Bridged chipamp for 16Ohm speakers

[blue eagle]

Problems like ground loops, instability and DC offset are some possible causes of abnormally high current. When one or more errors occur at the same time as inrush, the combination results in a blown fuse.

Do you have 1 ground wire, from power star ground, to each amplifier board, or do you have ground loop(s)?

1 wire to the star ground (1 star per channel)

Do you have 1 ground wire from power star ground to speaker jacks or is it convoluted?

The black speaker post goes straight to that channel's corresponding star ground

Did you install the RF filter cap, per the datasheet? That part also promotes stability, according to the datasheet. Stable amplifiers draw less current and run cooler.

Which capacitor is that? The amplifier circuit is the same as the most recent diagram I posted.

Did you put a multimeter to the speaker jack, set to read small amount of DC and confirm that output is less than 100mv DC offset? Zero would be good.

One reads 0 and the other 1mV, which is surprisingly good. When it was first completed I checked the DC offset before connecting any speakers and it read about 17mV for both, this seems to have disappeared. I also just unplugged one of the speakers and with nothing connected it still reads 0.

If you checked everything under the sun AND also have a cool running amplifier, then its possible that the fuse values might be optimistic, so you might check on that to see about correct fuses.

Yes, I thought I would get 3A time delays this time. And it runs cool, I have quite a large heatsink, it is 200mm x 100mm (with 25mm fins) and I've never felt it more than 'warm' to the touch.

And Michael, yeah they are definitely time delay fuses but when they blow it doesn't look like the filament has melted slowly, it is splattered around the edge of the glass container.

Also as for ground loops, I've used pretty thick gauge wiring for everything, I think it is equivalent to about 14AWG, so resistance and inductance throughout should be pretty low.

 

[AndrewT]

T3A seems a bit, no a lot high, for a 240Vac powered chipamp.
That will pass >700W continuously.
It will pass 6A for quite a few minutes, maybe even half an hour.
It will pass 30A for a second or two.
To be "splattered" all over the glass surface indicates it probably tried to pass >100A for a few milliseconds.

Fit a soft start.
Then you can use a close rated fuse. Try T1A for upto 250VA transformer.

 

[danielwritesbac]

Black Speaker Post >>> PG (power star ground)

The black speaker post goes straight to that channel's corresponding star ground

Normally, the black speaker posts go to Power star ground, which on your diagram, is located at the 0v output of the power supply board, where you tied PG+ with PG- (unified Power Star ground).

Which capacitor is that? The amplifier circuit is the same as the most recent diagram I posted.

LM3886 datasheet from National Semiconductor, the makers of the chip:
http://www.national.com/ds/LM/LM3886.pdf
Part Cc, a 220pF cap between input + and input - of. . . any non-unity stable op-amp. The job of that cap is to accidentally block RF while promoting increased stability on purpose.
That is an especially important omission to repair on most popular amplifier kits.

 

[blue eagle]

Yeah I think you are right Andrew, I need a soft start as much as I wish I didn't 😉 I'll get on that.

Daniel, the black speaker posts are then indeed correctly connected. And I suppose I can pick up a few 200pF caps to include when I over haul the amp. (And I'll probably get those 1500uF's and put everything in at once)