How does this U87 capsule look to you?

[samgraysound]

Hi All,

Someone brought me a U87 to work on. The complaint is that it is "sounding thin" and "distorts easily."

I checked voltages inside. They look pretty ok, the only concerning one is 15v at the drain of T1. Schematic (attached) says 11v.

Took a look at the capsule, it looks okay to me, a little dirty. But I have no experience with these so I'm not confidant in my assessment. Could someone with some more experience take a look at this capsule pictures?

https://drive.google.com/open?id=1Ik_SI4SBSCyCxFj3hUccgIxrlcH9g22L
https://drive.google.com/open?id=1pnrYkS74_MoiyVhnGORNyvdxtHkxlWjk

 

[samgraysound]

fixed the sharing permissions

 

[kingkorg]

It looks fine, that amount of dirt shouldnt be the issue.

 

[RuudNL]

Distortion could be caused by incorrect biasing of the FET.
(Neumann specifies the drain voltage at 11 v, but according to my measurements this does not result in the lowest distortion.)
However, this does not explain why the microphone is sounding 'thin'.
You could measure the frequency response by connecting a generator to the "Messeingang" and measure the output at different frequencies. According to Neumann, 40 Hz should be -3dB re. 1 KHz.
You could extend the LF frequency response by changing C5 (increasing the value), but this should not be necessary.

 

[pucho812]

have you listened to it to confirm that it is indeed thin sounding and or distorts easily?

 

[samgraysound]

Distortion could be caused by incorrect biasing of the FET.
(Neumann specifies the drain voltage at 11 v, but according to my measurements this does not result in the lowest distortion.)
However, this does not explain why the microphone is sounding 'thin'.
You could measure the frequency response by connecting a generator to the "Messeingang" and measure the output at different frequencies. According to Neumann, 40 Hz should be -3dB re. 1 KHz.
You could extend the LF frequency response by changing C5 (increasing the value), but this should not be necessary.

What level signal can I put into the Messeingang input?

 

[samgraysound]

have you listened to it to confirm that it is indeed thin sounding and or distorts easily?

I tested it with an oscillator through a speaker just for level and to look at the waveform, but I haven't done other listening yet. I'll do that now.

 

[RuudNL]

What level signal can I put into the Messeingang input?

You could start with 100 mV.

 

[samgraysound]

Okay. I checked the bias resistor for symetrical distortion and tweaked the value up from 5.1k to 5.6k.

I also ran a frequency response test. It does have a bit more low frequency rolloff then spec:

-1.21 @ 125hz
-3.74db @ 63hz
-6.45db @ 40hz

here is all my frequency response readings and a chart: https://docs.google.com/spreadsheets/d/1SHhfD74JGKzUaTOYB77ZjMb1OZjkIOMKi2ME9P8WDvQ/edit?usp=sharing

Looking at the schematic this seems like it could be C4, C7, or C8. Anything else I should look at?

 

[RuudNL]

With 33nF (C5) in the feedback network the rolloff starts at a relative high frequency.
If you want a better LF response, increase the value of the 33nF capacitor to 100nF (or even 150 nF).

 

[samgraysound]

With 33nF (C5) in the feedback network the rolloff starts at a relative high frequency.
If you want a better LF response, increase the value of the 33nF capacitor to 100nF (or even 150 nF).

Ok, I'll give that a try.

 

[Whoops]

Did you compare it in listening tests to a well know good working unit?

I would start there before swapping components, the mic might be fine or the fault could be somewhere else.

It has to sound good with the stock component values

 

[RuudNL]

It has to sound good with the stock component values

Well, in this case you have to define 'good'.
With my original Neumann U87ai microphones 40 Hz was at -6dB.
That is the reason that I increased the value of the capacitor in the feedback loop to 100 nF.
Also it is a known fact that the Neumann capsules produced during a certain period didn't have a very good bass response, because the tensioning of the membranes was (too) high. Klaus Heyne has complained about this and wrote many posts about it.

Here is the effect of the filter with different capacitor values.

 

[Whoops]

Thank you so much for the insight Ruud