Marshal AVT - See ya later TDA7193

[cjenrick]

some people like the sound of the Valvestate chip amps, until they blow up.

so they get a new chip and kabang! plastic melting like so much candle wax,

well my fun meter has expired. can hardly move in my work space there is so much TDA bone pile.

time to move on. don't wanna hear about discharging caps or circuit board jumpers, diodes,insulators, grease or no grease, this that and the other. don't wanna search endless forum threads where poor souls have suffered endless amp tech bills and digikey shipping charges, fed up waiting for fed ex.

don't you just love the way these amps start to squeal and feedback and break into chip destroying ultra sonic oscillation when the knobs are pegged?

here is one of the the perpetrators right here>

[cjenrick]

some of you old timers might remember the Acoustic 150. 1970's solid state technology,

pretty bullet proof power amp, the jury is still out on the preamp,
some like it, some not. maybe a little bit to hi-fi.

sure do like the sound of 2N3055's

here is one of those amps right here>

[cjenrick]

can you see where i am going with this?

that's right,
would it be possible to splice on the Acoustic power amp to the tail end of the Marshall preamp, thus bidding a final farewell to the troublesome 7193?

let's give it a try.

the biggest problem will be the innerstage transformer. they don't grow on trees ya know.

but if there are grounding issues with the TDA daughter board (there are) then i bet the isolation provided by this transformer would get rid of the nasty feedback,

we did a tear down on one of these guys a while back so we gots the specs...

[cjenrick]

here is the bi-filar wound secondary of the step down transformer, scatter wound, we can improve on that with some patient layer winding,>

[cjenrick]

here is the blue print for the driver xfmr if you want to roll your own of have somebody else do it>

[cjenrick]

luckily there is a ton of free space to work with in this amp,

we will need a turret board to lay down some components, the 2N3055 power transistors have been mounted to the old chip amp heat sink, and we have the new driver transformer bolted in>

[cjenrick]

we have our power transistor support components in place, we are hoping to drive the transformer with the TL072 opamp signal from the Marshall preamp so that we can keep the parts count down, we are checking DC idle current by using Ohms Law on the .33 ohm emitter resistors. power is being supplied to the output transistors by way of the plus and minus rails of the dual supply.

[Mick Bailey]

sure do like the sound of 2N3055's

Do you mean you like the idea of them, or do you think they actually have a 'sound'?

[cjenrick]

driving the transformer with the 072 opamp was a no go. there is not enough primary inductance to keep from dragging the opamp signal down. this is because the transformer has an air gap to prevent DC flux saturation.

so we add the rest of the Acoustic driver circuit which consists of two more transistors, a mid sized power transistor (RCA 40410 heatsinked model) and a small signal transistor to drive the base of the bigger guy. you might remember the 40410 from the Sunn Beta series.

since the 40410 is an obsolete part (can be sourced on ebay and amp supply shops) we want to upgrade to a easy to find modern part, but what about the built in heat sink? simple, just use an insulator kit and use the chassis.

we ended up with a TIP 30 (PNP) for the transformer driver and a MPS A06 to drive that.
you can see the TIP chassis mount below the transformer.

we chopped one of the turret boards ( two boards $7.95 from China free ship) in half since the parts will not take much space and we are running out of space inside the amp (bottom left)>

[cjenrick]

we had to play some games with the power supply grounds to get rid of some hum when adding the transformer driving circuit,

this consisted of using the CT/Chassis ground for the driver circuit and the plus and minus rails for the output transistors.

thus turned out to be a good thing as it dropped the Vcc to the driver circuit from 75 volts to 32, which means less heat in the dropping resistor and less heat on the TIP 30. The original Acoustic 40410 ran very hot as well as the power supply resistor. This circuit runs very cool. Here is a shot of the the 2N3055's >

[Enzo]

7293...

I'd recommend MJ15003 or other modern part over the 2N3055. They are more robust, and voltage will never be an issue, while 2N3055 might be marginal.

40409 and 40410 are readily available, I even have a couple bags of them, NOS RCAs.

Let us know how it works out. But I would make one further suggestion, gather your thoughts and present...

[Enzo]

...just one or two...

[Enzo]

...posts to describe the work.

[cjenrick]

here is a print of the modified driver circuit. the reason we choose the old school circuit is because it has a blocking cap feeding the speaker. this greatly simplifies the circuit as you do not need a complex short circuit/overload circuit which is required on modern solid state circuits which use direct coupled topology. Fuses in series with the output circuit are a pain as if they are sized right, the sometimes blow due to heavy transients.

there are only four transistors in this circuit, two 2N3055's, a TIP 30 and MPS A06, easy to get and cheap.

we upped the output cap from 1000 uf / 50 volt to 3300 uf / 63 volt Nichy which should be very dependable due to the reduced voltages of the Marshall (we went from 75 to about 63)

here is the modified Acoustic circuit, since the Marshall was rated at 50 watts, we can chop two output transistors from the 150 circuit, we do not want to overload the power transformer of the Marshall.

our TL072 output signal was too hot to drive the Acoustic circuit due to the extra gain of the innerstage TX driver circuit, we lucked out by finding that this problem is easily solved by increasing negative feedback by adding resistance in series with the emitter bypass cap on the first transistor. By adjusting the series resistance with a decade box at full volume, we arrived at 10 K as the resistor value.

a word of caution, there is - 32 volts on the speaker ground as compared to the chassis, so you must remove the ground jumpers on the plastic output jacks, this takes about two seconds with the soldering iron. any mishaps like touching the speaker cord ground the chassis will simply blow the 1.5 amp fuse with no component damage.

well, how does it sound?

fantastic! no feedback or squeal or hi freq osc. at max volume on both clean and overdrive channels! the tone is amazing, don't know why, could be the 3055's, could be the transformer, could be both, you want Mississippi Queen tone? plug in a Les Paul Junior and let er rip!

everything runs very cool so this thing should run for years and years with no problems.

see ya down the road Mr TDA7193!

[cjenrick]

for a complete analysis of the original Acoustic 150 output circuit, go here>

https://groupdiy.com/index.php?topic=64083.0

it includes some commentary from the original designer Duke Aguilar as well as ex Peavey engineer John Roberts and resident guru PRR.

here is the original Acoustic circuit. you use pin one of the Marshall connector with the 7 blue wires (+ signal) for your signal hot, and ground the pin 3 wire to the chassis which is the signal ground. we were able to ditch components that were used for negative feedback with no ill effects..