Wierd 1176 problem, expert help needed!

I am getting my ass kicked in the worst way right now. I have a rev d 1176 with a meter tracking problem I can not fix.

Both fets are matched.

All the transistors in the sidechain and meter driver amp have been replaced, twice.

The unit biases fine.

The unit cal's at 20:1 following the setup procedure fine.

the wierdness:

with no signal present in GR, at 20:1 the needle sits at 0 dB like it is suppposed to.

At 12:1, 8:1 and 4:1 I get an off scale reading with the needle to the right, like you would expect to see with all buttons in.

With 20:1 and 4:1 in simultaneously I get a -8dB reading which decreases as you add more ratios, the oppposite of what its supposed to do.

Thats all with NO signal into the unit, just looking at GR mode.

Now, with a 1K tone at unity on an EXTERNAL meter I get

20:1 0dB out and -10dB reduction
12:1 +1 dB out and -7dB reduction
8:1 +1 dB out and -7dB reduciton
4:1 +1dB out and -5dB reduction

20:1 +4:1 -7dB out and -7dB reduction
12:1+ 8:1 +1dB out and -7dB reduction

probing around voltages against a working unit I get even more frustrated, good unit on the left, bad on the right:

Q11 D-.81v -.76
S 12.7 11.2
G -.98 -.9

q12 E -1.4 -1.3
c 12.62 10.6
-.8 -.7

Q13 E -1.4 -1.3
C 12.7 10.7
B -1.4 -1.3

voltages look pretty similar everywhere.

This unit once worked fine...

Anyone ever experience anything like this? Is there someplace I should be looking that Im not? So far that I can tell, the sidechain and meter driver amp look ok, wtf?

dave

What happens when you rotate the release control? I had one once with a leaky side chain rectifier (CR 2 or 3), when the release control was moved the GR zero would move around. This had me going for an hour or more. This was also the time I learned about low leakage diodes... a regular 1N400x was as leaky as the "bad" original diode so when I put that in as a replacement the problem continued causing me to look elsewhere for a while.

I'd also look at C19 and C20 since it seems that there is a DC bias related problem or some kind of oscillation. I hate those.

Tim

I'd have to check the diodes, what part did you wind up using for a substitution?

Forgot to add that I had replaced C19 and C20 a few times with different cap types and values.

dave

A common 1176-problem is conductive-dirt-build-up around the FET's. (that obviously depend on very high impedances). Cleaning the board area around the FET's - possibly on both sides of the pcb - can often solve these "wierd-behaviour" problems..

Jakob E.

hmm. I have both fets socketed, perhaps I should solder them in, thats not the first time Ive been cautioned about dirt around fets. My only question is if dirt would have such repeatability in regards to the behaviour at different ratios- Ive put matched pairs of both bf245a and 2n5457's in there and I get identical results after biasing for the fets.

Hopefully I can dig into this again tomorrow. wouldnt be fun if it was simple, right?

dave

A common 1176-problem is conductive-dirt-build-up around the FET's.

Click to expand...

Definitely. One of the 1176 documents has the breath test you can do - just watch the GR meter as you breathe on the FETs. I was shocked when I tried that with one of my units - the gain meter swung from zero all the way to the left. After cleaning the board with alcohol it only dropped by about 1dB. I'm not sure if just soldering them in would help much - the socket and/or board has to be really clean and that should take care of any stray current sneaking around the FET.

I soldered in the FETs, no change and really no suprise either.

I dont suspect this to be a global problem, at 20:1 everything is perfect, sounds fine, compresses fine and tracks correctly.

What could be going on where the unit works ok in one ratio but is all fucked up in others? Maybe I have a short in the ratio switchbank somehow?

If I had a bad leaky diode, wouldnt that effect all the ratios?

Anyone have any other ideas, Im about ready to have leia fetch the droids to blast off a message proclaiming my only hope. I need a crazy old wizard right about now.

dave

Have you aligned the meter null pot, R75?

Here's the procedure:
SETUP
no signal
disconnect R44
centre R75 ( null pot )
put voltmeter across R74

CALIBRATION
a ) zero meter with R71 ( thru front panel adjust )
b ) adjust R75 for 00 volts across R74
repeat a and b until both conditions are met

The following shouldn't be necessay, but as a final check:
setup unit for 10dB of compression as measured with external meter ( or use the 1176 meter in output VU mode )
select R44 for correct tracking, if R44 is changed it may be necessary to repeat above calibration steps.

Certainly if R75 is not set correctly then the meter readings for 0GR will move around with different selection of ratios

Rob

yes, thats the cal I followed. Lines up perfect for 20:1, complete mess as noted for the other ratios...

dave

bummer

first up it's not clear to me from your first post if you are getting the correct ratios on the audio

forget the meter for now
you need to set up a jig to test the ratios
ensure that as you test each ratio that you push the unit into 3dB of GR, then up the input level by exactly 10dB without touching any of the 1176 controls and measure the change in output level. For 20:1, you should see about 0.5dB increase, for 4:1 you see about 2.5dB increase etc etc.

You need to prove the sidechain / ratio circuit first before investigating the meter. Make sure you reset the unit for the 3dB of GR when testing each ratio.

R

I would check the wiring across the ratio switches. One half of the bank is a voltage divider that sets the amount of signal to the side-chain amp, and the other side is a voltage divider that sets the threshold for the side chain rectifiers, where I suspect the problem lies. When 4:1 and 20:1 are pressed together, the other two switches are effectively shorted out, and they should not cause the meter to move. The fact that it does move as you add 8:1 and 12:1 makes me wonder if the the ratio divider is maybe floating on one end - check switches to the 150ohm resistor to the release pot, and check the continuity from that point to the bias pot.

Or I could be completely wrong.

You could also try swapping diodes cr3 and cr4 from the good 1176. Leakage will definately throw things askew.

By the way, hello everyone. This is my first post, though I have been lurking for several months. This is a fantastic forum, I'm glad I stumbled across it! :grin:

ok, had some time today to get back to this. Replaced both diodes int he sidechain, still have the same problem...

Testing the ratios, I get a 3dB change at 20:1, 4dB change both for 12:1 and 8:1 and a 6 dB change at 4:1.

This is getting really annoying. I wish I had the skill set to deal with this, grr.

dave

Dave,

Have you looked at it with a scope, or just a meter? Maybe something's breaking into oscillation when changing settings?