Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › M3 supply rail Troubleshooting
New Posts  All Forums:Forum Nav:

M3 supply rail Troubleshooting

post #1 of 12
10/16/09 at 11:31pm
Thread Starter 
I am trying to work the bugs out of my M3. I recently finished my sigma 11, which puts out a steady 23.5V. The M3 however, only measures 1.15V when hooked up.

I get 0.024V across pins 4 and 7, and +0.011V, and -0.011V from the opamp + and minus rails. So it looks like the rail splitter is not to blame.

I suppose it is then the fault of one of Q5+/-, Q6+/-?? I measured these and
Q5+ gives 0.61, 0.61, 0.29V (c,b,e)to the minus rail.
Q6+ gives 0.61, 0.61, 0.61V to the minus rail.
Q6- gives 0.63, 0.63, 0.63V to the positive rail
Q5-gives 0.63, 0.63, 0.35V (c,b,e)to the positive rail

I am not so sure what this tells me. What should I look for when scrutinizing BJT's, and JFETS? Perhaps I am on the wrong track altogether and it is something else entirely.

Thanks for any help
serge
Reply
post #2 of 12
10/17/09 at 1:03am
Serge - Since your voltages are equal, sounds like Q5+ and/or Q5- are blown. I had the same issue a few weeks back with an M3 build - replaced both and voltages came in-line.

Let me know what you find out.
Reply
post #3 of 12
10/17/09 at 1:24am
Make sure you didn't use Fairchild 2N3904Cxx for Q5+ and Q2[LRG]. The C-suffix part has the wrong pin-out.
Reply
post #4 of 12
10/17/09 at 6:57pm
Thread Starter 
Thanks for the replies.

I replaced Q5+ and Q5-, which btw were faichild parts with -k and -n(?)suffixes. they read 3904TAR and 3906TAR respectively.

The voltage now is 4.9V. So a definite improvement. I am wondering if Q6 also needs to be replaced or if these are just improved over the previous, clearly broken, parts.I have spare Q5's, but not Q6's on hand.
Reply
post #5 of 12
10/17/09 at 6:59pm
Thread Starter 
I forgot to get individual readings, I can if it will help diagnose. I , vote replace Q6 +/-, what do you guys think?
Reply
post #6 of 12
10/17/09 at 7:09pm
Quote:
Originally Posted by sergery View Post
I replaced Q5+ and Q5-, which btw were faichild parts with -k and -n(?)suffixes. they read 3904TAR and 3906TAR respectively.
Those are fine.

Quote:
The voltage now is 4.9V. So a definite improvement. I am wondering if Q6 also needs to be replaced or if these are just improved over the previous, clearly broken, parts.I have spare Q5's, but not Q6's on hand.
The Q6s rarely fail.

Do you have C4 mounted in the correct orientation?

Power off and let all caps discharge fully. Then measure your Q5+ with your DMM in diode-check mode. Put the Red lead on the B (center) pin, and black lead on the C and E pins, your meter should read about 0.6V. Reverse the probes and they should all read "infinity" (OL" or "1") depending on your meter.

Repeat the above for Q5-, except the results will be opposite polarity (black lead on B, red lead on C or E should give 0.6V).

If either test does not pass, then that transistor is blown.
Reply
post #7 of 12
10/18/09 at 12:19am
Thread Starter 
So on Q5+ with the red probe on b, I measure:
to e: 0.67V
to c: 0.15V
reversing probes I get:
to e: 2.1V (and keeps rising)
to c: 0.18V
So on Q5- with the Black probe on b, I measure:
to c: 0.17V
to e: 0.63V
reversing probes I get:
to c: 0.17V
to e: 1.925V (and keeps rising)

So it looks like I managed to burn another pair. Weird, I took them directly out of the static package with static safe tweezers - no hands at all, soldered with a grounded soldering iron, applied max 2s heat to each leg. I wonder if there is something else I am doing wrong. Is there such a thing as ESD clippers? anyway I will try a new set later.
Reply
post #8 of 12
10/18/09 at 3:40am
Did you remember to discharge all caps first before working on the board? Any stored charge may damage parts if certain parts of the circuits touch and make contact (such as by tools, soldering iron, or other objects).
Reply
post #9 of 12
10/18/09 at 8:26am
Thread Starter 
I shorted the output, and then checked at the output and at the test points near the back of the board that voltage was 0 before I did the diode testing (and replacing the BJT's this second time). I tried replacing Q5+/-. It seems that these parts are not to blame though because I can insert them into the circuit and test them - I got the same values, and then remove and test again, they test fine, ~0.68V or 'OL'.

I have looked over all the electrolytics, they are properly oriented. The joints look good under close inspection, I will check again though you never know.
Reply
post #10 of 12
10/18/09 at 8:44am
Thread Starter 
I wonder if C4 was damaged when running with damaged Q5+/-? Or perhaps D1 if there was a surge? I suppose I can use this diode check function for the, well, diode. Forgive my ignorance, Is there a simple way to check capacitors in a circuit?
Reply
post #11 of 12
10/18/09 at 11:06am
Quote:
Originally Posted by sergery View Post
I shorted the output
Shorted what output? That doesn't sound like something you should do...

Quote:
I wonder if C4 was damaged when running with damaged Q5+/-? Or perhaps D1 if there was a surge?
Not likely.

Quote:
Is there a simple way to check capacitors in a circuit?
Depends on capacitance ane where it is in the circuit. But in general the only reliable thing you could test is for a unwanted short circuit. For large capacitances that have a charge that doesn't get drained by the circuit quickly, you should be able to measure the voltage across it slowly drop after the power is removed. Electrolytic caps don't usually fail without some visible signs, lsuch as bulging, leakage, or downright explosion.
Reply
post #12 of 12
10/18/09 at 8:42pm
Thread Starter 
Quote:
Originally Posted by amb View Post
Shorted what output? That doesn't sound like something you should do...
Thanks for pointing this out. The headphone jack. I didn't think there was any concern draining all the caps at once (from 0.4V), but now that you point it out that that could still be a lot of charge in a small amount of time. I didn't use a very high value resistor either - 100ohm.
I was thinking more about resistor heat so i just used the largest size resistor I had.
Upon further thought I think I would short individually next time to keep current from passing through transistors...

Quote:
Electrolytic caps don't usually fail without some visible signs, lsuch as bulging, leakage, or downright explosion.
Well, no visible signs... I will continue to poke around with my meter/magnifier. I may have damaged some transistors though.

thanks for keeping the insight flowing
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › M3 supply rail Troubleshooting