Marshall MG250DFX - Smoke Generator!

[DetroitDan]

This thing had 2 burned TDA7293s.

All pre-amp circuits are working properly.

All values on the power amp plug pins seem correct.

When I plug in the power amp module (with a new TDA7293) the speaker (for that side) emits a loud tone (like feedback), and the chip get very hot, very fast.

I think there is some sort of runaway loop problem related to the feedback circuitry, but I can't seem to pin it down.

I will appreciate any ideas; thanks!




Dan

[52 Bill]

When I plug in the power amp module (with a new TDA7293) the speaker (for that side) emits a loud tone (like feedback), and the chip get very hot, very fast.

Don't try hot swapping these modules, as you will kill the new chips.

Are you reading any DC on the output with no speaker?

[Enzo]

ANy chance you have swapped the cables to the modules? That is, so Channel A is driving module B and vice versa?

AND GET RID OF THE SPEAKER. Don;t hook up a load until you are sure the amp is stable.

[52 Bill]

AND GET RID OF THE SPEAKER. Don;t hook up a load until you are sure the amp is stable.

And when it is stable, shut it off before plugging in the speaker to test the amp.

[DetroitDan]

Thank you for the suggestions.

The chip(s) get hot, immediately, even without a speaker. There is no time to measure DC out - The chip will burn

No hot-swapping. I plug the module(s) into the main board; hit the power; and instant heat. Every permutation has been tried, one, both, A-only, b-only; output jacks disconnected


I've tested each of the pins to the module, with a DVOM, and scope, without the PA module connected.

Nice 1vp-p signal; +36;-36v; grounds seem good(<0.1ohm); mute rises from low to 6.45v; Output not grounded (I think it was about 4.5k ohms w/no speaker, after a cap charge rise from 2.5k).

Crazy.

I picked this up with both power amps burned; so, I don't know the history. None of the main board circuitry seems to have been disturbed.

Dan

[DetroitDan]

I just read all of the threads I could find regarding the TDA7293; yikes!

I'll hook up my variac and bulb limiter, and try a little slower path.

I generally only work on tube amps, but lately, have ended up with some burnt Line6's and SS Marshalls.

[JoKeRz]

I just read all of the threads I could find regarding the TDA7293; yikes!

I'll hook up my variac and bulb limiter, and try a little slower path.

I generally only work on tube amps, but lately, have ended up with some burnt Line6's and SS Marshalls.

Hi DetroitDan - I know this thread is quite old, but I have the same amp model with same exact problem, exact behavior and readings. Did you manage to fix your amp? I have blown 4 TDAs already and no solution to the problem. Any further feedback from your troubleshooting will be greatly appreciated.

Thanks,

JoKeRz

[booj]

mg250dfx.pdf

[JoKeRz]

Thanks booj. I already had the schematics and used it to follow the signal throughout the circuit. Pre-amps are ok and I get sound on the headphones jack. Just wonder if DetroitDan was able to fix his amp and how. I've already unsoldered several components to check if they are good and so far no culprit. Thanks!

[TomCarlos]

I just started working on a MG250DFX for a friend. It was making a loud screech at turn on. So I swapped the speaker cables and was able to isolate the problem to one side. After a little soldering on the power chip leads, I tried to fire the amp up again. It seems to be working.

I cannot believe the SMALL heatsink for these two power ICs. The other thing I am not sure of (cause someone else worked on this amp before me) .... The fan... should the air flow be flowing into the amp or out of the amp? I think the fan was in backwards!!

Tom

[JoKeRz]

TomCarlos: the fan should be blowing the hot air out of the case, like an exhaust fan. I can confirm that was the configuration of mine when I first opened it.

In my case, something is clearly not correct in the main board / components. I can tell that because I have already meticulously checked all components and contacts in the little power amp board several times before and after attaching a brand new tda7293 (I ended up soldering an IC socket to make it easy to swap the 7293s) and still I blow up the TDA a couple of seconds after I turn it on (with smoke if I leave the main power switch on further). The power amp board and heatsink are not touching the chassis nor any other part (so no chance of short circuiting the +V and GND) and at this point I really cannot pin point the culprit, but haven't given up.

Anyone with similar problem on MG250s, MG100s MG50s, AVT150s, etc (Marshall SS based on TDA7293), please share your successful experience fixing this. Thanks!

[Enzo]

Hi, welcome.

So stop burning stuff up. Look up "light bulb limiter." Simple thing, make one and use it.

Whenever one 7293 blows up, I always replace BOTH of them.

Operate WITHOUT a speaker load until the amp is stable and not making DC on the output.

Disconnect the power modules. verify proper power supply voltages at the connector cable.

NEVER connect the cable to the power module without verifying the power supplies are completely discharged.

The little parts on the power module boards rarely fail, but problems elsewhere in the amp can cause problems with the 7293s.

[Jazz P Bass]

NEVER connect the cable to the power module without verifying the power supplies are completely discharged.

This is THE most important item.
You cannot connect the TDA ic with power 'floating' on the rails.
They WILL fail.

[Enzo]

For that matter, do not DISconnect themk with any power in the system.

[JoKeRz]

Hi Enzo and Jazz P Bass – I had read a lot (really a lot) of previous issues concerning the power module amps with TDA7293 and your great feedback in those threads. I am following all the tips you have provided in those and have a fairly good understanding of electronic circuits so let me give you a bit of more details of my situation and what I tried so far:

- Both power amps were dead once I god this unit, as well as the 3.15A fuse blown and the standby circuit zener (6.8V, called D19 in the MG250GFX schematics) opened.

- First attempt was to just replace the fuse, and that just made another fuse go away. Since I did not have an incandescent bulb at that time (just fluorescent ones),no go for a light bulb limiter and I could not wait until next day so I decided to just put a jump cable in the place of the fuse and do quicks turn-on/off in the main switch. I could notice the traditional silicon burning smell after a quick on/off and if I left the switch on for a few more seconds, smoke on both power amp modules.
(that is when I started researching on forums like this)

- I ordered several TDA7293 so I could play (and burn a few on the way). I ordered them from an ebay seller from which I had ordered parts before (seller located in the US) and latter on I read in some websites this IC can be fake. I really did not believe that, but just to make sure I used my multimeter to read resistance between some pins and it seem to match the internal schematics of it

- So I rereplaced the 6.8v zener (first removed it, and checked if there was still a short between transistor pins / capacitor, and there was none), checked all capacitors and connectivity of the power module board and re-soldered a brand new tda7293. I made sure to drain power from the big electrolytic capacitors before reconnecting the board. Once unit was turned on (with no speaker out) smoke again. Very quickly. Funny thing is that the heat sink was not even warm so the IC blew without generating heat. Oh, forgot to mention: the chassis of the 7293 (which has -VCC) is not insulated from the heat sink so I made sure not to let it touch any part of the MG250DFX chassis (isolated by a piece of 2x4 pine wood )

- Next I decided to solder IC sockets so it would be easy to remove and add another tda7293 as necessary (I am afraid to lose the PCB tracks after so many re-soldering). After I did that, I attached another new tda7293, confirmed all connections to the terminals of the power amp module cable (all 15 pins) and then turned the unit on (discharged capacitors before) and, again, tda7293 blown followed by zener blown.

- I could not find any other problems with the main board. I am testing with just one of the modules (which should be ok since this is a stereo unit so each power module is independent for L/R. If you see a problem with that approach, please let me know).

- I have also checked contacts in all 1/4in jacks, including plugging a cable from fx out to fx in (I read in same threads those could be causing some short circuits)

- Also forgot to mention: headphones output works like a charm, with both clean and overdrive channels working as well as effect processor unit and reverb. So the pre-amp / effects part of the circuit seems to be ok. I even connected the power amp board WITHOUT TDA7293, and made a direct connection from IC sockets 2 to 14 which pretty much connects the input directly to the output (without power amplification). Doing that and plugging a speaker in the speaker out jack I could even hear the input signal (very low volume, of course) and the main pot also worked controlling the overall volume. That is why I am a bit skeptical of a major problem in the main board. But, still, cannot pinpoint the where the problem is.

- For next troubleshooting steps I have two approaches:

1) I ordered a new power amp board, pre-assembled, from ebay, using LM3886. Once I received it I will test it isolated and then try to connect it replacing the original TDA7293 from marshall. I just read in many places that the LM3886s are less sensible and should not blow that much

2) in parallel, I am trying to use the original marshall power amp module with a new TDA7293 by just connecting +vcc/-vcc/gnd from the main board and feeding my own input signal (already pre-amplified) and checking the output with multimeter and then with an old speaker (yes, do not want to risk blow a good speaker). The goal here is to confirm the problem is not in the power amp board itself (neither a “fake” 7293) to then focus on the main board components. Then I can plug input from main board to see it still works so I can pin point if the problem is in the circuit before the power amp or after (where the main output pot is).

Well, that is pretty much a summary of my problem and attempts. So far did not find the culprit. I will make sure I use the bulb limiter going forward. Unfortunately I only have one digital multimeter as a tool (no scope, not even a variac to play with input voltage). Please let me know if you guys can suggest any other troubleshooting step. Tonight I will perform another inspection on the main board.

Many thanks,
JoKeRz
(sorry for the long post)