pot scratching with high gain amp

[sarv]

I won't be able to measure the DC voltage at the input terminals for a couple days (away from home atm), but will confirm that when I get a chance.

Otherwise, I've gone ahead and ordered a conductive plastic pot to test, we'll see if that makes any difference.

[Helmholtz]

I don't think it's real. Connect a 1M scope to the inverting input and it will show half the supply voltage (4.5V) because the opamp will drive its output to ensure there is no voltage between its inputs and the non inverting input is at half supply voltage. Connect the scope to the non inverting input and it loads it down to 2.25V (because of R1) which will cause the inverting input and output to be at 2.25V also.

Yes, it can't be real in a working circuit. The loading with 1M is a perfect explanation of the discrepancy, good point. My suggestion was to recheck with a DMM directly between the points of interest - in case of doubt.

[pdf64]

At 1k, R4 might not be sufficient to prevent the capacitive load of the cable causing instability; try 4k7.

[sarv]

At 1k, R4 might not be sufficient to prevent the capacitive load of the cable causing instability; try 4k7.

that did it! scratching gone!! I don't think I would have ever thought of that - thanks so much

[pdf64]

Great, pleased that worked out for you, now that's proved the root cause, you may be able to bring R4 down to 2k2. I had to work that issue out for myself when I started tinkering with this stuff (40 years ago!), as it's not something that seems to get covered often in circuit design texts. The learning point is that opamps, especially older designs, are far from ideal. My suspicion is that the open loop output impedance of these lowish current FET input designs is significant, perhaps being operated at a low supply voltage doesn't help. Other opamp types may be fine with 1k or less, eg perhaps 5532.

[sarv]

So although increasing R4 was the fix, I am still seeing some dependence on pot value and/or type. Either way, I've got a working solution now but might as well try to see this through all the way.

What I'm seeing is that R4 4.7k with a 100k conductive plastic pot is fine (no, or at least barely perceptible, scratching). 2.2k wasn't enough, scratching came back. I then bumped it back up to 4.7k and tried switching back in my 150k carbon pot and scratching came back. Now I understand that this isn't an apples-to-apples comparison, because as I had mentioned in a previous post I had already noted that there was a relationship between how bad the scratching is and the pot value (with higher values resulting in more scratching) - so I can't say just from that test whether the increase from 100k to 150k or the change from plastic to carbon, or both, were a factor. I plan to do more testing:
* will try a 100k carbon pot with 4.7k
* will try increasing R4 further with higher value pots (to a certain point - won't want the output impedance to be too high. might just try 10k)

Anyway, was just wondering if anyone has some insights on why the pot value would have an effect on the scratching, especially since the scratching is as it's being shorted out, not as it's reaching its max.

One final thing potentially of interest, I'm using a reverse-log pot (not sure if that matters at all).

[pdf64]

It's a margin of stability thing, set by the opamps gain/bandwidth/phase characteristics and the gain/bandwidth/phase of the circuit it's being used in (both the simple schematic design and its actual implemented circuit, that includes all parasitic/non-designed effects); it's almost certainly unrelated to the pot material/construction, just its actual set value. Though of course the parasitic characteristics of the pot construction (eg metal shell connected to circuit common) and the lead dress of the connections to it may differ with different pot types, and so may have some effect on the margin of stability.

In a nutshell, as the circuit gain increases, stability generally becomes more of an issue, especially with non inverting circuits. But paradoxically, very low gains (ie lots of NFB) can also cause stability issues.

Kids these days have got everything too easy, with their compensated opamps and all.