Gibson G35 ss Amp

[lowell]

I believe the driver IC MFC8022A is shot here. Anyone have one of these they wanna sell me? Or know who might? Only other option I see is redesigning the PA with LM3886 and doubt the customer will want to pay for that.

http://www.gibson.com/Files/schematics/G35(REV%20A).pdf
http://www.gibson.com/Files/schematics/G35(REV%20B).pdf

[52 Bill]

Only other option I see is redesigning the PA with LM3886 and doubt the customer will want to pay for that.

Can the LM3886 chip run off a single sided power supply?

[g1]

Hard to tell if this place has them or not: ENGINE ROOM

Someone else was having the same problem with that model amp. It didn't get solved, but this may give somebody an idea: Audio Asylum Thread Printer

[Enzo]

How about get rid of it and the output transistors and stick a TDA2040 in there?

[lowell]

Good call Enzo... that looks like it'd be implemented quite easily! Just a matter of removing some components and then running the 5 wires to the right spots. Easy as pie.

52Bill I'd think you could just make a simple voltage divider to get the input to V+/2 and put a series DC blocking cap before that in the circuit. Someone please correct me if I'm wrong.

[Enzo]

The input already comes through C11, so no worries there about DC. And C17 does the same for the output to speaker.

Then look at R15,16,17. R17 is 1 meg and r15 and 16 are in series and add up to...1 meg.

Looks like it is already there waiting for you.

[lowell]

Yah indeed!

[lowell]

Ok so I got it in there and the amp works. Used a TDA2050 cause had it in stock. However, it is distorted at all volumes. Not terribly distorted, but not clean either. There is a "bump" on my scope mid-wave that looks like crossover distortion. Signal is good at TDA input. I replaced the output coupling cap to no avail. I also replaced the 2n2484 cause it was noisy, with a 2n3904. I tried new filter caps too but it didn't seem to help.

[lowell]

Dang... well now the thing just sits there and oscillates!! Even with all knobs at zero. And the zobel 22ohm resistor is smokiing. Shoot guess I need to see what's up here. I should mention that I put the output cap after the zobel. Is that a bad idea?

[Enzo]

Look at the 2050 data sheet sample circuit, and make sure you didn't leave out any of the little circuits around it.

[J M Fahey]

Agree.
I was chiming in to suggest TDA2050, but I see Enzo already did that.
It will work, but is choosy about decoupling for stability.
1) Zobel must be changed to 2r2 + .47uF , "stronger" than original 22r+.22
2) add a good .1uF ceramic from TDA pin 5 (+B) to pin 3 (gnd) ; at the chip itself if possible
3) be tidy and neat with wires from PCB to TDA, which I guess is mounted offboard, straight on heat sink (in place of one of Q4/Q5?)
4)it should work very well, simply it's slightly unstable now.
Scope it with a speaker load and check for fuzzies, don't think it's fully oscillating or it would overheat big time.
Speaker loads are nasty.
If nothing else does, try to lift C15 (50pF) on the power amp feedback or raise it to 100/200 pF
Compensation for one IC not necessarily fits a different one.

[Jazz P Bass]

I should mention that I put the output cap after the zobel. Is that a bad idea?

Which output cap are you refering to?
C17?
It is not required for the TDA chip.

[lowell]

JPB in this circuit it IS required because using a single supply rail. There is DC on the output of the chip.

JM, when the thing was working but was distorted there were definitely "fuzzies" on the scoped output. I'll try those component adjustments. Of note is that if I put a .1uf cap from pin 2 to ground, the oscillation stops.

[lowell]

Put the 2r2/.47uf zobel in there and reassembled and the thing is banging!! Works great, better than before.

I also moved the zobel to AFTER the output cap. I get a CLEAN 6.5vacRMS 1khz at 4ohms resistive load. Input is ~250mvRMS so maybe I'm clipping a preamp stage. That's (6.5^2)/4 ~11watts clean, then it breaks up.

[Jazz P Bass]

[QUOTE=lowell;281558]JPB in this circuit it IS required because using a single supply rail. There is DC on the output of the chip.QUOTE]


Oops.
Forgot about the single rail