Cmoy DC offset, help needed!

[axiom]

This is my first project, so just to see if I could do it, I bought all my parts from Radio Shack (possibly part of the problem

). The only opamp they have is the TLO82. The test source is a 4th gen Ipod photo from the headphone out (so I have volume control).
I have put everything together according to tangent's website, but I have a few questions.

1) When I measure the DC offset with no source (from Lout to ground) I get .0430 V, and Rout to ground is .0068 V. The Rout gets lower the more 9V I hook up in series, but the Lout doesn't change. Is it just me or is the Lout DC offset a little high?

2) Possibly another symptom of the same problem, if I power the amp using one 9V after about a second or two the left channel develops a lot of distortion and crackling. This goes away with 2 or more 9Vs.

I checked all of the solder joints and it looks like they are fine, and I made absolutely sure that I hooked everything up right.
When I hooked up some cheap phones anyway just to test it (

) it actually works , but I am a little nervous about hooking up my Beyer 880s with a possibly dangerous DC offset... If someone needs more measurements, just let me know and I'll be happy to oblige!
Hope someone can help!

 

[nikongod]

have you checked for obvious shorts? any "less obvious" shorts? any stuff that should not have been connected that was? i once saw a cmoy with the input perfectly soldered to the output... it was hard to troubleshoot as the thing was built beautyfully, and it was a DELIBERATE joint.

after that, do you have a different opamp to drop in? if the circuit checks out (whcih based on one good and one bad channel it probably will) you may have a bad opamp.

 

[Pars]

Have you gone thru all the measurements on Tangent's site (as well as his troubleshooting article)? What does your voltage split at the opamp look like (V+ to vground and V- to vground)? Does it change if you plug some phones in? Does it change with volume setting?

The noise in the one channel is bothersome, as is the 43mV of offset.

A couple more things you can check:

1) Are the electrolytic caps in with correct polarity?

2) With the opamp out and the amp not on, ohm out each pin of the opamp socket to vgnd. Check the resistance reading you get against the schematic. For example, the input pins (3 and 5 IIRC from memory) have a 100K resistor to ground. If you see something else on your meter, figure out what/why and fix it. The power pins (V+ and V-) should show the electrolytic caps charging (low initial reading rising to infinite or very high... M ohms). If you see a steady resistance here, something is wrong with the cap connections. You can also ohm adjacent pins out... you should not see any real low (<100 ohms) readings here. Also double check the input and output jacks and the pot (if you have one... probably not if this is all from ratshack).

If everything checks out fine, post some closeup pics of the top and the bottom of the board. The behavior you are seeing is subtle enough that it could possibly just be the opamp, but it doesn't hurt to verify everything with Mr. Meter

 

[picklgreen]

Check the DC offset agin, this time with a source! Dont plug in any headphons but plug your ipod into the input so the CMoy is actually doing something.

 

[axiom]

Thanks for the help so far! What a fast response!

I don't have another opamp and RS is out of them. Perhaps later today I will go to a different RS somewhere else and see if they have any.
Now I've checked just about everything I can think of to check. All connections look good and there does not appear to be a short anywhere. I checked and all the proper resistances from each pin of the opamp to ground are correct (per schematic). For power caps I am using 470uF electrolytics. When I measure the resistance from V- and V+ to ground (with power applied) they increase rapidly from about 3.35kOhm (V-) and 2.15kOhm (V+) (both after settling) to 10.05MOhm (V-) and 9.96MOhm (V+).
The voltage split from 2 9Vs is V-:9.002, V+: 9.011 from 18.06V total. Looks like I lost .05V or so, but I think this is probably not important.
I measured the DC offset under various circumstances (all in mV):
Just power (no source or hp):

L: 43
R: 8

w/line out no hp:

L: between 8 and 43 changes a lot mostly below 15 but spikes frequently to over 20
R: 7-9 pretty steady

w/hp no source

L: 40
R: 6

both hp and source

L: 8-43 from either output (varies quickly)
R: 5-6 from either output

Should I change the input caps from .22uF to 1uF? or should I add output caps? Perhaps the left channel input cap is messed up?
The thing that makes me the most confused is the good performance of the Right channel compared to the rapidly changing offset of the left...
Hopefully this makes sense everyone can help me some more!
Thanks again!
(pictures to come tonight)

 

[Pars]

Probably just the opamp...

 

[FallenAngel]

Quote:

Originally Posted by Pars /img/forum/go_quote.gif Probably just the opamp...

Second that, 43mV is kind of bad, but I'd almost expect it with CMOY amp and very bad opamp.

 

[frankeleveld]

I'd first change the op-amp, try a Burr Brown OPA2132 and retest.

If that doesn't resolve the problem, check for bad joints, shorts, etc. - as others have also mentioned. I experienced a large offset with a CMoy I built recently and I couldn't put my finger on the problem until I replaced the electrolytic caps - despite their polarity was as it ought to be, they were the culprit - the problem was resolved when I put new caps in.

But I'd recommend changing that op-amp first.

 

[axiom]

I did change the opamp, and now the offsets are a much happier 20.5mV in the right channel and 16.5mV for the left. Is it common to have such a large variation between the same model of opamp (the right was 7mV before)? Also if I switched to a 2132 would the offset improve? (I know SQ would)
Thanks for everyone's help!

 

[FallenAngel]

SQ difference will be more than worth it and yes, specs are a lot better on it so you should get your DC offset a bit better. Please also don't forget that considering how the CMOY is designed, DC offset will always be there. Tangent explains it pretty nicely in this article.

 

[Mizerable]

If he had not gotten a new, working opamp, what type/size of output dc-blocking caps would he have needed?