PS for LM3886 amp [diagram?]

[petemoore]

  I have an LM3886 amp, which works.
  The PS is from a CT transformer, DB/caps...prividing about +/- 18VDc.
  I read it should have regulated +/- supplies.
  I am looking for confirmation that this supply will improve performance...amp works, but when cranked a bit exhibits a bit of strangeness...like the power supply is 'moving around'...
  And for a diagram/schematic which shows the wiring assignments and applicable regulator chips for each side [ _ / + ] of the supply.
 

[R.G.]

Don't believe everything you read. An LM3886 works fine with an unregulated supply. There are situations which call for a regulated supply, all right, but not in general.

Ideal for the LM3886 in terms of most output watts and no heating problems from the chip is +/-31Vdc, made by a 44Vct transformer.

The LM3886 has a low voltage shutdown inside it, so if your +/-18V is sagging a lot, that may cause you trouble.

See if you can get a 40Vct to 45Vct power transformer. I think Parts Express has a suitable one for about $25.

[MKB]

IIRC the LM3886 does NOT like overvoltage on the supply rails.  They go boom very fast when that happens.  So be sure you are operating well in the safety range if you do use an unregulated supply.

We used the LM3886 in a product at work that operated the amp with an unregulated single supply.  The part was basically indestructable and very reliable and consistent.  They do start freaking out in a very strange way when the thermal shutoff starts kicking in, part of the waveform gets chopped in an unusual way.  So you might make sure your impedances aren't too low and you are getting the heat out of the part correctly.\

I used to run these things at 90W into 2 ohms on the bench, they would just take it and smile at you.  Obviously that is way out of their proper operation, but did show how rugged the parts were.

[petemoore]

The LM3886 has a low voltage shutdown inside it, so if your +/-18V is sagging a lot, that may cause you trouble.
  Sounds about right, or, is probably what's going on here...a bit of instability when pressed, it doesn't 'want to go past a certain amount of output.
  so..amp works fine 'until'..and will continue to do so.
  still 'scouring the bottom' for more iron...lol.

[soulsonic]

I highly recommend this National Application Note: http://www.national.com/an/AN/AN-1192.pdf
It's been extremely helpful to me.

[jrc4558]

I'm running mine at 41-0-41 unregulated and so far had no problems. I didn't have it on longer that 20 minutes though, so no idea bout long-term stability...

[R.G.]

The chip will withstand higher voltages, but it will go into thermal shutdown more easily with higher voltages because the chip just has to eat the extra voltage times the load current. And if you go over the rated voltage (94V? +/-47V?) the chip becomes a smoke emitting light display. Big heat sinks or short run times make it live with high voltages, but that's not really the best way to run it.