D-87 Noise: Wind - like rustling noise from D87 and drifting bias voltage

Hi
I've built 2 u87 clones. I've used the syt5 body, cinemag 13112 (?) transformer, parts from mouser, PCBS from Dany Bouchard and capsules from microphone-parts.com. I followed chunger's and others' threads paying close attention to neat soldering,  cleaning the board as I go and wiring the transformer correctly.

They sound amazing to my ears, having adjusted the trimpot to get a drain voltage of 11.5v. I didn't use a scope but it still sounds fantastic. BUT one day one of the mics just stopped working. I suspected dodginess in the biassing department,  and measuring the drain voltage I noticed it had slid upward maybe to 14 volts. This was after I had substituted in 1% metal film resistors for the trimpot and double checked that the voltage was ok after doing this. So I took those resistors out and put the trimpot back in and re - biassed it. Again it worked but now I became aware of a rustling sound, like the sound of wind blowing on a mic without a windshield. Not very loud compared with the signal but just enough to be annoying. I checked that the capsule was clean and it looks immaculate. Meanwhile I have been noticing that the bias voltage just won't stay still. My other mic has been perfect all along and its bias voltage has stuck on 11.5v.

So I'm wondering: is the drifty bias related to the noise?  If I can stop the bias from drifting will it stop the noise? What is causing the drifty bias? Could an electrolytic capacitor be leaking? I have known leaky elwctrolytics to cause voltage changes and noises in audio circuits before. I don't think it's the zener because its voltage has been stable. I note that the tantalums are working very close to their maximum voltages so perhaps it is one of them. Has anyone else seen this? Any ideas before I go pulling everything out and testing capacitors out of the circuit...

Not quite sure if / how much it might help, but i'd like to point you to two of my posts:

http://groupdiy.com/index.php?topic=48030.msg776029#msg776029
http://groupdiy.com/index.php?topic=61079.msg773776#msg773776

And for the record, my order of trust in capacitors goes like this: film caps, japanese electrolytics, "mid-brand" electrolytics.

Anything else, especially in the signal path, i hate with a passion, from an engineering standpoint I've witnessed ceramics short out internally, i've read numerous stories of tantalums crapping out or blowing up, as well as chinese electrolytics bulging even in the signal path (in the case of the output coupling caps in one or two dbx DriveRack digital crossovers).

Sounds like a leak current in the high-Z part of the circuit, or even pollution on the capsule.

It's interesting, I have been noticing the issue for a month but only using the mic for short sessions of say half an hour at a time. Now I've put the post up, I turn the mic on and the issue has gone. I test the drain voltage and it's 12.5v. I wind the pot (because of the issue I took the fixed resistors out and replaced with a trimpot) to get 11.5v and this time it is easy and doesn't vary. Tested it again for two days in a row and it's still working. Leads me to suspect perhaps I was missing some moisture etc on the hi -Z board. It's spring in Australia and our temperature varies from 20 to 40 degree Celsius,  humidity 70 to 90 percent. Thanks for your comments. Also if it does occur again I'll post a recording of good mic vs bad mic.xpecug

Have now fixed drifting bias voltage. Discovered one of the polystyrene 470pF capacitors on the Hi Z board had the wrong voltage on it compared with my good mic. Seems it was leaking and behaving like a varying resistor. Probably I heated it too much whilst soldering in. Now bias voltage is stable and with adjustment of the variable resistor on the FETs source it is now a lovely 11.5v. HOWEVER, WIND LIKE NOISE STILL REMAINS. I guess it could be the capsule so I could just replace that with 2 47pF caps and see but when will I have the time..?

A 'dummy capsule' (one or two capacitors with the same value as the capsule capacity) is always a good test to exclude certain problems. Of course take care that the capacitors can handle the polarisation voltage! (And have a low leak current.)
The 'wind' sound still makes me think that it is caused by moisture on one or more Hi-Z points.

After a long pause I have finally got back to this mic! I tried substituting the capsule for capacitors. It was difficult to tell but I thought perhaps it was quieter. Then I swapped the capsules. Now the other mic has got the noise! So it presumably is the capsule. Looking at the "offending" capsule, there does seem to be a tiny indentation as if it is damaged. I wonder if I damaged it when I had to do a capsule swap after the initial build when I was supplied with a capsule that had a short between the two backplates. The supplier promptly sent me a replacement for that shirted capsule. I doubt he would send me another after all this time! If I have damaged the capsule is there any way it can be fixed? Or am I better off just replacing the capsule. At $129 it's not a small thing to do though it's not the most expensive capsule either.

Jonathan_D said:

After a long pause I have finally got back to this mic! I tried substituting the capsule for capacitors. It was difficult to tell but I thought perhaps it was quieter. Then I swapped the capsules. Now the other mic has got the noise! So it presumably is the capsule. Looking at the "offending" capsule, there does seem to be a tiny indentation as if it is damaged. I wonder if I damaged it when I had to do a capsule swap after the initial build when I was supplied with a capsule that had a short between the two backplates. The supplier promptly sent me a replacement for that shirted capsule. I doubt he would send me another after all this time! If I have damaged the capsule is there any way it can be fixed? Or am I better off just replacing the capsule. At $129 it's not a small thing to do though it's not the most expensive capsule either.

Click to expand...

Did you test your microphone many times talking to it from close distance and not using popfilter?
If so, probably there's too much saliva on the diaphragm and you could try to clean it.
If not there's few other options too check.
You can also connect capsule in reverse and try.
If the capsule is faulty, you can always buy exactly the same capsule directly from china for less than half of price you paid.
If you need information about where purchase it, let me know on PM.

Hi, thanks for your response. When you suggested connecting the capsule in reverse do you mean connecting the front ("vor" on the u87 schematic) diaphragm to the rear ("hinten") circuit and vice versa, or do you mean swapping the front wires around and swapping the rear wires around? And if I do one of these manoeuvres,  what would be the expected result if the capsule is damaged? Sorry if my questions appear stupid but I have to say the u87 schematic hi-Z area has always baffled me in terms of how it manages to swap capsule polarity without using a dc to dc converter. As to your suggestion about an alternative way of sourcing another capsule I would be interested in the details, my email is [email protected]
Thanks again for your interest in my problem.

Send you a link, check your email.
You need to reverse connection of diaphragm wires and corresponding backplate wires. With backplate you need unscrew connectors, with diaphragm desolder from the circuit. Side (one diaphragm/one backplate) which was previously on the front, need to go to the back.
Mostly people tests cardioid pattern more often,  spitting on the front diaphragm 
If you will swap the capsule sides and if the noise will stop - then you will know that your capsule is dirty.

Tried swapping the capsule sides and the noise stopped in cardioid pattern. Noticed the wire from the backplate of the front capsule appears frayed. I replaced the wire and now the noise is gone altogether. So my current theory is that it was the wire all along. However this problem has been somewhat intermittent so I wonder if I've really got this right!