Question about bias adjustment.

[Buckeye]

I built this amp from a kit:

MX50 Schematic-2.jpg

There were no instructions. The bias adjustment pot (R13) does absolutely nothing. I see that it's supposed to adjust the quiescent current through the emitter resistors for AB operation. However, I'm using a 200mV meter across the 5W resistors and see absolutely no variation in Voltage as I adjust the pot.

Is it possible that Q9 is biased off? Does anyone have any insights into what is going on here and why?

Thanks in advance.

[mrfeedback]

What is the base-emitter voltages on the output transistors and does it change ?.

Eric.

[MJL21193]

Are you running the amp at full rail voltage?

[Buckeye]

1) Didn't check the Vbe of output transistors, but I do know that it doesn't change. I looked at that yesterday. I just checked the Vbe of Q9. It is -0.674 and that would mean that Q9 is operating in cutoff and that's why the adjustment does nothing.

2) No, I am not operating the amp at full rail voltage. The amp is claimed to run from +/-15V to +/-50V. I am running it at +/-18V. My 50V supply has left the building and I will cobble up another to see what happens.

Would you expect the bias adjustment to be non-functional at low rail voltages?

[sregor]

It looks like there are some errors in this schematic - first in the (I assume) current mirror (q2 collector and base connections should be reversed), and some of the components and connections don't make sense. The .22 ohm resistors in the collector of the negative rail transistors don't do any thing - the use of a darlington pair for bias transistors is very unusual, separate drivers for each output is odd and would seem to be counter productive, and the diodes across the emitter resistors of the negative drivers seem like it would increase driver failure, with only possible benefit of possibly better clipping (I wouldn't know for sure). The bias circuit should work unless the other problems cause insufficient current to flow through it. Good luck with it.

[ostripper]

Originally Posted by Buckeye

1) Didn't check the Vbe of output transistors, but I do know that it doesn't change. I looked at that yesterday. I just checked the Vbe of Q9. It is -0.674 and that would mean that Q9 is operating in cutoff and that's why the adjustment does nothing.

2) No, I am not operating the amp at full rail voltage. The amp is claimed to run from +/-15V to +/-50V. I am running it at +/-18V. My 50V supply has left the building and I will cobble up another to see what happens.

Would you expect the bias adjustment to be non-functional at low rail voltages?

A stable amp w/ ccs, you should get easy bias adjustment all the way from 15 - 70v (my amp). C10/12 - 13/14 are to stabilize an unstable amp. My guess is that it is still oscillating. It has no emitter degeneration on Q1-2 /4-5 (100R ??) C4 should be gone and C7 should be 47pF.

BTW , I had 1 amp that just would not bias. It did this intermittently as it was oscillating (wrong degeneration resistor). I made this point yesterday about chinese Ebay amp kits, most are poorly implemented .

OS

[Bonsai]

The Vbe Multiplier adjustment range is 3.84 to 1.92V, with the mide point at 2.55 Volts. Since this is a double emitter follower, you cna expect a stand-off voltage somewhere in the region of 2.3-2.4V.

You should be able to bias this amp from OFF to virtually short circuit with the adjustment range.

Recheck the voltages around the circuit, and also measure the output device current and tell us what you are getting.

[magnoman]

In addition, I think Steve was pointing you in the right initial direction.
Its likely that Q2 and Q4 were supposed to form an active current mirror load.
Bases to bases then the collectors to the dif input collectors, Q4 collector tied back to its base, emitters to ground ... I think you recognize it.

Hope this helps
-Antonio

[Buckeye]

That was quite a variety of replies... I never expected it. Thank you all. It took a while to digest what was said. Plus I also cobbled up a +/-50V supply for more testing. I was able to get about 130W out of this amp at clipping into a 4-Ohm resistive load. I'll get more specific info later today.

Good eye, Sregor. The schematic was in fact wrong in (at least) two places. The minus side current mirror was drawn wrong as you said. The darlington bias pair was also wrong. I corrected the schematic and attached it below. The corrected schematic clearly shows the cause of the non-functional bias adjustment. The emitters of the bias transistor pair is connected to nothing but themselves and a bypass cap. See here the corrected schematic:

MX50 Schematic-3.jpg

So maybe I should add a jumper to the top of R17?

As for the Q13 & 17 collector resistors, they are used to sum the two collector currents. I've seen this done either in the collector and emitter path. Obviously the emitter makes more sense because it includes the collector current.

ostripper, this amp is not unstable and does not oscillate. It had that problem but I fixed it by adding some capacitors to the rails. (I posted another thread about that.) Can't yet comment on the frequency response but flipping through the frequencies on my function generator and watching the scope, it seems pretty flat. Maybe I'll mess with the filter caps later. Yeah, for the most part ebay stuff is sometimes hit or miss... Chinese or not.

Although the bias adjustment is non-functional, the amp is not terribly biased. I'll get some numbers on that in a bit. At clipping I get around 130W, but the plus rail clips a little before the minus rail.

I'll also get the voltage range at R13 wiper.

Thanks again.

[Buckeye]

The last schematic I posted for the MX50 appears to be correct and is how the board is wired. It works. I'm a bit embarrassed because the problem was the 1k pot. Apparently one of the pots got run off the end. Bourns quality it ain't.

When I checked the 2nd amp. it was in a range where adjustment had very little effect. With a bit more patience, the adjustment worked. I fixed the bad pot by pulling the adjuster up while turning clockwise. After the internal worm gear engaged, the potentiometer adjuster worked fine.

Using a +/-48V power supply, I got 138W at clipping. Because of my lousy heat sink, the transistors got hot, fast.

All is good.