Check for oscillating amplifier

[regiregi22]

Hi mates!

I have read a lot about oscillating amps, what it seems to be the most common construction problem. I have been looking in the forum and on the net, and I couldn't find any place where is it explained the correct method to look for oscillation.
I have built some amps already, they were gainclones and headphone amps wich sounded quite nice and were not suspectfully hot. Despite I assume they are alright, I would like to check it anyway.

I have a 20Mhz scope. How should I test the amps? Should I just hook up the output to the scope and look for "something"? Wich pattern are we looking for? Any capture would be great...

And on the input side, may I short them or should I inject a test signal?

Any info will be gladly welcomed. As a novice, I have soooooooo much to learn from the "amps-masters"

Thanks in advance.

 

[Conrad Hoffman]

Aside from obvious problems where you just look at the output and say, "holy cow, it's oscillating!", I find that with sine wave drive there is often some particular amplitude and maybe load, where you'll see just a bit of widening or fuzz on the top and/or bottom of the trace. It's usually near full power, but doesn't have to be. IMO, amplifiers that do that invariably sound different than amplifiers that don't. It's often the values of the output R & L that will cause or fix this.

Conrad

 

[regiregi22]

Aside from obvious problems where you just look at the output and say, "holy cow, it's oscillating!", I find that with sine wave drive...............
Conrad

That's right where I am, in the problem of identifying obvious oscillation. What are we looking for in the scope? What are your methods of looking for oscillation?

Thanks for your quick response

 

[sumaudioguy]

technique

Abuse the circuit like this- use your oscillator to select a sine wave first and square wave second. Turn the scope down in a darkened room and adjust trace to a fine clear line. Start with like 10kHz. Look at the waveform such that one single cycle pretty much fills the screen. Now all a load and look for ringing on the flat part of the square wave or like the guy says a "Fat line" anywhere in your clean trace. Zoom in on that fat part if one is seen. Now add loads with capacitance starting at 100pf and have a look. Then do 300pf and repeat. Then 1000pf. A power amp should drive a .1µF and a headphone amp should drive a .01µF easily. One of these may make the wave quite strange.

 

[fredlf]

Hi all,
Coincidently, I have an Adcom GFA-555 that I believe is oscillating in one channel. I posted a picture of the trace today. You can pretty plainly see the "fat" clipping at the top of what should be a smooth sine wave.

I'll try and get a better pic next time I'm running tests.

cheers.

 

[regiregi22]

Hi all,
Coincidently, I have an Adcom GFA-555 that I believe is oscillating in one channel. I posted a picture of the trace today. You can pretty plainly see the "fat" clipping at the top of what should be a smooth sine wave.

I'll try and get a better pic next time I'm running tests.

cheers.

But what you show in the image seems to be a channel clipping. The amplifier cannot cope with the load and it clips the top of the signal.

I thought oscillation was a different thing.

 

[fredlf]

No, it's not just clipping. If you look at the picture of the trace full size you can see that it turns into a thick, "fat", fuzzy line. I admit it's a poor picture.

The amp had no load on it and was only running at half power (60vac).