Op amp output loading, pot values and more

[Eddododo]

I'm trying to solidify my understanding of a couple things regarding the output load of op amps etc.

I 'know' we have to watch for the output sourcing current specs of an op amp.. but is this just a DC characteristic? If we use a coupling cap before say a volume pot on the output, does that make the Output sourcing current irrelevant?

And then as far as the AC load on the output.. I'm not sure how to determine what is an appropriate parameter, outside of gleaning rules of thumb from other designs.. is it simply a matter of output impedance, and staving off high frequency loss?

And then this specific scenario from a circuit I'm working with.. the dotted ovals are 20kb pots for the EQ


Looking at the output of U38.. with the R596/r597 potentiometer as shown, the load U38 is seeing is just the 2k2+5k6+10k into the virt ground of the next - input..? But that's the DC current consideration, right? How do I determine If I'm loading U38 appropriately if I were to, say, make R604 1k, or a short even? I know that 10k is a fairly low output volume pot to see, for example, so I'm assuming less than 10k to 'ground' on the output of u38 becomes a chore, but *when,* or better yet, *how*?

(Also I know there is some offset to deal with between the stages.. I didn't exclude the Cap for a reason, I just hadn't put one and the offset hadn't been a problem as far as i can tell.. I plan to add one but please share anything that makes that relevant in the context of this topic! It seems that sourcing current is usually a non-issue with the Op amps I've looked at, but if it were somehow, is a coupling cap all you would need to make that problem go away?)

I hope this makes enough sense to actually form a question to be discussed..

[ElectricDruid]

I 'know' we have to watch for the output sourcing current specs of an op amp.. but is this just a DC characteristic? If we use a coupling cap before say a volume pot on the output, does that make the Output sourcing current irrelevant?

No, it doesn't. You can pass currents with AC too - otherwise the power grid wouldn't work!

Adding the cap to the volume pot simply turns what was a simple resistance into an impedance. E.g. now it has a frequency response that we need to consider, rather than just a resistance which is true at all frequencies.

In the case of the coupling cap with the volume pot to ground, if you think about the connection at the top of the volume pot, you've got a highpass filter (C, R to ground). This happens because at low frequencies, the cap acts like a *huge* resistor and the voltage divider is therefore very "top heavy", so very little output. As the frequency goes up, that capacitor's effective resistance drops and the voltage divider becomes more "bottom heavy" and the output increases.

Looking at the output of U38.. with the R596/r597 potentiometer as shown, the load U38 is seeing is just the 2k2+5k6+10k into the virt ground of the next - input..? But that's the DC current consideration, right?

Yes, if you ignore the cap C327, then that's roughly true. The whole lot is actually in parallel with R544 so it's a simplification, but it's a fair one. I think the situation is actually even more complicated than that in reality, but we don't need that level of detail.

How do I determine If I'm loading U38 appropriately if I were to, say, make R604 1k, or a short even? I know that 10k is a fairly low output volume pot to see, for example, so I'm assuming less than 10k to 'ground' on the output of u38 becomes a chore, but *when,* or better yet, *how*?

I'd do it the way you have - a basic "DC to daylight" analysis. At DC, all caps are open circuits, so what does it look like then? At "daylight" (some amazingly high frequency) all caps are shorts, so what does the network look like then?
10K might be fairly low for a volume pot, but it's still a reasonable load for most op-amps. While I might start to be concerned if the load impedance goes much below 10K, I wouldn't think it was critical until we're down to a few K or less. Obviously it depends on what op-amps you like to use. Old analogue mixer circuits often use much lower resistor values than we're used to to kee noise down, and then you see heavy use of op-amps like the NE5532 which can cope with that.

(Also I know there is some offset to deal with between the stages.. I didn't exclude the Cap for a reason, I just hadn't put one and the offset hadn't been a problem as far as i can tell.. I plan to add one but please share anything that makes that relevant in the context of this topic! It seems that sourcing current is usually a non-issue with the Op amps I've looked at, but if it were somehow, is a coupling cap all you would need to make that problem go away?)

No, as stated, adding the cap doesn't make the problem go away. It just turns it into an AC analysis problem instead!
What the coupling cap *will* do is block DC ("At DC, it's an open circuit", remember) so if your output is riding on a 4.5V bias and your volume pot has one end grounded, it'll block any current that would have flowed because of the bias voltage. Which is usually why they're there, and why they're often known as "DC blockers" as well as "AC coupling caps".

HTH

[PRR]

What is the difference AC or DC?

It would take infinitely long to show a signal is "pure DC". And if it changes in that time, it isn't DC.

OpAmps usually go by what we used to call "instantaneous" conditions, too quick to show on mechanical meter (if you remember those) but maybe not-caring about infinitely fast either.

"All" the utility opamp chips will pull 2K load nearly to +/-15V rails. On single 9V power the worst-case current will be a third or a quarter and sub-1K loads may not make trouble.

Trial-and-error is safe. I managed to kill 709 chips with short-term shorts but nearly all newer chips will do as best they can and then sit there getting hot.