Green pre troubleshooting

I did a 2ch version. One side is really quiet. The other side has a much higher noise floor. It's not hum, it's like white noise. The phase invert switch seems to make it louder. It is there even when there are no mics plugged in. Don't know where to start looking for the problem. Any ideas?


imageshack.us

I switched around the opamps and the same result. One channel has the same noise level, regardless of the phase switch position. The other has it much louder in one position and nearly dead quiet (!) in the other. I tried both dynamic and condenser mics. Why would the + and - sides sound different? CMRR? Opamp biasing?

Where in the circuit is the phase switch?

at the output.

The switch on mine is at the output. I listened to the + and - before the switch and it seems like the noise is there on the "-" side. I guess I'll try to trace the signal.

Is there a good point in the circuit to check for noise? I'd like to avoid going through the whole PCB looking for the problem, but I'll learn the hard way if I have to. I wish I had noticed it before I got the thing all mounted in the case. I made my own PCBs from the version on Peter C's
site, which I guess is V1.

Since you have a good one and a bad one, and it is quick, power up and probe the DC voltages everywhere, comparing good channel to bad channel. You don't have to know what you are probing, so long as it is the same point both sides and you get the "same" voltage.

There are some AC-only faults that could give improper operation and excess noise without screwing-up the DC voltages, but DC is easier to check, and any significant difference is a bad sign and a good clue.

Some places, a tenth-volt difference between channels might be "normal". I'd first look for large differences, big enough so I didn't have to squint at the meter to see "different!".

What happens if you short out the inputs?

Try shorting the phase switch.

Take out the 6k8 phantom resistors to make sure there is no interaction with the 48v supply.

I've had noisy channels with bad input zener diodes which in turn made the first NE5532 noisy. There may have been a big spike that took out the zeners, but I replaced them & the IC all has been fine since.

Peter

Follow your nose. I smelled something cooking. The NE5532 "IC1" is blazing hot. None of the others get close to this warm. I'm going to check the voltages on that one right now.

DC voltages on IC1 on the good channel are:
Pin 1: 356 mV
Pin 2: 11.65 V
Pin 3: 11.59 V
Pin 4: -15.03 V
Pin 5: 11.59 V
Pin 6: 11.59 V
Pin 7: -0.798 V
Pin 8: 14.74 V

DC voltages on IC1 on the bad channel (hot opamp!)
Pin 1: -12.73 V
Pin 2: -0.90 V
Pin 3: -11.85 V
Pin 4: -14.07 V
Pin 5: -11.85 V
Pin 6: -10.96 V
Pin 7: -12.96 V
Pin 8: 13.92 V

Definitely not the same. Going to check PSU connections and swap opamps. Any other thoughts?

Ahh, Peter. There's no substitute for experience. I found a "two-way" zener diode at the input. I'll replace it and report back. Do you think that fried the opamp? It got pretty hot, but it wasn't on long. Thanks for the guidance.

Yikes, now I found another bad zener on the other board, the one I thought was quieter. That's two of eight zapped or defective. Then I put 4 good zeners on the "good" board and all hell has broken loose. White noise on one side like before, higher gain positions on the rotary switch totally distorted, and all the zeners still read good. I'm out of steam for today. Just when I thought I had it fixed, I have made it worse!

Hello again. The results are in. Zeners and 5532s replaced and we have a winner! Looks like all is normal now. Thanks to Peter for the clue.

I spoke too soon. Still one good channel. Constant loud white noise on the bad channel. Phase switch now does nothing to change it. It's passing audio, but with lots of noise. I'm running out of ideas.....

Hey Man,

I know this might not be what you want to hear, but check yours tracks!

If you remember on my one channel I was having problems with my Phantom Power and it was fixed, then it wasn't, then it was... anyways long story short it was a bad track... I had to resolder it and now I have no problems, couldn't really see it with the naked eye, but at a point my 48V dropped to 5V, bad track.

As far as noise, I was getting a white noise on my channels and it seemed to be that it was grounding issue, which also induced a noisy ground hum. Also is the noise there on dynamic and condenser mics? I only seemed to get it when the phantom was one. Have you grounded the pres and the XLR inputs to the AC ground?

Fear not, I am sure we will all try and troubleshoot with you until it is right. Re the version the one on Peter's site is V14.

Cheers

Matt

Another suggestion, it may sound basic but things happen... Are you sure you have orientated the NE5532's the right way around? I think you would know if you didn't... they often tend to blow up... so I am told, but you may have fried one and it is still passing power... Just a thought.

Matt

OK, here is the latest. The voltages on the IC's are off I think. With no NE5532s inserted, I get 4.84v on Pin 1 of IC1 and 9.97v on Pin 2. These should be Ov and 14.8v respectively, so tracing back from there I get to the Q1 transistor. I'm hoping this is the culprit.

Not the culprit. So why do I have a voltage on Pin 1 of IC1? Thats the output of opamp 1a, so the voltage has to be coming from somewhere else since the 5532 is out of the socket. Doesn't it have to be coming from the Q1 emitter? How is it getting there? Why do the voltages on Pin 1 and 2 add up to 14.8v, which is what I should be getting on Pin 2 only?

It looks like I used 2n5210's instead of bc550b's as Q1 and Q2. Is this not a good substitute?

Not a good substitute especially if you assume the leads are the same. According to the spec sheets the collector and emitter are reversed, which I did not take note of. I haven't fixed the boards yet, but I'll let y'all know if it works. I hope all of this posting replys to myself is entertaining and helpful to others trying to build this stuff.