quick bias tolerance question...across source resistors

[Mattyo5]

Ok, i have set bias after inserting Q10 correctly on my anthony holton symmetrical 400 ...and temperature is stable and everything. I thought I matched the devices pretty well, so the bias voltages across the source resistors should be close right? How close? I set voltage at 18mV for one...then went back and checked when the amp had stabilized. Some were 13mv ...one was 37mV! . 13mv is ok...but 37? Now...i have pretty big heatsinks...so the amps are probably not even reaching 100F...is 37mV too high? I cranked the bias down a bit...so I don't think its that high now, but should I take that one out and replace it and hope for the best? or just leave it b/c these irfp240's can take a beating? thanks!

-Matthew K. Olson

 

[Andypairo]

It also occurred to me once, did you match the fets with the current you are using now?

My suggestion is to try to obtain the closest voltage value possible in-circuit (swapping the Mosfets or the source resistors for a first try and then using some other Mosfets in case the first try doesn't work well).


Cheers

Andrea /who is going to have a weekend

edit: to be safer place Q10 on the top of the fet with highest bias

 

[Mattyo5]

Can't really place Q10 ontop of the fets...they are bolted down w/ an aluminum bar...just like Anthony has his av800 amp and av1000. So Q10 is bolted to that...just next to one of the output transistors. Can I really changed the gate resistors...? Say from 470 to 560? think that would help? or the source resistors...maybe stick a .47ohm in there...that wouldn't screw up other stuff? That'd be much easier than swapping the whole transistor out. Thanks!

-Matthew K. Olson

 

[DrG]

Changing gate resistor will make no difference. Increasing source resistor could equalize current but that's not the way to go. Bite the bullet and whip out the FET!

 

[Mattyo5]

ok, so now i'm trading in and out the fets....lovely....how do I know that the one that i get in there will be a match? i can't possibly know...unless I rip out all 6 outputs and rematch them, so then it leaves it to trial and error.... what about just leaving it...plently of heatsink...37mv is high but not like the thing is going to go into thermal runaway ...i have plenty of heatsinking....
they run way below 100F.


yes no?

-Matthew K. Olson

 

[Nelson Pass]

I set voltage at 18mV for one...then went back and checked when the amp had stabilized. Some were

Don't get so excited. Turn the bias up to where both
channels idle at 45 deg C. and sleep well at night.

 

[Mattyo5]

So, the "large" difference from one fet to the next in bias voltage isn't all that significant? ie, some are 13mv others are up to 37mv? and thats ok?

and also you are saying crank up the bias more?...so the heatsinks get a little hotter? this is class AB amp

thanks!

-Matthew K. Olson

btw, the heatsinks...to the touch ...feel about the same temp ...barely warm :-D