Help with detecting/reducing ocilation on a gainclone

[PaulHilgeman]

Hi,

Right now I am working on a gainclone amplifier. I have a 100K input potentiometer, the gain resistors and the filter caps all wired up. I heard that the best way to test for oscilation was to leave the input open, and put a resistor (1-10 ohms) across the output, and see if the amp gets warm. Sure enough, after about 5 minutes, it is too warm to hold for more than 10 seconds (the chip and the heatsink).

I can take it into the lab on friday to look for the oscilation, so I am wondering what types of tests I should run on it. I have access to a funcion generator, and 100MHz scopes.

Please outline a few tests that I could do, and what I can do to fix the oscilations that I probably already have.

I do have a 1 ohm resistor and a 1uF capacitor that I could use for a zobel across the output, but I hooked it up and I was still getting oscilations.

Thank you so much,
Paul Hilgeman

[Equilibrium]

Quoting Master Kuei Yang Wang:


The circuit I originally proposed relies on the Volume control AND the input to ensure stability. This is not the most ideal solution as some have discovered with regards to reliability.

Hence I would now suggest to have a resistor of 10k Value on the non-inverting input of the Amp, no capacitor and a 22k Resistor to ground on inverting input. This way the Amplifier will not oscillate even with no volume control and no source connected.

and

The KEY issue is with the input UNCONNECTED. In that case the amplifier will be instable.

"To "cure" the "open input" issue I suggest you place a 22k resistor to ground on the INVERTING INPUT. This way, no matter what external impedances and connections are present the Amp always operates with a noisegain of 11 and will be stable. Then simply make the resistance/impedance on the non-inverting input match, 10k - 15k should be close enough.