The legendary / infamous Marshall AVT-150 power amp issue

[timgman]

Hello,
I recently opened up my "blown" Marshall avt150 and the power amp boards are dead. I was looking a tthe schematic and I think (depending on the way things are) I can use a set of "prebuilt TDA7293" power amp modules instead of replacing the chips on my boards. These modules seem to only have 6 connections...POWER > IN > OUT.
so if I were to "ignore" the other connections on my avt (standby / mute / ?) could I just put one of these generic modules in instead?
Here's a sample of one of these modules.

https://www.google.com/search?q=tda7...96505545354103

[timgman]

There may be functions on. the amp such as "noise gate" or standby things that use the other vhip functions but overall.. it may work and be a sustainable cheap modular fix. the marshall outputs seem to be around 80$ shipped each IF you can find them. that's a 160$ bill for a amp that's possibly worth 200$. I'd prefer to not toss the amp. I donlt mind using mine as a test bed for the betterment of the AVT world.

[teemuk]

Your module is a full 75W mono amplifier; TDA chip and all the external parts required to make it work in a specific way. The Marshall module is nearly just the chip and all the external parts to make the chip work are on the main board. The little TDA chip board just contains local power supply decoupling caps and other stuff for inhibiting oscillation. So, the other module is not going to be a direct fit replacement and you may have to modify the main board as well.

For instance, AVT150 uses two 75W TDA chips to generate its rated output power. The way the main board is wired results to either 2x75W (of AVT275) or to 1x150W (of AVT150). How are you going to wire your mono modules? Also the "common" terminal of the load "floats", which you don't see in generic audio amp modules but is essential for specific operation of Marshall's circuit.

The mono amplifier module is also designed for Hifi audio. Marshall circuit is not. It implements current feedback to alter damping factor so that the power amp responds to reactive loads similarly to a tube guitar amp. The module will not have same tone, probably not even same gain overall. It's not a direct replacement.

You need to really ponder how much rework there will be to fit that full amp module there. It's more complicated than you likely thought.

How much does a TDA chip cost today? I bet not $80 like the replacement Marshall module. So, I'd rather buy the chip and clone the little Marshall board. You even have the schematic for it. Better yet, build few for backup: Marshall's design is not thermally reliable and with insufficient heatsinking provided nothing similar will run reliably there.

[Jon Snell]

What is wrong with replacing the chip?
Only 15 pins and with care will save you a fortune.
DO NOT I repeat DO NOT purchase sub standard boards or even chips through fleabay/amapong etc. You have no idea what you are getting!
It takes me less than 10 minutes to replace the chip and check for damage.
Purchase the chips from a known seller; RS, Mouser, Farnell etc.
https://uk.rs-online.com/web/p/audio...er-ics/0218478
If you can't do it, any TV/Radio engineer can.

I have only ever had one TDA fail and that was due to user error. They are very robust both over current and over temperature safe.

[timgman]

I think that the boards have had the chips replaced a few times and there's a jumper wire already patching a weak / trace. I'm not certain oif the boards are conductive or not. I wonder if a breadboard makeup would be fun. I'm not certain I have the time to go that far. I will attempt a chip replacement first. Another thing of note. the left chip was replaced at some point with a NTE7186 chip. is that compatible ?

[Jon Snell]

I think that the boards have had the chips replaced a few times and there's a jumper wire already patching a weak / trace. I'm not certain oif the boards are conductive or not. I wonder if a breadboard makeup would be fun. I'm not certain I have the time to go that far. I will attempt a chip replacement first. Another thing of note. the left chip was replaced at some point with a NTE7186 chip. is that compatible ?

They are identical.

[g1]

Like Jon says, make sure you buy from a reputable seller. There are a lot of fakes out there.
Any of those modules (aside from actual Marshall ones) I would not trust.

[teemuk]

I have only ever had one TDA fail and that was due to user error. They are very robust both over current and over temperature safe.

Yes they should be. They have reliable track record in countless domestic audio amps, powered monitors and PA speakers, etc. Yet in Marshall amps they are notorious for failing, so I think the common denominator is Marshall's design, and its inadequate heatsinking.

Maybe Enzo has some deeper insight, IIRC he refuses to service the Mode Four with it's four TDA chips running in chip's built-in master/slave setup and bridge. One module fault blows all four.

[Enzo]

Mine is from a practical point of view. No matter what I do to it, I can get it running, but by its nature it will never be reliable. SO I just refuse to fix them and put my name on the repair.

To make it intuitive, when I looked at what the four module circuit was doing, I saw it as bridged, and one of the two on each side was slave to its mate. In my head I called it a "power follower". I don't know if there is such a thing, Maybe there is a better term. In any case, when one fails, most times it takes others with it. My rule was always if one fails, replace all four. And that is not the way I like to work.

Even not working on the Mode Four, I still ordered TDA7293 by ten each. In the AVT series, I don't know that they were any less reliable than other amps, but I surely replaced my share of the TDA chip.

Particular to Marshalls, If your power supplies are not fully discharged, if you plug the cables onto the module it can blow it up right there. Since the four modules work in bridge mode, neither side of the speaker is ground. SO if anyone connects anything to the output that can get to ground, it blows the modules. I found that one must never connect or disconnect the load while the amp was powered. And the various impedance jacks on the rear are there for a reason.

There are just too many ways I can screw up my own repair and have to start over. And ther eare too many ways for the customer to screw up his own amp. I'll pass on these.

[Jon Snell]

Mine is from a practical point of view. No matter what I do to it, I can get it running, but by its nature it will never be reliable. SO I just refuse to fix them and put my name on the repair.

If your power supplies are not fully discharged, if you plug the cables onto the module it can blow it up right there. Since the four modules work in bridge mode, neither side of the speaker is ground. SO if anyone connects anything to the output that can get to ground, it blows the modules. I found that one must never connect or disconnect the load while the amp was powered. And the various impedance jacks on the rear are there for a reason.

There are just too many ways I can screw up my own repair and have to start over. And ther eare too many ways for the customer to screw up his own amp. I'll pass on these.

I completely agree.
It takes only a few left over volts from a semi discharged capacitor to destroy a TDA or most other sensitive chips for that matter.
Far too easy to break for a coffee and forget to check before reinstallation.

Another issue these have, similar to most older car radio amplifers, the output stage can be bridged and those unknowingly connecting a loudspeaker lead that has one side of the lead ground, (another amp is plugged into the spare socket on your 4 X 12 cab), ... BANG!

[Mick Bailey]

Did anyone come to any firm conclusion over the spec sheet changes? See post #70 here; https://music-electronics-forum.com/...owing-up/page5

[J M Fahey]

My personal view, others may differ, in the "3886 good reliable/729x not that much" debate is:
BOTH have exact same case, so exact same heat dissipation capability, I mean the Real World end to end type, at real world temperatures, not fairy dust "25C case temperature" which is physically impossible (would require sub zero heatsinks, just do the Math) BUT one is speccd at 50/68W tops, with strict +/-V rails , while 729x is foolishly speccd at +/- 50Vv rails, or 100W RMS, which is NONSENSE.

Maybe in a Lab,never ever "out there"

I am certain that IF they are used at LM3886 ratings, they will last "forever" or nearest decade, he he.

Special points for the stupid parallel active voltage sources idea.

The very ultra-care needed to achieve that (0.1% matched resistors, for Godīs sake!!!!) , unnecessary lossy 0.1 ohm series resistors *outside* the NFB loop. etc, screams that.

No problem at all with bridged outputs, as no problem with series voltage sources.

Oh well.

[mikeskory]

Wondering if I can comment on this older thread. I'm working on a AVT150 head right now. It's the AVT150 head with 2 amp boards. The "B" output connector appears to have -50v going through pin 3. However at the "A" board pin 3 is ground. Am I completely mis reading the schematic or could this be correct? Thank you

[The Dude]

The schematic is here:

https://elektrotanya.com/marshall_av.../download.html

I believe the amp modules are the same, so reading pin numbers on the amp boards should be the same from one to the other. However, at the other side of the connector from the "send" side, pinout is different from connector to connector. Pin 3 is ground on one, but not the other. Wires must be crossed in the cable? It's odd- even though I've repaired plenty of these, I've never had cause to notice this before.