Hi Ian,
Thank you very much for explaining things, this is what I was after.
If I have understood correctly, the pot will adjust the bias current through the FETs. This current can be determined by measuring the voltage across the 0R22 resistors on the FET sources (2 x 5 of them) when Vin=0V. Broadly speaking, higher bias current = hotter but better sound.
I'm a bit hung up on the 'correct' setting for the bias because one of the reasons I stopped using the amp was that it seemed to be running hotter than before and smelling hot too - but that could have been dust (now gone). I don't really trust the pot to be doing what it was originally set to do 35 years ago and am tempted to replace it but that means knowing how to set it.
I'd be very interested if anyone knows what the factory set bias current might have been.
Thanks too for the facebook/Audio Tech Service Network link - I wish my board was that clean and the big yellow "Specially made for Musical Fidelity" caps were as free from heat damage!
That pic is likely taken after a full clean-up. Don't worry too much about having shiny, bright metal and Lego coloured parts in architecturally balanced positions. It impresses but adds nothing to performance. When you bolt the cover in place, there's nothing to see anyway.
You have the setting procedure idea down OK but before touching anything, you might use a vacuum cleaner with the aid of a soft brush to suck out that dust and grime, taking care to avoid bumping things or disturbing the pot. settings. You need to record how things are set now, in both channels, as a reference point that you can return to with certainty.
Provided no one has fiddled with the bias setting pots, you should be able to establish the present settings as roughly correct just by comparing one channel's bias current to the other. Measure the voltage across each resistor and calculate the total of 5 currents, one from each series pair of resistors as explained. Identify and record the voltage across each OR22 resistor on a copy of the schematic or a sketch so there is no chance of error when you need to refer to the results of specific parts measurements again. Remember, the series currents don't get summed as you would be double counting the current. Just add the 5 parallel currents as calculated from either one of the 2 series resistors, or from the voltage across the total resistance of both resistors in series (i.e approximately 0.44R, if you like). The current should work out to be the same either way, give or take a gnat's whisker.
We are looking at several possible faults together here, and failed Mosfets is a possibilty that could cause the remaining Mosfets to run hotter than intended. There is still a lot to check there and I'm assuming you have a DMM that can read low DC voltages with precision, down to 1mV or less, even if not strictly accurate. I would not try to make do with anything less.
You have the setting procedure idea down OK but before touching anything, you might use a vacuum cleaner with the aid of a soft brush to suck out that dust and grime, taking care to avoid bumping things or disturbing the pot. settings. You need to record how things are set now, in both channels, as a reference point that you can return to with certainty.
Provided no one has fiddled with the bias setting pots, you should be able to establish the present settings as roughly correct just by comparing one channel's bias current to the other. Measure the voltage across each resistor and calculate the total of 5 currents, one from each series pair of resistors as explained. Identify and record the voltage across each OR22 resistor on a copy of the schematic or a sketch so there is no chance of error when you need to refer to the results of specific parts measurements again. Remember, the series currents don't get summed as you would be double counting the current. Just add the 5 parallel currents as calculated from either one of the 2 series resistors, or from the voltage across the total resistance of both resistors in series (i.e approximately 0.44R, if you like). The current should work out to be the same either way, give or take a gnat's whisker.
We are looking at several possible faults together here, and failed Mosfets is a possibilty that could cause the remaining Mosfets to run hotter than intended. There is still a lot to check there and I'm assuming you have a DMM that can read low DC voltages with precision, down to 1mV or less, even if not strictly accurate. I would not try to make do with anything less.
Last edited:
Thanks Ian - I'll get measuring. I'm fairly comfortable about what to measure now. Will see how the left and right channels compare. It will probably be the weekend since the day job is shouting and there's quite a bit to do before I can start measuring.
Thanks again
Thanks again
The 2SJ119/2SK414 are not lateral mosfets which the originals are, that is why the drain and source are swapped over. To get back to the original you could use exicons but if you are happy with the sound then no point in changing.
Yes, I should have looked at and commented before posting that link....my bad for confusing matters. I imagine that since he bought it new, nigelm's amp still has lateral mosfets and it probably would mean replacing all of a given type/rating and polarity in a particular channel if replacements were necessary. Adapting to VFETS might prove to be a difficult process here. If Profusion PLC's laterals are available, it would be simpler but the originals are HPAK (2-bolt flanged and obsolete type) and the repairer shows a flat aluminium bar is used to adapt the TO3P style, 1- bolt replacements.
I don't much like that idea but if that's what it took to repair it in the circumstances, needs prevailed, I guess. The originals are low cost, 8 amp, 100W rated and Profusion PLC has TO3P types ECX10N20 or 10P20 types or newer TO247 versions that should fit the space without problem. However, these would all require new bolt locations, new holes and threads tapped for 3 mm bolts.
Much easier and better, would be 2 rectangular or channel bars spanning all the Mosfets on either side and just using the original tapped holes to clamp the FETs firmly with long bolts passing through the bar and screwed into the holes.
I don't much like that idea but if that's what it took to repair it in the circumstances, needs prevailed, I guess. The originals are low cost, 8 amp, 100W rated and Profusion PLC has TO3P types ECX10N20 or 10P20 types or newer TO247 versions that should fit the space without problem. However, these would all require new bolt locations, new holes and threads tapped for 3 mm bolts.
Much easier and better, would be 2 rectangular or channel bars spanning all the Mosfets on either side and just using the original tapped holes to clamp the FETs firmly with long bolts passing through the bar and screwed into the holes.
Russel - thanks for the clarification about the FETs. As Ian notes, I've had the amp from new so it has what MF fitted which in my case are J82 for the P channel and K226 for the N channel.
I'm really hoping I don't have to replace any of them - that's going to be a different topic but it's interesting to read the thoughts here.
Would I be right in thinking that if any of the FETs has failed then the drain / source current of that FET will be 0 (and the FET will be cold)?
Thanks to everyone for the input on this.
I'm really hoping I don't have to replace any of them - that's going to be a different topic but it's interesting to read the thoughts here.
Would I be right in thinking that if any of the FETs has failed then the drain / source current of that FET will be 0 (and the FET will be cold)?
Thanks to everyone for the input on this.
It looks like the FETS have been put in with self tappers and there is no washer to the epoxy package which might affect reliability. The profusion single die parts should be a drop in replacement size wise but I dont like the idea of self tappers. It would be just north of 100UKP to replace all twenty, so not insignificant, although profusion can match for current balancing.
No need to butcher the job with self tappers if you utilise the old tapped bolt holes and add aluminium clamp bars of about 6 x 10 mm or so section, for rigidity. It woudn't be difficult to cut these from standard section offcuts at a window, caravan or hardware supplier store if they were the helpful type, then drill some straight and plumb 1/8" or maybe 3.5mm holes at matching spacings.