TIP 147's paralleled with no degeneration resistors?

[SQLGuy]

Hi Folks,

I'm reparing a Coustic Power Logic 460. It uses pairs of TIP142's and TIP147's (TO-3P) for outputs. These pairs are completely in parallel (i.e. bases, collectors, and emitters all directly connected between the pair), but then share one 5W .1 Ohm resistor per pair on the way to the output (I guess to allow measurement of bias). The one I'm working on has one blown TIP147 - its "partner" is fine. I'm not surprised. My question is, did this setup ever work? How did Coustic expect paralleled Darlingtons to current share without emitter resistors?

Also, while I'm on the subject of design I find questionable... the temp feedback Vbe transitors are TO-92's effectively placed next to the heatsink and bent towards it a bit; there's nothing to hold them in contact. Did this ever work either?

Skeptically yours,
Paul

[jackinnj]

Time for a POOGE -- parallel Darlington's are a recipe for disastre (savez vous?)

[ppia600]

I never had any problems with my power logic 380 or 190 and they both had the same design. Cool amps IMO

Originally posted by SQLGuy
Hi Folks,

I'm reparing a Coustic Power Logic 460. It uses pairs of TIP142's and TIP147's (TO-3P) for outputs. These pairs are completely in parallel (i.e. bases, collectors, and emitters all directly connected between the pair), but then share one 5W .1 Ohm resistor per pair on the way to the output (I guess to allow measurement of bias). The one I'm working on has one blown TIP147 - its "partner" is fine. I'm not surprised. My question is, did this setup ever work? How did Coustic expect paralleled Darlingtons to current share without emitter resistors?

Also, while I'm on the subject of design I find questionable... the temp feedback Vbe transitors are TO-92's effectively placed next to the heatsink and bent towards it a bit; there's nothing to hold them in contact. Did this ever work either?

Skeptically yours,
Paul

[SQLGuy]

Originally posted by jackinnj
Time for a POOGE -- parallel Darlington's are a recipe for disastre (savez vous?)

I think I'm going to separate the emitter paths and replace the single .1 Ohm with a pair of .22 Ohms (one for each device).

Oui, je sais.

[acid_k2]

Originally posted by SQLGuy
These pairs are completely in parallel (i.e. bases, collectors, and emitters all directly connected between the pair)

Paul

some EEs are bad
and their design may be even worst...

Originally posted by SQLGuy
the temp feedback Vbe transitors are TO-92's effectively placed next to the heatsink and bent towards it a bit; there's nothing to hold them in contact. Did this ever work either?

Paul [/B]

the contact on the heatsink is optional; without that, the thermal feedback is a bit slower, and the bias current stays more stable on pre-fixed value.

[Eva]

Car audio amplifiers wouldn't fail so often if they were well engineered...

Two darlingtons paralleled directly will never share current: The hottest one will exhibit lower Vbe and carry more current becoming hotter, the coolest one will exhibit higher Vbe and will carry even less current becoming cooler.

Don't worry too much about thermal feedback, most car audio amplifiers are plain class B (with crossover distortion) to avoid the cost of the bias adjustment process at the factory, and yours probably is class B too.