Transistor Pwr Amp Pot

[JohnRoberts]

The diode in question is KB362, known as bias diode (about 2V).

Thank you,,, clearly intent is thermal bonding to heatsink/power devices...

Keep in mind many schematics are drawn by draftsmen not engineers, or even technicians. "And" old hand drawn schematics do not have to agree with reality...

JR

 

[CJ]

yes the diode is heatsink mounted as Moamps correctly signified, , and Q710 is close to the heatsink and bent over in an anemic attempt to provide some type of thermal coupling.

here is fwd vs i

 

[CJ]

yes, the backwards V+/V- thing got me a few times and eliminated a few devices unintentionally (they blew up)

still getting tricked,  hooked up a test circuit backwards,



funny thing is, it still works!  but the diodes do nothing, here is the erroneous test (diodes should go from base to collector)>

(also note that overdrive clippers in upstream posts were eliminated from schemo for clarity)

we are running the correct circuit test right now...

and Winston, you are right, the breadboard can go into osc depending on voltage ground leads, and weather we touch certain parts with our hands (which actually stops the osc) even hanging an alligator clip off the pot body can stop the osc, and we accidentlly had a wire wound 5 k pot which caused osc,

 

[CJ]

ok here is the revised test with the diodes in the right place,

not a ton of variance no on the Vce,

collector current stays pretty constant while the pot is turned,

we are getting the same numbers as the test on both channels of the actual circuit board in the receiver.

idle current goes from 0.85 ma to about 20 ma through on of the .2 ohm resistors

 

[CJ]

ok we got this thing fixed!

turned out to be the 4.7 ohm base resistors that feed the power transistors, they were all open. no visible signs of stress, overheating or anything, just open.  obviously we will replace those in the other channel.

thanks for all the help, i learned a ton on this one,!! 

the problem with this circuit is if the output stage goes funny, pin 2 on the input opamp floats around which upsets all 7 transistors in the driver circuit,  so if you try to fix the driver circuit without fixing the output section, you will run around in circles and get different voltages every day you fire it up. there are clipping transistors in the circuit but they do nothing to center that feedback node thru the 22K resistor because they are essentially cutoff.

here is the schemo on the bad channel with the feedback path in grn and the offending resistors in red. thanks again! this guy took me about a month but the next one will take ten minutes, if we ever get another one of these which i doubt. but the knowledge will apply elsewhere. those mv readings across the 4.7 ohm resistors should have triggered my brain but the receptors must have been busy. 

 

[JohnRoberts]

4.7 ohm is a common value for fusible resistors that are designed to open under heavy load.. One thing I liked about old ROHM carbon film is they could turn red hot and when they cooled off they still worked.

I have used the fusible resistors,  I think they were called "flame proof" or something like that.

JR

 

[Winston OBoogie]

That you built a breadboad to better understand the thing, and persevered for a month to fix it says a lot.  You have my respect.

those mv readings across the 4.7 ohm resistors should have triggered my brain but the receptors must have been busy. 

Stuff like that is always easier to see after the fact 

Going by PRR's explanation of that Vj adder, seems like a decently healthy amount of bias was possible for the OP transistors.  Where did it end up?  Anything close the 118mA 'ideal'? 

 

[CJ]

thanks for the kind words John, I am glad you are back!

here is another multiplier used in a Sansui,

looks like the draftsman forgot a junction point (blue arrow)

i thought that cap across the transistor would short out the signal, but both sides are moving in the same direction so i guess not, they discuss that cap in those links upstream,

 

[Winston OBoogie]

Yep definitely a junction missing for that miller comp.  22pF

Is this the next one to get the C.J. treatment?
It's definitely a more refined circuit.

 

[CJ]

no, thankfully i am done with that one.

btw, the current spread for the trim pot averages about 1.5 ma to 20 ma,  when everything is working that is,!

 

[JohnRoberts]

thanks for the kind words John, I am glad you are back!

here is another multiplier used in a Sansui,

looks like the draftsman forgot a junction point (blue arrow)

Part of the drama figuring out old schematics is finding the mistakes/omissions. I suspect most of these are innocent but who knows? 

i thought that cap across the transistor would short out the signal, but both sides are moving in the same direction so i guess not, they discuss that cap in those links upstream,

Caps across the Vbe multiplier stiffen up the base drive to both upper and lower output devices (AC short). You want low impedance in both directions to not only turn on devices quickly, but turn them off quickly again to prevent mutual conduction (when both outputs are turned on pulling in opposite directions). A little bit lowers distortion, too much releases the magic smoke.

JR