How to test for oscillations?

[Jye]

Hello,

I have been making a 100W amp using the parallel circuit in the Overture Series High Power Solutions application notes and fired it up for the first time last night. It worked first time and everything is fine apart from some crackling that I am putting down to a bad RCA connection and the fact that it heats up to about 80 degrees C in about 30 seconds I have a feeling it could be due to some high frequency oscillations, since there was no audible noise.

So my question is how do I test for these? And I have access to a CRO and signal generator.

Cheers
Jye

[jackinnj]

one thing about the heat-sink you picture == they are one heck of a lot LESS efficient than one in which the fins are not "interrupted" -- and if you aren't using a fan the C/W rating is much less than you would suspect.

you can see the oscillation on an oscilloscope -- it helps if you have a scope with FFT -- then you can peg the frequency in a couple seconds -- if it is oscillating make sure that the lead lengths are as short as possible (tack the resistors directly to the chip pins), incorpoate the 0.7uH choke which National recommends, make sure that the filter caps on the power supply pins are bypassed with 100nF mylars. Bypass capacitors as seem to be shown in your photo (being several mm from the supply pins on the amp chip) are about as useful as teats on a bull.

[Jye]

Thanks Jackinnj,

I have placed some filter caps on the supply pins of the chip but the oscillations are still present. On the scope the oscillations are at about 150kHz and 5vp-p. One thing I did notice is if I turn the power supply down to about +/- 9v the oscillations disappear and all is well

I will try and get some chokes and try that out tomorrow. But in the mean time any other comments are much appreciated.

Cheers
Jye

[thespeakerguy]

I only see the bulk electrolytics. Any low ESR high frequency caps on the supply pins?

[Jye]

I have just placed some on the under side of the board

[sek]

Hi,

One thing I did notice is if I turn the power supply down to about +/- 9v the oscillations disappear and all is well

As of the National data sheet, the chip switches off (activates it's undervoltage-protection) at +-9V. This would mean that your oscillations are entirely independent of the supply voltage. Is the output (test/audio) signal still present when the oscillation stops at +-9V?

If not, you already have determined that the oscillation is affected by the protection circuit, which could narrow the suspect components/wires down when looking at the internal block diagram.

Have you tried improved input filtering (series 1kR, parallel 220pF) to rule out any EMI induced swinging? Next would be bypassing the feedback series resistor (e.g. 100pF in parallel upon it).

Hope this helps,
Sebastian.

[jackinnj]

are you using the 0.7uH choke? You can wind 15 turns of #20 enamelled wire around a 2 watt carbon resistor -- if you don't have magnet wire you can use the regular RS, PVC insulated wire, it just won't look pretty.

you could try placing a 220pF capacitor across the input pins -- but this is mainly to keep Radio Frequency Interference out of the chip.

[sek]

Erm, didn't I just write that? Except for the winding suggestion!
Edit: ah, we've been editing in parallel. So the suggestion must be very good.

[jackinnj]

we've been editing in parallel -- sometimes I leave the message open while I see how something is working. sorry to be duplicative.

[BlackDog]

Originally posted by jackinnj

you could try placing a 220pF capacitor across the input pins -- but this is mainly to keep Radio Frequency Interference out of the chip.

Not really related to this post, but I can confirm that the 220pF cap across the inputs works a treat if you live next to the regional radio transmitter, like me!

Steve