LM3886 shutdown

[Robin755]

hello,

I've build a LM3886 amplifier recently.
When I start to play it works perfectly but after a few minutes the LM3886 turns of. After letting it rest for a minute, it works perfectly again, but only for a couple of minutes like before.
First I thought it was the powersink that was to small, so I placed a bigger one. Still, the same problem.

Then I read on a forum somewhere to measure on the output of the amp with a DC-voltmeter.
I measured while starting to play on the LM3886 and noticed that the voltage started around 1Volt to end up to 30Volt when it shut down.

Does this mean my amp is oscillating and what can I do about it?
Could it be because I haven't placed the inductor parallel with a resistor at the output?



Thanks, Robin

[teemuk]

You need to reference that non-inverting input to ground. Without that reference the input will float in any potential it sees the best. This causes the huge drift in DC offset. It's a common mistake people make with these chips. A resistor from non-inverting input to ground will cure it.

Since R2 is 22k (and feedback path through R3 is AC coupled), use a 22k resistor from non-inverting input to ground as well. Why? This will make both inputs see the same DC resistance to ground. + input sees the 22k and - input sees the 22k + speaker's DC resistance, which is low enough to be negligible. This furtherminimises DC offset.

If feedback path would not be AC coupled the input resistor would have to be equal to R3. Makes sense?

[Robin755]

I understand, and thanks for the quick responce.

[Robin755]

Well, I've added the resistor and it works perfectly now.

Thanks a bunch.