Help me fix API-clone preamp problem

I bought an OSA API clone off eBay, and it had a REALLY rough shipment. This is a very old OSA model, which uses a Jensen input trafo, 2520 opamp, and original API output trafo. So it's basically an exact API clone.

I've basically resoldered every lead and joint in the whole card but I'm still having a problem.

Essentially, there is no volume change throughout the first 80% of the gain pot's range. It is a pretty static 30db of gain. This is not the pot's fault - I tried replacing it with another one, testing it with a multimeter, and most importantly, disconnecting it altogether with the volume leads not connected in any way... and I STILL got that constant 30db of gain.
The highest 20% of the pot's range exhibits some change, but, when you crank the gain pot to full, or simply short the two wire leads connected to it (zero resistance), the preamp just puts out this big "pop!" then refuses to let ANY sound through at all, except when you hit the mic or make a really loud sound into it, then it will just POP momentarily then fall totally silent again.

What could be going on here?

Also I should note that I tried an alternate wiring configuration for the gain pot, by shorting pins 2&3, and it behaved better... it didn't POP, I got more gain... but only for a few seconds! At maximum gain, it was really loud for a few seconds then gradually turned itself down to that 30db floor again.
Also this didn't fix the 30db problem in the first 80% of the range, either.

when you have all the action happening in the top 30% of the throw of a pot, that usually denotes in my experience an audio taper pot in place where a linear pot should be. I know that circuit as standard wants a 25K reverse log pot in the feedback loop, but not knowing what the OSA exactly is, its hard to say wether or not you should have a linear or log pot there. Have you tried a linear pot and gotten the same results? Regardless seems like something more than that is going on, but perhaps this will be of some help.

dave

It's not that there is LESS difference in volume in the first 80%, it's that there is NO difference, whatsoever. Not even a single decible. And like I said, with the wires to the pot completely disconnected, I get the same amount of gain as anywhere within that first 80%.

what is the value of the pot?

have you tried a linear pot?

what is the value of the feedback resistor in series with the pot?

Ive been in that scenario with a log pot that was out of the range for what the feedback loop of the circuit I was working on wanted. But like I said, not knowing what the OSA designers did, who knows. Sounds like theres more to it than that.

dave

[quote author="soundguy"]what is the value of the pot?[/quote]
22K linear, with a fixed resistor in parallel to make it "reverse log"

have you tried a linear pot?

Click to expand...

Yup. I've been running without the parallel resistor for a while now. It makes things worse.

what is the value of the feedback resistor in series with the pot?

Click to expand...

20k

Here's the deal... I "fixed" the problem... sort of. There is a 250uf DC-blocking cap in series to ground from the feedback loop. My symptoms were definitely acting like a cap problem. I took out this cap and simply jumpered the points, making the loop connect directly to ground.
Now, there is no more "popping" and I can maintain maximum gain for as long as I wish.

I have more questions, though.
First of all, I tested this 250uf cap... and it charges fine. I'll try replacing it with a new one ASAP, but I have a feeling it's gonna put me back where I started. Any thoughts on why this could be happening but it tests OK?
Secondly, will running without this DC-blocking cap have any adverse effect, aside from pot noise?
Third, the MINIMUM gain is still rather high. Perhaps I've been mistaken all along - could this just be how API preamps generally are? Quite loud unless you engage the 20db pad? Because, even when the pad is engaged, and the gain pot is at zero, there's still a decent level coming out just talking into an SM57.

Also, I would appreciate it if someone could detail how exactly the pot should be wired on this preamp. Both of the wires were torn off upon arrival and I've just been running them on CCW and Wiper.

The new DC blocking cap arrived in the mail today and I installed it.

I'm pretty much back to where I started, except the syptoms are slightly different now. There's still absolutely no gain change in the first majority of the pot's range. Actually I understated just how big that range was last time. It's more like for 95% of the range, there is no volume change. The next 4% gives me a pretty big boost in gain, but then when I turn the pot to absolute maximum 1% (or short the two wires going to it), the gain DECREASES dramatically and everything gets really distorted. If, while in this state, I bang the mic very very loudly, the signal will completely cut out until I turn the pot back down and BANG on the mic again. Weird stuff. This doesn't happen when the capacitor terminals are shorted, removing it from the signal path.

I also tried reversing the capactor's polarity, but that seems to sort of reverse the symptoms as well - I get no signal at all until the gain pot is at maximum.

I don't understand this at all. What the heck is happening?

We are going to need a circuit for this one.

Schematic:

http://www.waltzingbear.com/Schematics/API/API_312.htm


It's exactly the same as this, except with an added DI, Phase switch, -20db pad, and Phantom.

have you tried a different opamp in there? Are you certain that the problem is in the outer circuit and not in the opamp itself?

hate to be the voice of doom, but you did get it from ebay...

do a search for "1731" and see if you can order up a pcb for yourself and put one of those amps together and stick it in there and see if the 2520 is bad. If its the opamp, api will sell you a new one if you send in the broken one.

dave

I guess the thing that makes me think it's not the opamp is that the problem goes away when I bridge the capacitor connection. Also these opamps are made by stuffing a bunch of actual components in epoxy, right? Pretty much impossible to mechanically damage that.

I do have some OPA627's lying around so unless someone else agrees with my logic here, I suppose I'll give one a try in its stead.

Pretty much impossible to mechanically damage that.

Click to expand...

They do die, though. I have one lying here to prove it.

Good luck - sounds knotty.

:thumb:

they definitely die, just like any other opamp. I have a dead 2520, one of the freaking pins fell out...

dave

You need that cap in there, otherwise, you are essentially putting the bases of the two input transistors at ground. (one thru the pri of the input, which is about 1.5 kohms resistance.
This will mis-bias the input transistors.

This is going to mess up your gain pot.
I have a schematic with a 1k lin penciled in. Don't know if this is right or not.
Leave the cap and try a lower value pot.
The manual says 40 db of difference with the trim adjustment.
Starting from where, I do not now. Unity? 20 db?
Might depend on the model.
There ought to be some API gurus that can do better than I.
cj

Something very wrong, and I bet you can see it with a DC meter. Get some DC voltage readings, both when working and when not working.

Where should I be taking DC readings from?
I did some readings at the output of the opamp, as someone suggested to do on another forum (gearslutz.com) but it measures zero. The only thing that shows up is about 5V AC when the gain is at max.

> Where should I be taking DC readings from?

Everywhere. You said it had a rough trip. Are you sure the supply power really gets to the opamp pins? The +In, -In, and Out pins should sit near zero volts DC, but 10 or 100 milliVolts may (or may not) be significant. And don't believe that both ends of a PC board trace are connected: if it has been treated roughly, there may be hairline cracks breaking the trace.

It really should be a simple diagnosis: it is a simple elegant circuit. But I can't make sense of it, which is why I'm wondering if there may be cracked connections that you still have not found, despite all your efforts.

OK, I went nuts checking voltages everywhere. The Power inputs were the ONLY pins on the opamp showing ANY DC in any gain state... EXCEPT for the pin labelled "T" on the opamp, which showed a constant +15VDC. Is this normal?

The only other thing out of the ordinary that I measured was on the output pin. An AC voltage of 5V appeared the instant the gain pot was cranked (or the leads shorted) and that POP noise occured. Then this completely dissapeared to 0V (up to 0.1VAC when a signal's coming out) as soon as the pot was turned down even slightly, indicating that the opamp acts normal as long as there's SOME amount of resistance in the feedback loop. We're talking even like 50ohms as being enough here.
Any ideas as to what's going on?

Have you tried swapping with a good, known, working opamp?

That opamp is just like a little black box. You don't know if internal components have been damaged. You're just getting readings from it's input/output pins.

Aha! Now HERE's a funny thing... the Output pin is shorting to ground (did a continuity check on tons of traces)!

The only relevant place it goes to is the Phase Reverse switch, where ALL SIX PINS are short to ground, somehow! I can't tell if the problem is in the switch itself or one of the preceding traces, though it's probably the switch because they're all shorted together and to ground regardless of whether the switch is in or out. I suppose I'll try getting rid of the switch, though I don't know how to test when it's out, because I don't know which pins the switch controls to short.
Am I right in thinking this is the culprit of all my problems?