A perplexing signal-level reading - troubleshooting query

[Mark Hammer]

I'm trying to bring an EHX Polyphase back to life for someone, and it's driving me nuts.  The unit has no bypass or phased signal output.

All op-amps tested so far show appropriate supply voltage readings.  I plug in a signal generator and I'm reading about 2v p-p at the input jack.  When I read the output pin of the first op-amp stage, the amplitude reads the same but quickly drops down to near 0v within the first few seconds.

My inclination is to look at caps when something varies over time like that.  I've already replaced the largest caps, but am disinclined to start yanking and replacing all the electros unless warranted.  Same goes for desoldering chips and replacing them. 

What might produce a drop in amplitude like I describe?  Given that there is no bypass OR effect signal, is it likely to be as simple as changing the first op-amp?

[Govmnt_Lacky]

Hey Mark,

Unfortunately... without a schematic it will be hard to identify the culprit.

Seeing as you did not post it I can only assume that...

1) You do not have it -OR-

2) You want to keep it close to the vest for some reason.

Either way, based on the limited info in your post, I would have to say go with the op amp.

Good Luck  

EDIT: Sorry MArk... Google was my friend!    I concur with earthtones.... try the op amp. Hopefully it is socketed 

[earthtonesaudio]

The signal is never good.  The gain after the first opamp should be four or something, not unity.  If it drops after that it may be another problem or the same.  If the circuit has no obvious physical defects like a broken trace or blown cap I would say replacing the opamp is a good use of your time.

[Quackzed]

if that voltage divider setup after the input cap isnt working right it could be pulling the input too high or too low... 

[Mark Hammer]

Thanks, folks.  I'll try injecting the signal a few other places downstream from the input, and see if I get output from that.  But for the moment, it's starting to look like that op-amp may well be the culprit.

[Gurner]

I'd measure the DC on all pins (first opamp) not just the supply voltage ...if nothing obvious amiss on any of the pins, then I'd lift the 1st opamp output (point A onwards)  therefore lift one of the legs of the two follow on resistors after point A.

That way you've isolated the opamp from anything loading it down at the output & can simply focus on the immediate 1st opamp stage.

[R.G.]

As a generic debugging process, I'd vote with Gurner - measure all the pin voltage.

For the moment, I'd first replace the two 1uF caps on the first opamp. If the one on the feedback path was high impedance, it would go to unity gain, as you describe. The other one is on the bias for the opamp, and if it goes funny, it could throw off the bias point.

My experience is that if an opamp passes signal at all, it's probably fine. Not always, but usually.