Q13 is starved. All transistors except for Q13 read good for VBE. Q13 measures good with a diode test so it is just that it isn't getting enough voltage to turn it on. Spice says I should have 6.23V on the bases of Q2 and Q4 but I have 40.6V there. Something is not right but I can't find it. The whole front end is not pulling enough current. Spice says the front end pulls 25mA but I only have 13mA.
R5=0.1V
R9=1.2V
R12=0.017V
Those voltages are all on the schematic. You just need to do the subtraction.
R9=1.2V
R12=0.017V
Those voltages are all on the schematic. You just need to do the subtraction.
It's usually a waste of time to subtract voltages off a schematic, because the voltages are usually wandering so much it doesn't really tell me anything.
Yes I know but I checked them as I went to make sure they were accurate. I can actually adjust the output to zero. However, I have zero voltage across the output emitter resistors.
It's looking like your VAS current is a little low, but should run. Something is off with Q5 though. It's not passing any current, throwing everything off balance. I'm suspecting Q4 may be bad.
Q4 has a vbe of .612V. It measures perfectly with a diode test as well. BTW both channels are acting the same way. Q4 is fed by R11 which drops 1.2V, the same as R4 which feeds Q2. It looks like Q3 is not on. -48V on the emitter and -8.3V on the collector. Supposed to be -15V. I'll go through again and measure vbe on each transistor and mark up a schematic. The circuit works in spice so it should work in real life.
Yes Q4 looks fine by your voltage readings. R4 and R11 actually feed Q7 and Q8. The input stage steals current from them. That's how it operates. If everything's in balance, R5 and R12 should be passing the same amount of current. This will be the amount Q2 and Q4 are taking from the VAS section. R12 is only passing 0.7mA. If everything was working properly it should be passing around 7mA by you sim file. Most of Valery's input stages are passing 3mA, so something may be off there. It's possible this is all off because you have no bias current in the output to make it operate though. What is your collector to emitter voltage on Q11?
RE: the LT Spice file
FIRST, i'll confess upfront I didn't bother to load the correct transistor models in; i just took the defaults and fired up LT Spice for a VERY quick look.
i see 20V (~48v - ~28v) across 22ohms (R20).
Right away, that sounds like a bit much (~1A!).😱
And a 12v difference between Q7 & Q8 emitters with no signal applied.
Makes me a little leery about some of the other values.
Of course, some of this may clean up when i use the right models, but this rings enough bells to make me say "hmmmm..."
mlloyd1
FIRST, i'll confess upfront I didn't bother to load the correct transistor models in; i just took the defaults and fired up LT Spice for a VERY quick look.
i see 20V (~48v - ~28v) across 22ohms (R20).
Right away, that sounds like a bit much (~1A!).😱
And a 12v difference between Q7 & Q8 emitters with no signal applied.
Makes me a little leery about some of the other values.
Of course, some of this may clean up when i use the right models, but this rings enough bells to make me say "hmmmm..."
mlloyd1
Attachments
Something's wrong with your sim. I'm reading 25mA on R20. Terry had a sine wave running on the input of the file.
It looks like something is wrong with Vbe spreader.
We've got almost 5mA through the VAS, which is fine, but too low voltage across the spreader.
We've got almost 5mA through the VAS, which is fine, but too low voltage across the spreader.
Terry, if R25 is 910R now - try to decrease it gradually - 820R, 750R, 680R...
At some point Q11 should start closing - its Vce will start growing, increasing the bias.
You should see around 3.5V between collector and emitter of Q11.
At some point Q11 should start closing - its Vce will start growing, increasing the bias.
You should see around 3.5V between collector and emitter of Q11.
Last edited:
I guess i have to wear my donkey ears for a few days. Seems I checked everything except the 100R trimmer. I put a trimmer in parallel with R25 so I could reduce it. The only way I could get it to do anything was with R22 set to zero. That's what finally got me to pull the trimmer and measure it. Doh, 100k. Serves me right for working on too many different amps at the same time. Anyway, It's working now. What bias do you suggest? R 25 needs 910R. 1k won't get it close enough.
Thanks, Terry
Thanks, Terry
Hi Terry,
If you increase C7 value from 2.2pF to 10pF - that will eliminate the overshoot.
Cheers,
Valery