problem with lm3886 parallel

[laki101]

so made my first ever amp based on this PA100 DIY 2x LM3886 in parallel gainclone audio amplifier

YouTube - VIDEO0016

this what i get so it plays, the bad sound is because of the crappy speaker but the problem is that the speaker is fully pushed in (from the voltage) so like when you conect a battery to a speaker and then somethimes it goes all the way up like you change polarity

and the part where the speaker goes crazy is when i disconect the signal, so it has no input signal but the speaker pumps

now the only thing that i can think of is the problem that the power isnt regulated so that makes the problem but not sure is it that

 

[Aucosticraft]

Have you checked them in individual amplifiers? ( remove them connected in parallel at output.)

Have you procured components as recommended by the application notes? Cause leakage in capacitor can lead to offset voltage.

One more thing while you are testing each amp. isolate input too.

 

[laki101]

so disconect the input and output and test one at a time?

 

[Aucosticraft]

yes......

 

[laki101]

did it one chip has 0,5v and the other 35v or the suplly so i gues that one is toasted

 

[Aucosticraft]

See you got answer ...

 

[laki101]

changed that one and now it works when it has a signal but has u buzz and after a few minutes the one chip that i changed overheats and it starts to "pump" the speaker
and also when there is no signal it pumps it so have no idea what could case that

 

[Aucosticraft]

This is how the chip must have got damaged previously. Well don't load it with speakers this moment. use some dummy load. check the output of both the chips pin to pin for analysis. do not combine their output unless they both are same. Parallel connection of amps need close matching between those two amp.what component have you selected? are resistors tolerance as mentioned in the application notes? capacitors are of the type as mentioned in the application note?

 

[laki101]

i was thinking can I make this using just one chip per channel? so use the same components but just one time? so when you look the circuit you see 2 same parts around the lm3886 so just use one and leave the rest the same?

 

[Aucosticraft]

tell me when you remove them from parallel mode. and measure voltages then what are those? also if you connect speaker at the output of each does it work ok ?
if yes, then sure that two amps are not matched correctly. check each component value and its tolerance closely. feedback resi need to be 0.1% tol and others should be 1% follow the AN-1192.

 

[laki101]

one chip works when speaker connected directly, the other one not
one has abou 5-6v when there is a signal and 28 when there is no signal connected and thats the one that works
the other has a constant 0,5v

and you did not say will it work these components on a single chip not a parallel configuration

 

[Aucosticraft]

Measure outputs in this conditions.

1) ground Input
2) Input not grounded but no signals at input.
3) voltage at pin10 of LM3886 when input is left floating.

I asked if you verified components again. cause this looks like there might be some wrongly component placed by human error.

well also tell me if you have web cam we will do this online.

 

[AndrewT]

one chip works when speaker connected directly, the other one not
one has abou 5-6v when there is a signal and 28 when there is no signal connected and thats the one that works
the other has a constant 0,5v

and you did not say will it work these components on a single chip not a parallel configuration

Hi,
is that a second chip amp damaged?
Did the same one survive both tests?

A single chip amp is much simpler to build.
Get a pair working with all the optional components in place and both as AC coupled.
Then start experimenting with these to hear what some of the suggested modifications sound like.
Save paralleling and bridging for your next life.

 

[Aucosticraft]

Hi,
is that a second chip amp damaged?
Did the same one survive both tests?

Dear Andrew,

I guess the chip is not damaged yet. The reason why I feel so is,when there is no input, output is 5 to 6 v. when he gives input, the output voltage goes to 28v so I feel its problem of excess gain. that it amplifying stray inputs signals when no input is provided. also he says its 5 to 6 V that means its rapidly swinging, another sign of stray pickups amplified. the reason why i asked him to ground input and measure the output voltage. Whats say?

Regs.