Marshall AVT150 keeps blowing up TDA7293

[vintasonic]

Hi

I have repaired a bunch of the Marshall Valvestate with the TDA7239 outputs before, usually just a matter of replacing the bad TDA7239 and cleaning up the board.  I know that there are replacement modules but I had a half dozen TDA7293s in stock so just thought I'd replace the bad ones.

However, the replacements keep blowing up.  No load, just slowly bringing them up on the variac and then all of a sudden - poof! - and it's a goner.  Checked all the connections, voltages are all stable, disconnected the amp out just to make sure there wasn't a short on the main pc.  Everything checked ok.  I would suspect a bad batch of 7293, I have had them for a while and don't remember where I got them but I think it was either Mouser or DigiKey.  I would buy some more but everyone is out of stock right.  That and I think there is something obvious that I am not doing in my troubleshooting.  In the 150 the amps are used in a master/ slave config, so I got to the point of only trying to get the 'A' amp working first.

Any ideas?  Here's a schemo                     https://mega.nz/file/xfwnXSzK#5oLqw1Yq0R8qULoGikdnj4GlZJEpY1X_pdQNvFbVTms

[Latole]

Not easy to follow this schematic
Where CN1 go ? To CN102 ?

[sluckey]

Not easy to follow this schematic
Where CN1 go ? To CN102 ?

CN1 goes to CN102 or CN103 on pdf page 8.

[Latole]

Not easy to follow this schematic
Where CN1 go ? To CN102 ?

CN1 goes to CN102 or CN103 on pdf page 8.

Thank's Sluckey, that is what I tough.

Very "funny" way to name male and female connectors......

[sluckey]

Not very intuitive. Probably make more sense if you were also looking at the amp.

[Latole]

Not very intuitive. Probably make more sense if you were also looking at the amp.

Right !

[vintasonic]

Sorry - I been staring at it for so long it's become pretty familiar to me.  On Page 8 you can just look at Amp A, pin 4 & 6 don't really matter much, 4 is stand-by and 6 is for paralleling the signal of the TDA7293 to amp B, the B amp is for better performance into lower impedance loads. Pins 5, 3 &1 are power ground, ground for signal and grounding one side of the balanced input. Pin 7 & 8 are for the bootstrap cap, the links are for changing the value when two amps are in place but there is always a bootstrap in circuit even with only one amp.  The only real difference between amps A & B is pin 3 - on the B amp it is tied to Vs- to tell it to be the slave.  So, I should be able to run just the A amp for testing.

What I don't get is the sudden short that happens as I power it up.  I was monitoring the rails and everything seems OK.  As I bring it up on the variac it's fine and then all of sudden there is a massive current draw and the chip blows up.  Again not connected to a load so the chip is blowing itself up for some reason

[Keppy]

I was monitoring the rails and everything seems OK.  As I bring it up on the variac it's fine and then all of sudden there is a massive current draw and the chip blows up.

What are the rail voltages when it blows up?


I'm wondering if the variac is helping you here. Most of the amp innards are running off the +/-15v regulators, which can't really regulate until they receive an input voltage that's a couple volts higher than their output voltage. I'm not sure what their behavior is below that threshold.


Have you measured the voltages on the PCB pin holes at full power without the chip installed?

[PRR]

I'm reading a lot of posts about fake power amp chips. Typically they work at lower voltage, but go up and POOF.

So how sure are you about the pedigree of your chips?

[vintasonic]

Thanks for your input.

The suggestion about the internal +/- 15 is interesting but the chips blow up just when I am at or near full rail voltage, so the regulators should be working at that point; the rails read around +/-35.  There is a little crud on the pc board near the filter caps which I didn't give much mind but will revisit when I get new parts.  One thing I ordered was full amplifier module that takes the transformer AC, so that will eliminate most of the Marshall power circuitry.

The quality of chips is questionable, I've had them a long time so I am not sure where they came from, they may have been a eBay purchase.  Tjhey are marked ST but that doesn't mean that they aren't conterfit. Unfortunately, Mouser and DigiKey have about 40 week wait times so I ordered a couple off eBay and will replace the filter caps and see what happens.  Very possibly the fix will be to use the amp module, I saw a post somewhere else and the tech used a similar kit to fix the problem. 

These amps blow up the output chips often, this is not the first repair that has been done to the outputs on this amp.  So if the stand alone module works, it will actually be a better and cheaper solution.

[Keppy]

the rails read around +/-35

As in 35v above ground and 35v below ground? If that's the case, then the rails are 70v from one another and your regulators are fried.


If you measured 35v from one rail to the other rail (not to ground), that's probably close enough to work fine.

[glass54]

Hi Vintasonic
Please be aware that according to ST Semiconductors Data (v8, 2010), the absolute maximum ratings for the TDA7293 Power chip are +/-60V DC rails and peak Current of 10A (repeat max 10A pk). Your rails of +/-35V is well in the suggested operating zone as specified by manufacturer. Also looking at the chip summary, a proper ST device is very well internally protected.
Having replaced some of these in Marshall's some ten years ago, I suggest as did PRR, that you either have dud chips OR less likely a possible power supply fault. I would check C3 and C5 to confirm good PS decoupling at the TDA device. You could also run the amp with TDA removed, check PS Volts (will probably be a couple volts higher) and possibly try running a pair of dummy loads on the +/-Rails to confirm a clean PS (say 470R rated at 5W or 10W) and they will get quite warm . Try this for 30min or so confirming clean PS. Also while the TDA is out just check the resistance of CN1(3), IC pin 14 to ground, but I don't think you'll have a problem there.
If your PS is good and no other PCB/wiring problem exists, I would suggest only one scenario, which has already been covered.
Love the Valvestates!! 
Kind regards
Mirek

[st]

Little help to you, but when I was experimenting with these chips recently, I had the exact same problem with a few of them but not others. In the end, I concluded it was due to fake chips (sourced from different places). It must be a plague ...

[shooter]

"Love the Valvestates!!   "


I fixed 2, after that told the guitarist to throw them away n buy a broke 60's/70s tube amp, THEN i'll work on it!!!!

[vintasonic]

Hi


Just and update on my lack of progress.  I also dug into the power supply some more.  As I said, I purchased a TDA7239 power amp module assembly, just for test sake.  I also have a couple more TDA7239 that I got from eBay, not counting on them to be legit.

So I tried the module in the AVT150, it has a rectifier and filtering on it so that would leave most of the Marshall circuitry out of the picture.  I started to bring it up on the variac and same thing - it starts to draw power like a dead short.

I then dug up another transformer same secondary voltage that I was going to use on another project.  So, nothing but the module and the transformer with no load it starting drawing current like a dead short and suddenly popped.  The line voltage was around 30 volts.

Is there some reason I can't bring the TDA7239 up on a variac? I am pretty sure I have done it in the past...

[thetragichero]

instead of a variac have you tried with a current-limiting device (lightbulb limiter)?

[shooter]

So, nothing but the module and the transformer with no load it starting drawing current like a dead short

No-Load as in just rectifier module n PT?  (No PA section)
If so;
should be able to OHM a dead short before you even apply power