only to shine a ray of sunshine on non-optimal business practices by humor. Reference to the original human group has been euphemized to further avoid offense.

[Insert image: fence post lying on the ground next to an empty hole; caption reads "Why was this post removed? To prevent a fence."]

No person, and no horses, were harmed in gathering this information. The sentiments are not mine, and I did not originate this document. It is presented here only as a humorous recollection of the state of the business world decades ago.

=========================
How To Ride A Dead Horse

[Plains horse-nomad] tribal wisdom says that when you discover you are riding a dead horse, the best strategy is to dismount.  However, in business we often try other strategies with dead horses, including the following:

1. Grimly continuing to beat the dead horse while saying things like "This is the way we always have ridden this horse."
2. Buying a stronger whip.
3. Changing riders.
4. Changing to another dead horse.
5. Appointing a committee to study the horse.
6. Visiting other sites to see how they ride dead horses.
7. Increasing the standards to ride dead horses.
8. Appointing a tiger team to revive the dead horse.
9. Creating a training/educational class to increase our riding ability.
10. Doing a comparative study on the state of dead horses in today's environment.
11. Revisiting the performance requirements for horses to determine if dead horses might not be acceptable for the immediate needs after all.
12. Changing the definition of "dead" so that the horse does not qualify under it, making the horse not technically "dead".
13. Hiring contractors to ride the dead horse more effectively and cheaply.
14. Harnessing several dead horses together for increased speed.
15. Hiring or appointing a team of expert consulting riders to see if a highly focused team approach can get the dead horse to work better.
16. Declaring that "No horse is too dead to beat."
17. Providing additional funding to increase the horse's performance.
18. Doing a Cost Accounting Study to see if contractors can ride it cheaper.
19. Purchasing a product to make dead horses run faster.
20. Declaring that the horse is actually "better, faster and cheaper" or "a better value" if it's dead.
21. Forming a quality circle to find uses for dead horses.
22. Saying that the horse was procured with cost as an independent variable.
23. Promoting the dead horse to a supervisory position.
24. Hiring consultants to review dead horses on other continents to see if riding dead horses can be off-shored.

[Matthew Sanford]

#23 is a standard business practice

[antonis]

Can't imagine what will happen in a thread dealing with bias current..

[duck_arse]

Can't imagine what will happen in a thread dealing with bias current..

current bias? very right leaning, I'd say.