Trace Elliot Bass amp repair

[SLEEAudio]

Hi guys,

My first post here...

Im looking at a Trace Elliot for a friend. AH300, schematic canbe found on this thread

http://music-electronics-forum.com/t19560/

The issues is, is that there is about 55v at the output terminal, reading with an osciloscope, it appears to be AC. This voltage only appears when the relay (shown in the power amp schematic) turns on. Could someone explain the reason for this relay please?

The negative PS cap (4700uF, 100V, -70v) has some damage to it, but it still works , however, it take a very long time to discharge...

....so first things first, ill replace that. But other than that, I need help determining the cause for this dangerously high voltage at the output.

Thanks for your help

Charlie

[Steve Conner]

Are you sure it's AC? The classic failure mode of all solid-state power amps is a high DC voltage on the speaker terminal, caused by one output transistor shorting. Take a multimeter and read the voltage using the DC volts range and also the AC volts range, and post the two results here.

The relay is for disconnecting the speaker. The purpose of this is to:

1- Mute any weird pops, thumps, squeals etc. that the circuitry makes when powered up/down.
2- Remove the load if the amp overheats
3- Save the speaker from being fried by DC if the amp blows an output transistor.

Many amps only implement one or two of these three functions. I can't remember off-hand which the Trace has: but if you have DC on the output and the relay still closes, I guess it's 1 and 2.

[SLEEAudio]

Hi Steve,

Thanks for your reply.

Your right, I think I jumped the gun saying it was AC, it reads 55v DC. With a scope its very similar to the 70V at the ps cap (lots of ripple).

I am not able to test it again as the ps cap has been removed and doesnt look safe to put back into the circuit.

So the relay closes and this 55v appears on the output. So you are suggesting that the amp is over heating and the relay is closes? So maybe I have a shorted output transistor? Could this be the reason for the high DC at the output?

Apologies, I am a lot more confident with valve circuitry than SS. Need all the help i can get.

Regards

Charlie

[Enzo]

-70v and takes a long time to discharge? Sounds like a good cap to me. I suspect the OTHER cap is the bad one.

Now about the other cap, Does it also have about 70v on it? Or is it a lot lower. And flip the meter over to AC or just scope it. Is that smooth and filtered, or is it pulsing unfiltered DC?

Both caps should have about 70v on them, one per polarity. And not a lot of ripple on either.
The relay drive is complex, it is associated with the fan circuit, so there is a thermal control aspect. I think it also puts a turn-on delay in there. I don;t see any overcurrent or DC detection offhand.

[Enzo]

If one 70v rail is unfiltered, it will read low on a meter, and since that rail will now be pulsing, so too will be the output. Repair the power supply, and I bet your output will settle down.

[SLEEAudio]

Sorry, i think ive been looking at this wonky.

Caps read +64, and -64 DC

AC is negligable, meaning low ripple (right?)

And at the output, 55V DC, 0.1 V AC

[Steve Conner]

Yup, it's looking like a blown output transistor. Pull the transistors and check them.

Edit: Although, if it really was a blown tranny, surely the output voltage would be 64, the same as the DC rail. Maybe it's something else, like a blown rail fuse or bad filter cap as was suggested.

Note on replacement output transistors:

They don't make the 2SJ50/2SK135 any more, but the BUZ901/BUZ906 and ECF10N16/ECF10P16 are suitable replacements. You can get the latter from Profusion in the UK.

I'd be tempted by the uprated dual-die versions myself: the ECF20N20 and ECF20P20.

[SLEEAudio]

Really helpful Steve, thank you.

Ill try all of the above and post my results.

Charlie

[Enzo]

Oh well, to bad it wasn;t a simple filter cap failure.

And a shame too, those SJ/SK power amps in the old Traces sound really good to me, very smooth and effortless. I have really missed those old transistors.

[Steve Conner]

Aw cheer up Enzo, the Magnatec/Exicon replacements are pretty much just as good, and the silicon dice inside are made in Scotland.

I think when Hitachi launched the 2SJ50/2SK135 they were a bit over-optimistic about the capabilities of them. "Short circuit proof" and all that. Designers took them at their word and made amps without any current limiters, and poof. I've seen a C-Audio PA amp that actually ran the 2SJ50/2SK135 off +/-100V rails, putting up to 200V across 160V devices. Those go poof quite a lot, and they have about 10 pairs in them. A friend was trying to fix one, and he tracked down the guy who designed it. He admitted that it was a mistake and C-Audio went bust.

That's why I like the dual-die ones, they're basically two 2SJ50s or whatever in parallel inside the one TO-3 can. So by installing them you can give the amp twice the number of output devices and make it almost bombproof. They cost over $10 each, but for a repair job that's not too bad compared to labour costs, etc.

There is maybe a reason to the smooth, effortless tone of the old Trace stuff: those MOSFETs with their simple driver circuit don't saturate and "hang up" the way BJTs do, when they're overdriven. They clip very cleanly and recover instantly. I think Trace added a further soft clipping circuit, so you can probably overdrive the power amp on signal peaks without even noticing it.

I have an old Maplin Mosfet Amp kit with one pair of 2SJ50/2SK135, that I'm keeping as a collector's item.

[Enzo]

I have a quad power amp of some brand of other in the warehoouse. SOme fatal thing has happened to it. But the otput sections of the four channels are those SK/SJ parts. There are enough TO3 thingies in there it looks like a Peavey CS1200 or something. All MOSFETs. I am hoarding it. Each one of those parts will enable a repair one day.

Those old Trace outputs rarely came in burnt up. Swap a transistor and off to work it goes.

[SLEEAudio]

Can someone walk me through the proceedure for testing these transistors, so I can single out the faulty one.

Thanks

[Enzo]

Are those outputs in sockets? Pull the two screws and the transistor can be pulled from a socket? If not socketed, then two screws come out and unsolder the two pins to free the part... if you have to.

In any case, those things have two pins underneath and the case is the third "pin". No pins should be shorted to other pins. If you measure what seems to be a shorted transistor, remove it from the amp and test it while it is loose. That will reveal if it is the transistor or something else on the board that was shorted.

[Phostenix]

Can someone walk me through the proceedure for testing these transistors, so I can single out the faulty one.

Thanks

Here's a link to some good info, including pinouts & testing.
Transistors

[SLEEAudio]

excellent, thanks for your help