Strange lm3886 distortion problem

[Oliphant]

Hi gus I'm a long time reader of this forum and I finally have a reason to sign up and post something.

I've built an LM3886 circuit and over time (5-10mins) the signal slowly swings to the negative rail.

There are four circuits and they all do exactly the same thing, so I'm certain that it's a fault in the design and not with the chips themselves.

Attached are pics of the circuit, layout and signal.

Any help would be fantastic and thanks in advance.

[sangram]

One resistor is missing - from the +input to ground, it is probably what is causing the issue.

[Oliphant]

Thanks Sangram I'll give that a try. What value would you recommend?

[Oliphant]

Sweet, that worked perfectly. Thanks again

[sangram]

For some strange reason I wasn't notified of replies...

Anything between 10 and 47K is good, the closest you can get to Rf (20K in your case) the better your chances of low output offset.

Good to see you got it sorted out

[danielwritesbac]

Since you have four circuits, that is sufficient to create a parallel amplifier. See the PA100 diagram in National Semiconductor's AN1192.PDF document.

On your schematic, input load is missing, and input load isn't supposed to be decoupled. For success on that, just imagine that you had C3 located upon the RCA jack of an amplifier enclosure instead of upon the circuit board.

On a parallel amplifier, you will want a unified load and a single C3--the exact same signal goes into two chip amps else the difference comes out as heat. One doesn't want to turn desirable signal into heat.

Conversely, with the LM3886, harmonic distortion increases as frequency rises and does occur within the audio band. Fortunately the distortion differs slightly between chips and a parallel amplifier will turn this noise into heat. Its great to turn undesirable signal into heat.

Temperature reduction can be done be slightly changing the resistor value of "Rout" that is mentioned in the PA100 schematic of AN1192.pdf.

http://www.national.com/an/AN/AN-1192.pdf
See page 8.

[jasonmbeamer]

Ok I had several setups working great. yes they were a bit unstable (hahaha) But I started having this same problem as Oliphant after i designed a new board and my chips started burning up. Could this resistor I left out of my new design cause my chips to burn up ??? Granted I was running at +50v and -50v (open voltage) but I very rarly burned up my chips with my first design.

The thing is I downed the voltage to +45v -45v (open voltage) and I am really missing that kick *** thump these amps use to make

[Oliphant]

Originally Posted by danielwritesbac

Since you have four circuits, that is sufficient to create a parallel amplifier. See the PA100 diagram in National Semiconductor's AN1192.PDF document.

http://www.national.com/an/AN/AN-1192.pdf
See page 8.

Thanks danielwritesbac, I'm running these amps as part of an active crossover system so I won't be able to try out the parallel amp - it does look like something I might like to try in the future though. This is my first experience using the lm3886; is the slightly harsh top end normal for this kind of amp? Even when using neutral sounding speakers, listener fatigue sets in pretty quickly. Is there any way to reduce it?

[Oliphant]

Originally Posted by jasonmbeamer

Ok I had several setups working great. yes they were a bit unstable (hahaha) But I started having this same problem as Oliphant after i designed a new board and my chips started burning up. Could this resistor I left out of my new design cause my chips to burn up ??? Granted I was running at +50v and -50v (open voltage) but I very rarly burned up my chips with my first design.

The thing is I downed the voltage to +45v -45v (open voltage) and I am really missing that kick *** thump these amps use to make

+/-50v is 6 volts above the absolute maximum supply voltage (datasheet page 3) so I'd say it's safe to assume that, that was destroying your amps.

The resistor that Sangram suggested adding to the circuit that I had totally fixed the distortion problem that I had. I'd add it to your circuit if you're having a similar distortion problem. For more power try running them in parallel as danielwritesbac suggests - or just bi-amp your speakers if you're able to. Doing either of these things should allow the amp to grip the woofer better and deliver a lil more punch.