Neumann U87ai needs repair and testing

[Spencerleehorton]

Hi,

I have a Neumann U87ai and it had a problem with corrosion to the capsule and apart from making bubble and popping noises was completely dead.
I have first replaced capsule and it did work but noises persisted.
I then got the ISP out and gave all the pcbs a clean and noticed that the XLR connections were not working correctly once I had cleaned everything. I then replace XLR connections and after another clean the mic seemed to at least have correct voltages going in and noises seem now to have stopped but capsule doesn't seem to work?
When I switch between Omni and fig 8 for a split second the mic works perfectly then stops again?
Tapping the mic on the body and having the gain up I can hear the tapping.
I have checked capsule connections and all seems fine.
I have swapped out both of the 470pf caps and nothing seems to change.
I'm not sure whether I have any spare 1G resistors to check to see if they are the problem but I'm looking for other options of how to get this working as I'm starting to run out of ideas.
I was going to put in a sine wave from the capsule connections just to see what I get out.
I have a signal gen, scope, anyone got any other ideas of how to move forward?
Pin 2 and 3 = 45.5v
R10 = 53.6v
R13 = -53.7v
D101 = 31.77v
T1 = gate = 0v, source =1.719v, drain=1.868v
T2 = 42.8v, 44.6v, 42.7v
T9 = 26.20v, 29.38v, 25.60v
T10 = 9.77v, 0v, 25.54v
(Base, collector, emitter)
D7 = 26.22v

Some of the voltages

Regards

Spence.

 

[pvision]

I'n mo microphone expert but would think that, as it works for a brief moment, the problem lies around the pattern switch

IIRC the pattern switch changes the polarisation of the diaphragms . I could be mistaken - maybe it switches the output phase and leaves the polarisation the same. Either way, there are two diaphragms and their output is combined to give the polar response required

I'd look at that switch and maybe disconnect one diaphragm and test again

Nick Froome

 

[Spencerleehorton]

The problem seems to be further down the line I think as when I insert a 1k sine wave it doesn't appear on the output that much, I can hear something but it's very faint.
The waveform is all over the place.
All I can see is an upside down  waveform at the R109 point, no nice sinewave anywhere apart from at the gate of T1?
Will try and bypass the switch and see if this helps.
Any other ideas?

Regards

Spence.

 

[moamps]

DC voltage on drain of T1 doesn't look correct to me.

 

[Spencerleehorton]

I have replaced T2 and after much tinkering and unplugging and plugging back in the mic works again?
Still have popping and clicks but at least the mic is working, if I take the capsule off the noises are still there?
Could this be a DC/DC converter problem?

Spence,

 

[tubestation]

Hello Spence,
I just checked the Schematics of the 87 ai. There are no Referencevoltages shown. The  Frontend FET Stage of the 87/87a is quiet similar to the 87ai. The referencevoltage for the FET Drain  in the 87 is 11 Volt. Not sure If it s similar in the 87 ai but If You meassure  1,8 V there could be something wrong with the FET or T2. If You  need  to Replace the FET  You need to adjust the Source Resistor to Bias  the FET correctly for lowest Distortion and max Headroom. This is a good Way to do it.
https://dl.dropboxusercontent.com/u/43869772/U1621/BiasingFETScopeMethod.pdf
I just was reading that You Replaced T2. Do You still have 1,8 V at the Drain ? Probably not , hopefully now 11 V.
If You disconect the Capsule the + 60 Volt Part of the Capsulebiasvoltage from the  dc/dc Converter is still connected to the Gate. If any ac Ripple or Noise comes from the Converter, it will reach the Gate.  With the whole Headbasket removed the dc/dc Converter  is not  in the Signal.
If You have a Skope, You could  look at ( before, not After the 1G ) R 103/104 if You have the +/. - 60 Volt and If there is any unwanted  ac  coming from the Converter. If there is any ac, there is Something wrong with the Converter. There is also a Reasons why T 2  was faulty.
Greatings,
Lothar
Edit, some BUll  S...  removed.....

 

[Spencerleehorton]

Hi,

The bias resistor was label RS and wa 3.3k, I have replaced for the moment for a 10k pot.
I will measure voltages again but I have measured at the transformer BL and get 45.4v and WS = 45.3v.
I have changed T2 for BC547B.
At pin2 going to T2 emitter I measure 42.8v
Did you mean measure at R103/R110? I get 53.5v here.
I will try lifting R103 and see if this removes noise (AC ripple)

Spence.

 

[tubestation]

Hello Spence,
At R 103 the + 60 Volt for the Capsulebias for Cardioid and Omni.
At R104 the - 60  Volt for Figure of Eight.
To meassure  Drainvoltage there should be a Referencepoint between R 109/ C 106 or meassure at  C110.
You should find something like 11 V.
There must be a  Reason why T2 burned, maybee a shorted Capitor ?  C108/109 ?
Greatings,
Lothar

 

[Spencerleehorton]

I've built a 30v DC Psu and I'm going to disconnect the DC/DC converter and see if it behaves, if it does its the regulator circuit at fault.

Spence.

 

[tubestation]

Hello Spence,
Wouldn t it be more easy to  control the relevant Voltages first ?
I think after all  Your  nice  Micbuilds  You have a Scope  to look at R103/104  If You  have a clean, ripplefree  + / - 60 Volt dc  coming from the dc/dc Converter and If You have the correct Drainvoltage. This can be  done with the Capsulehead  disconnected .
If You use Your Method ,lift R 103 and connect Your external dc (  the 1 G is in the dc Rail )  with the Headbasket mounted ,
But how do You want to handle it ? How to connect the external Cable ?  With the Mictube removed ? This could be another Source  for Problems because the mega high Impedance Circuit is not shielded anymore and  Hum and HF Terror may reach the Gate.... Nice Radio. It could work If You have Two small 22 Volt Photobatteries  in Series and a Capitor , placed inside the closed Mic, but I personal wouldn t do it like this.........
Greatings
Lothar

 

[Spencerleehorton]

Hi,
I have
already stated that I only get 53.5v at r103 and -53.7v at R13, these
points so regulator looks off, but want to verify DC/DC converter is ok.
I will do what you suggest first though.

Spence.

 

[Spencerleehorton]

So with lots of help from Matt Nolan! I have built a separate 30v dc supply and removed the connection to RT and then I connected the 30v do supply to RT.
I am only getting 43.3v at R10 and the negative supply is all over the place!!!
So  it certainly looks like the dc converter is the problem.
How I proceed I'm not sure now but Matt has suggested I look at the caps?
Has anyone else got any ideas of how to fault find this dc  converter as it's pretty tight on this pcb and I don't really want to be heating to much up so I don't damage the pcb.

Regards

Spence.

 

[tubestation]

Hello Spence,
I used this Schematic to study the Mic.
http://groupdiy.com/index.php?action=dlattach;topic=44805.0;attach=7894
Do You use the same ? I ask because i don t want confuse with wrong Partnumbers.
Greatings
Lothar

 

[Spencerleehorton]

Yes that's the schematic for this mic.
Running external 30v DC supply .
Raised the voltage to 36.2v and I'm getting 53.7v @ R10 but at D8 it's 59.1v
Negative side looks good as well
T9 voltages are E-20.94v, B-21.50v, C-21.04v
T10 are D-0v, S-6.68v, G-20.96v

Any ideas?

Regards

Spence.

 

[moamps]

already stated that I only get 53.5v at r103 and -53.7v at R13, these
points so regulator looks off...

These voltage are OK so your converter works as it should. Your problem is too low drain voltage at T1, IMHO.  Please reconnect the phantom power and recheck and measure once again test points as you did in first post. 
Remember that you can get measurements errors when you are measuring high impedance circuit. So please measure DC voltages of the converter before R10 and R13. 

Please measure voltages on both side of R109.

 

[Spencerleehorton]

Hi,

Here are voltages
Cable tested at 48.2v
Xlr pin2 45.4v
RT  42.7v
R20 = 35.9
D8 = 58.8v
D6 = 10.63v
D4 = -59v
D5 = -11v
D1 = 58.8v
D3 = 23.44v
D2 = -18.92v
T1 = drain =3.87 gate = 0v source = 8.86v
T2 = emitter = 44.3v base = 42.6v collector = 42.4v
At start of R10 = 58.8v

Matt had suspected the regulator network, any ideas?

Regards

Spence.

 

[moamps]

Here are voltages...

Everything looks good, more or less.
Now, disconnect the capsule head from the body, apply 0,1V, 1kHz signal to pin 2,  gnd to pin 7, and measure AC voltage at T1 drain and C110 right side using a scope.

 

[Pip]

Here are voltages...

Everything looks good, more or less.
Now, disconnect the capsule head from the body, apply 0,1V, 1kHz signal to pin 2,  gnd to pin 7, and measure AC voltage at T1 drain and C110 right side using a scope.

Yes this sounds good. Any chance to see the condition of the capsule and top plate pin array?

 

[Spencerleehorton]

The popping noises are there without the capsule on, so it can't be anything to do with the capsule.
The Pin connections I have tested for continuity.
You have said "everything looks good, more or less" but I am not getting  +\-60v and on the cable I'm getting 48.2v on pin 2 when the xlr cable is inserted it comes down to 45.3v?
So something is obviously wrong here?
Matt did suspect the regulator section, so far nothing has been said about this?

Spence.

 

[Pip]

The popping noises are there without the capsule on, so it can't be anything to do with the capsule.
The Pin connections I have tested for continuity.
You have said "everything looks good, more or less" but I am not getting  +\-60v and on the cable I'm getting 48.2v on pin 2 when the xlr cable is inserted it comes down to 45.3v?
So something is obviously wrong here?
Matt did suspect the regulator section, so far nothing has been said about this?

Spence.

Firstly the popping noises lead me to believe bad caps not the DC to DC converter. The output voltage of the converter is coupled to the capsule by 1MEG resistors thus making it a hi-impedance affair, very hard to read voltages with an ordinary meter. Sounds to me like its working though. If the DC to DC unit was not putting out I think you would still hear something though rather low. Check C107 thru C112.