Myst amplifier

They probably both just used the MOSFET maker's reference circuit with minor mods.

Anyone who still uses such a piece of kit isn't going to put up with a repair like that. ;-) The original output devices are 2SJ82/2SK226 which are obsolete, but can be bought for 40 quid each. The current substitutes cost 4 quid. That is causing some sucking through teeth,

Could be - but they are closer in design to each other than to any of the basic circuits I've seen. Including the Hitachi one.

Since I'm not that familiar with this sort of circuit - anything obvious to look for when one of the output MOSFETS has a hole in it ;-) other than that driver transistor too?

John Linsley Hood ?

Have you tried J Birkett?

you serious? a real hole?

VERY unusual

MOSFETS may well NOT take the drivers with them. There are generally

Yup. Would you like a pic? ;-) Happened to the originals and some replacements fitted earlier. Have them here.

Yes - 240R resistors in series with the gates with a pot between them.

I was thinking of trying to test things without the MOSFETs in place - the replacements haven't arrived yet. To make sure there isn't some vast offset or whatever?

ah the bias pot presumably..hang on.. if the pot is there there IS no predrivers - the FETS are doing the phase spliting

that's a plan..but the overall feedback will come from the sources of the FETS and if they aren't there it wont be in circuit to stabilise the driver...so put a couple of resistors between gate and sourec where the fets ought to be.

I designed a really simple FET amp like this years ago - hitachi fets - and its pretty easy to get a relatively decent performance without going mad. The bigger problem is usually running the drivers off a slightly higher voltage than the FETS to get a better peak output..

I've decided to trace out the schematic, so will post a link when done.

Right. Will do.

I have the Maplin ones in this workshop. Driving some home made speakers. Which sounds as 'clean' as anything. Pure fluke, obviously.

Not really. driving 2n3055s hard and fast was a real challenge..driving a par of complementary fets hard is about as easy as driving a pair of EL34s. Its 'just works' Virtually no crossover distortion and what there is is very smooth.

And not much in the way of hard limiting either.

Tried it with 1k resistors, off my +/- 15 volt 1 amp supply and it seems fine. No offset, and it amplifies a sine wave cleanly.

The new MOSFETs arrived, and I repeated this with them installed. Drives a speaker ok - and again nothing looking wrong. Next thing is to try it with the proper supply. I've cleaned all the burn marks off the PCB in case it was simply tracking. ;-)

assuming they are lateral FETS the bias should be between 20mA and

Oh and those clamps will vary depending on which devices you actually use..so substituting a lower gate voltage device will mean the thing can overcurrent......

I'm told by the owner it blew up after having the output pair replaced with no load at all.

then the idiot probably didn't recalibrate the idle current. Or what he put in had a different leadout configuration. :-)

Here's the circuit if anyone is interested. Speaker switching unit left out.

Right. That looks simple enough.

I'd say that pot has gone AWOL in some way. You should be able to bias the idle current to nothing with that arrangement.

check what voltage is across it.

Or else the thing has gone RF unstable. Checkk the 27pf caps and the 6.8nF.