Loudbox mini

[lowell]

This amp came in sounding like a blown tweeter. After subbing tweeter and speaker too... I believe this is a circuit problem.

I got out the scope and found that with a 1k sine at 200mv RMS into the input, at a certain point there is serious crossover distortion at 10.3vAC RMS output (no load). This is with the drive very low around 2.5 and master volume at noon. Maybe this is because I have no dummy load.

Thoughts? Is it typical for a cold biased power amp to sound better at low volumes? Or vice versa?

I do believe the limiter FETs are working correctly as I don't see any clipping of peaks.

[J M Fahey]

Maybe this is because I have no dummy load.

Please try it with a proper load.
Worst case, use the speaker.

I got out the scope and found that with a 1k sine at 200mv RMS into the input, at a certain point there is serious crossover distortion at 10.3vAC RMS output

Not sure what you mean:
a) do you have "crossover distortion" (although you are not driving a load ŋŋ?? )at all levels, up to 10V RMS?
or
b) you have clean signal, which distorts at 10V and up?

In any case, besides answering post a scope screen picture showing the problem.

[Mick Bailey]

I find plenty of these with problem solder joints on the output transistor emitter resistors on the power amp board. First check I do with every one of these is to flip the board and inspect every joint. In most cases the resistor leg is loose and the solder looks like grey mud.

[lowell]

Thanks Mick will do.

JM. I tried it with a 4ohm resistive load. Same result: clean UP TO 10.3vac, at which crossover distortion begins.

[J M Fahey]

Ok, please show what this "crossover distortion" looks like

[Jazz P Bass]

Why are you using a '200mv' input signal?

I rarely go over 100mv's.

[lowell]

Let me clarify a bit more. The crappy crackling blown speaker sound is mostly on transients when picking...and I'm using an electric guitar...so I think the transients are probably less pronounce than on an acoustic.

I'm using 200mv because I believe transients far exceed that...so it's a decently sized sine wave, but still conservative in regards to transients. I also remember from some of my experiments 10yrs or so ago that generally single coils put out ABOUT 200mv and humbuckers 400mv. Maybe my memory is off.

I'll post some pics of the crossover distortion.

[lowell]

Here 'tis. First pic is at onset.

[lowell]

Hoping you all have a schematic. They sent me one but I don't think they want me to share it.

I checked the output transistor bias and all looks good.

Q12/13 Vbe
+/- 600mv

Voltage from Q12b to Q13b
+1.8v

[lowell]

More info:

When i first power on and inject signal it's a nice sine to a certain point. Then it clips hard. As I bring the input level down the clipping remains as the signal gets smaller on the scope. Makes me think "rail sticking." Thoughts?

Currently using U2 pin1 as scope point.

[J M Fahey]

Weird.
Iīd expect such "crossover at high power" waveforms in an amp *designed* to get them, such as a Peavey Transtube which has a circuit, between real preamp and power amp, *designed* to do that.
Donīt understand why that should be present in an acoustic guitar amp.
Important question: does power amp do that? (even if fed a clean sinewave) or it already gets that from preamp and it just reamplifies it faithfully??
Another possibility: by any chance itīs an LM3886 power ap? ...or maybe a TDA7294?
Because such chipamps include a so called "SPIKE protection" which clamps output to protect them and sometimes makes similar waveforms.

[lowell]

I'll inject DI into the PA and report back.

It's discrete not a chipamp.

In my last post I stated that U2 pin1 hard clips at a certain point. And then continues to clip as i bring the signal down. The clipped signal gets smaller but remains clipped. U2 is a preamp op amp. The tone stack recovery stage. TL074.

[DrGonz78]

So what did the solder joints look like on the main power amp module board? The one I worked on recently the emitter resistors R71 & R72 both had bad solder joints, one even ripped off the trace pad. The solder pads on those circuit boards just fall apart like cheap thin foil. Not only were the solder joints horrible on both ends of the emitter resistors but also the solder joints on the emitter legs were bad too. In fact there were half a dozen joints that needed to be re-flowed on that power board. Here is the schematic so others can follow along.

[lowell]

Havent disassembled yet. Wanted to narrow my search before blindly soldering stuff.

I am not having the issue if signal enters U2 pin 1 or 2. Which is the last opamp stage before the diff pair.

I then injected signal at input again, got the problem to happen. Upon pushing the phase switch it went away. Now I cannot get the problem to happen. Naturally I'm worried I'll give it back and it'll happen right when they get home.

I think ill do some resoldering on the phase switch. But not too confident that will be the ultimate fix.

Currently able to get 14VAC CLEAN output at 8ohms with 1khz. 25 watts. Then it clips but seemingly in a more normal hitting the rails kinda fashion.

[J M Fahey]

1) what is the clean output into 4 ohms, the proper rating?

2) so it does not happen any more? and the phase switch affected it?
Then itīs not a powr amp problem, which too boot is quite conventional.

I suspect, in this order:

1) you overloaded the input and got that weird waveform.
The problem was not in the 10 something output voltage but in the signal feeding the amp.

2) that Fet limiter is crude to say the least, more of a gate than a linear volume regulator, and it might create weird waveforms.

3) I was going to suggest you inject signal straight at the power amp input, on top of C31 and with the link to "limiter" disconnected and viceversa, rise input level and gain until you get the "crossover" and then scope power amp input , again on top of C31, but of course fed through the preamp, to check the power amp does the dirty deed (I doubt it) or it just reamplifies what itīs fed (what I suspect) but guess now itīs too late.

Just deoxit that phase switch well and resolder its legs.