[cmscott6]

I built a meta42 and am having problems with what I assume to be oscillation: there is no output until I turn the volume up about 2/3, then the signal is completely distorted to unintelligibility. I tried a cap at c5, but this had no effect at all. Is this typically a gain problem? Should I try bypassing the opamp power pins? The stacked buffers are slightly warm, but nothing that seems excessive. I've been over the circuit board with a magnifying glass and can't seem to find any shorts or bridged traces.
Any troubleshooting pointers?
Thank you!

 

[Voodoochile]

Hi there- Are you using a panasonic pot? They are sometimes known to be buggy.

You have checked your soldering all over for bum joints and bridges between the pads, that's good. What are you powering the amp with?

I don't think you have an oscillation problem, offhand.
Are the cascodes installed already?
What OPA are you using?
What is the gain set to?

Sorry to bombard you with these questions, but we will need a bit more to chew on. Your buffers are stacked already?

If you have any extra, try installing a single in each location. If you have no other buffers, you can remove them and jumper the signal in to the signal out. If you do this, you need to be very careful that you are not getting your pinout confused, as the consequences are not good.

Also, searching will turn up a TON of threads about debugging metas... something to peruse in the meantime.

Hang in there! It was not too long ago that someone was on here with a meta that was fine at first, then he washed the board with alcohol to remove the flux. When he tried it again, nothing. After a day or two of removing his hair one handful at a time (I assume that part), he realized he was using a different wallwart for power. Evidently it looked about the same. It is often something that seems fairly obvious in hindsight. Like using a 4.7K resistor where a 47K goes, etc.

 

[cmscott6]

Thanks for your response! I'm powering the meta with a regulated dual supply: +/- 15v. The op-amp is an opa2132pa. Originally I had the cascodes installed, but I removed them (thereby destroying them ) thinking that it might have something to do with the problem. No change. The gain is 100x inner loop and 10x outer loop - I'm using Grado's, but I do want to add the crossfeed, so I figured I'd start there.
I am using the pcb mount Panasonic pot - I tore that out as well, but it tests ok, so I put it back. Is it possible that it tests fine, but still has a problem? It's my only 50K pot. Could I try a 100K? I have a nice Noble sitting around...
The last thing I could think of is that somehow I mixed up a resistor while stuffing the board, but I was pretty careful while populating: R3 220R; R4 2K; R5 1.2K; R6 120K.
I know these are pretty elementary problems, and I did peruse the posts on meta's - there's some great info, but nothing that seemed to help this problem...

I'll try removing the buffer stacks this evening - those stacked four are my only ones...
Thanks Voodoochile!

 

[Voodoochile]

So if you have a dual supply, then you omitted the TLE and the vground buffer. Two less things to look at. The 50k pot is a good choice. Is R2 500K? The schematic shows 10k pot and 100k R2. With a 50k pot you want to have a 499K minimum R2. Up to 1meg should be OK.

Also, this is not your problem, but if you remove your resistors to test the values, you migh consider briging the gain down a bit while you are at it. That is going to be wickedly loud with Grados! Maybe consider 4 or 5. I hate to complicate the matter.

What OPA are you using, also? edit: nevermind! I see. Should be fine with the BB OPA. Do not install an AD8620 with that 30v PSU, though, it maxes out at 26v, and is happiest at 24v

 

[stereth]

Go over the board again, with a multimeter instead of a magnifying glass. It's hard to believe how small a solder bridge can be.

Check ground connections - put one lead of a DMM on ground and touch the other to every component that should be connected to ground. I had horrible distortion on my multiloop amp whenever I touched the volume control - turns out I had a bad solder connection on input ground.

 

[tangent]

What's in R7-R9?

Can you trace a test tone through the amp to find out where it stops?

Take out the chips and test the voltages at the pins relative to the ground of your power supply.

Is the PS ground hooked to the ground ring on the board, are the 2001G and the TLE removed?

 

[cmscott6]

Thanks folks. R2 is 470K; should I up this greater than 500K or is this close enough? R7 is 100R; R7 and R8 are unpopulated and jumpered.

PS ground is connected to ground loop; TLE and 2001G are not populated. I will check voltages tonight.

 

[cmscott6]

All right. I tested all the voltages and found out that I had connected the positive ps in lead to the wrong positive pad. Somehow I knew it was something incredibly silly like that.

I resoldered and all sounds great and very loud. I think I will drop the gain, now. Then I'll see if I can salvage the cascodes and wire the crossfeed.

Sorry for all the trouble, folks, but thank you very much for all of your help!

 

[Voodoochile]

See what I mean? I hate it when that happens!

The 470K is close enough. If it is socketed, and you have the other value, I'd still change it.

What do you think of gain 10 with efficient headphones?

You might even try gain of 3 or 3.5. I use senn 580s, and even with them gain of 6.5 to 7 is plenty powerful enough.

edit: Don't be shy about sharing a picture, either!