Question for Jerrold Tiers

[Stonie]

Hi Jerrold,

I've a '95 SVTIIIPro in my possesion (but not immediately right now) since new and I've been reading the various posts on here regarding the mosfet matching & bias adjust issue, ie. how to do it etc. and although I'm not an electronics guy by profession, adjusting the bias is well within my capabilities.

However, I've read a number of other posts on hear and other forums were it is being recommended, in order to "so called" help stabalise mosfet operation, to change the 47Ohm 1/2Watt (R28-31/40-43) and 0.47Ohm 5Watt (R32-39) resistors to a tighter 1% tolerance allround... Is this something you would recommend ?

In addition, for selecting and matching the mosfets... Is there any simple DIY way of doing this ?

TIA

[T-Bird]

Hi.

Originally Posted by Stonie

for selecting and matching the mosfets... Is there any simple DIY way of doing this ?

Originally Posted by Stonie

I'm not an electronics guy by profession.

I'd say You answered that question right there.

Matching MosFets isn't rocket surgery, but the thing is, You have to have a crapload of 'em in order to match to any meaningful tolerance. And You end up with a lot of matched pairs, quartets etc. that You can't probably use, and no-one will buy them as matched from non reputable source.

Just the same as it's with tubes, can't really match them without an ample supply of that particular tube.

Regards
Sam

[audiomitch]

The way I heard it is that only pairs have to be matched, and Ampeg has already done that. Matching all of them to each other is a waste of time and money. I'm sure someone will correct me if I'm mistaken.

[T-Bird]

Hi.

Originally Posted by audiomitch

The way I heard it is that only pairs have to be matched, and Ampeg has already done that. Matching all of them to each other is a waste of time and money. I'm sure someone will correct me if I'm mistaken.

Depends on the circuit topology, if a pair share the same bias circuit, then that pair has to be matched, if a triplet share the bias circuit, that triplet has to be matched, and so on.
Preferably, one wants both banks of push and pull to be matched as well to reduce distortion though.

I have deliberately avoided anything MosFet because, quite frankly, most of the circuits are a PITA to DIY'ers, so I can't remember what circuit Ampeg chose for this particular application.

I'm sure that Mr. Tiers will chime in soon and clarify it for us in any case.

Regards
Sam

[Jerrold Tiers]

OK......

Matching is by GROUP...... all the parts of one type in a group should match, but they do not have to be the same as the opposite group. So all the IRFP9240 should be matched, and all the IRFP240s should be matched, but the two types don't need to be.

Tighter tolerance 47 ohm does zip, zilch nada. Tighter tolerance 0.47 ohm theoretically makes a slight difference if they were at tolerance limits to begin with..... but is not a cure of any bias issue.

I don't recommend doing the biasing yourself, unless you are a decent electronics tech. Most who have done it were OK, but a few have toasted the amp, and that gets expensive. it also requires a decent digital voltmeter, one that measures millivolts accurately.

Matching Mosfets is not that hard, but requires stuff you are unlikely to have unless you are an electronics tech, power supplies, meters , etc.

I recommend not trying this stuff at home.... if you need to ask, you shouldn't do it, most likely. If you CAN do it, you already KNOW you can.......

[BassmanPaul]

+1 to both Sam and Jerrold posts. If you don't know what you are doing it's wise to stay out of an amplifier.

Matching requires a little jig with a load, a power supply and a voltmeter. It also requires a large batch of parts from which to choose matching devices. I had to repair three stereo MOSFET power amps. I bought twice the number of devices that I needed. Out of those I was able to match the six banks of devices plus another set that I still have. The remainder were not close enough for use.