Bias Drifting Higher?

[toufu]

Hi. I am building an amp using this schematics (mbl 8006 clone), however, the bias testing point on one channel, keeps drifting higher and higher (ts3 and ts4), could somebody give me some pointers on how to diagnose this? Thanks!

[Mooly]

If the bias (with no load) drifts it can only be because the vbe generator is allowing an increase in volts across itself or that it is not correctly positioned to track the output stage temperature changes. Does the voltage across the collectors of T7 and T12 increase as the bias drifts and how does this compare to the good channel.

[toufu]

Thanks for the help, Mooly. Actually, the collectors voltage across T7 and T12 for both channels decrease slowly, only that for the good channel, TS3 and TS4 stabilizes and for the bad channel, the bias voltage increases slowly. Any idea as to why?

Also, how important that the MOSFETS to be matched for this circuit?

Thanks.

[sakis]

yammi yammi .... i was looking for a thing like that long time now !!!!!

first of all what you are talking about is not mosfets !!! BDV xx family is the worst transitor ever used in audio amplifier .....its a darlington and its the king of unstables !!!!

i think what causes your bias drifting is either oscilation or bad thermal design

now .... this the first amplifier ever presented in the forum that has more than one pair of outputs designed arround BDV 65-64 or 66-67 family ( first cousin if not brother of TIP 142 147 ) .... the only one ever presented is the A40 from Nelson pass that features 2pairs per chanel ....

this happilly arises a few very serious questions to be answered ....

1) where the hell did you get BDV 65-64 transitors ? i belive that this impossible to get nowdays ... i could presume that you have BDV 66-67 pairs or TIP 142-147 pairs that are easy to get

2) none of the amplifiers ever presented by many designers ( and that will be only with one pair of outputs) was stable enough to operate with large or very large miller or stabilzation caps and that will be 560 pf from B to C one the outputs ...under this logic the amplifier you have is impossible to operate .Yet again the only exception to this rule will be again the A40 that is biased at 400ma per device ( if memory serves properly ) and that will support my theory that if a darligton is biased high is too bussy to oscilate and becomes actually stable .

3 ) the schematic states 36+36 volt rails ... in this case 3 pairs of outputs to produce 50 Watts is an overkill ( for class AB ) unless your amp is highly biased and closer to class A What is the bias of your amplifier ????

finally ....if transistor are proper ( you may find interesting this thread about what is a proper transitor A Darling.......ton Story .... ) then thermals have to be perfectly designed ...pcb have to be perfect ....and bias have to be high ...then yes you might have a chance ...other than that ...i see it as good as a ticking bomb !!!!

experts please join in i enjoy this so much

Thank you for your time

Kind regards sakis

[sakis]

one other thing i missed ...

If you are trying to repair this amplifier and any of the ouputs is blown and you replaced one BDV64 with a BDV 67 or a TIP 147 ( or even much better an NTE replacement )
DUCK NOW !!!! in a few minutes it will be too late !!!!

(I presume that we can share the same sense of humor )

kind regards
Sakis

[Tekko]

Heres a tip: throw away that useless thing and build a proper amp like the new DX Precision or one of Apex's designs.

[toufu]

Thanks for all your help and support. Actually, I am using MN2488 and MP1620 in place of the BDV's. Does anybody have any opinions on how to get this to work properly or help me diagnose the problem? Thanks again.

[sakis]

well done Tekko .... we all learn from things like that .... obviously you share the same sense of humor ....

[sakis]

what is the bias of the amplifier ?

[nigelwright7557]

Sounds like the bias could be too high.
This is cooking the output transistors into thermal runaway.