U87 high frequency problem

[samgraysound]

Hi All,

I'm working on fixing a U87 (a different one then the other I was posting about). I ran a frequency response test on the electronics using the calibration input and there is quite a bit of rolloff in the high end. Here's my test results:

https://docs.google.com/spreadsheets/d/1tIViAjrMqbHKXJrqvLFdGkMSjK95WlDL7nyug_VTVQM/edit?usp=sharing

What could be causing this?

Sam

 

[pucho812]

is this without the capsule and just the electronics?

 

[samgraysound]

is this without the capsule and just the electronics?

yes

 

[RuudNL]

The HF rolloff is the de-emphasis for the capsule.

 

[Whoops]

The HF rolloff is the de-emphasis for the capsule.

Yes, the capsule  has an high frequency boost and the circuit is a De-Emphasis

You need a setup to test it with the capsule

 

[samgraysound]

The HF rolloff is the de-emphasis for the capsule.

Strange, I was comparing it another one I am working on which did not have the high frequency cut.

Sam

 

[gurkan75]

I’ve seen U87’s with the deemphasis removed (one side of styroflex cap desoldered). Maybe that was a thing.

 

[samgraysound]

I’ve seen U87’s with the deemphasis removed (one side of styroflex cap desoldered). Maybe that was a thing.

which one is the styroflex cap?

 

[Whoops]

which one is the styroflex cap?

Styroflex = Polystyrene

This is a Polystyrene cap, but I don't know if it's the one Gurkan75 was talking about.
Polystyrene caps are normally used for picofarads values

 

[samgraysound]

Styroflex = Polystyrene

This is a Polystyrene cap, but I don't know if it's the one Gurkan75 was talking about.
Polystyrene caps are normally used for picofarads values

I did a visual exam of all the polystyrene caps and no pulled legs.  I suppose I could retouch all the solder joints...

 

[samgraysound]

Ok. So it seems like the consensus is that the U87 without the high frequency rolloff is the one that is not working correctly. I could definitely use some more guidance tracking down why.

Sam

 

[alhe]

-3dB down sounds about right for a u87 capsule. The resonant frequency should lie somewhere in the 3-5dB range and centered somewhere around 10k.

 

[chops]

-3dB down sounds about right for a u87 capsule. The resonant frequency should lie somewhere in the 3-5dB range and centered somewhere around 10k.

Yeah, but unless I’m mistaken, he’s saying he’s got another 87 that doesn’t have the deemphasis roll off in the amplifier.  He’s looking to troubleshoot why the hi freq roll off isn’t present on that one.
- Not the 87 in the graph in his first post of this thread.  (OP, Maybe you should show us a freq plot of the one you are trying to troubleshoot?)

 

[samgraysound]

Yeah, but unless I’m mistaken, he’s saying he’s got another 87 that doesn’t have the deemphasis roll off in the amplifier.  He’s looking to troubleshoot why the hi freq roll off isn’t present on that one.
- Not the 87 in the graph in his first post of this thread.  (OP, Maybe you should show us a freq plot of the one you are trying to troubleshoot?)

Oh snap, sorry,  I thought I had already posted it:

https://docs.google.com/spreadsheets/d/1SHhfD74JGKzUaTOYB77ZjMb1OZjkIOMKi2ME9P8WDvQ/edit#gid=646074354

 

[Khron]

http://audioimprov.com/AudioImprov/Mics/Entries/2014/2/7_Mic_Electronic_Eq_Pt.1.html

No / too little high-cut could only(?) stem from the pF-range cap having lost some capacitance, i would think.  Or becoming disconnected altogether.

 

[samgraysound]

http://audioimprov.com/AudioImprov/Mics/Entries/2014/2/7_Mic_Electronic_Eq_Pt.1.html

No / too little high-cut could only(?) stem from the pF-range cap having lost some capacitance, i would think.  Or becoming disconnected altogether.

Just to be super clear we're talking about C6? http://cdn.recordinghacks.com/images//mic_extras/neumann/U87-schematic.png

 

[samgraysound]

Ok. I tested C6. It is supposed to be 220pf. It tested at 225pf. I also retouched and did a continuity test on all of it's connections.

No effect on the frequency response. Still flat in the hi end.

What else can I look at?

Sam

 

[samgraysound]

Hmmm. So for kicks I started increasing the value of C6 just to make sure the feedback circuit was working at all. And indeed increasing the value increases the high frequency rolloff.

I need to up the value of 820pf to get -3db (well -2.77db) at 16k.

This U87 has had the JFET replaced with a 2N3819. Do you think that high frequency response of that transistor could be different enough that the value of C6 would have to be changed?

 

[RuudNL]

820pF for -3dB at 16KHz is a very unusual value. (In my U87's there is only 160pF for -3dB at 16 KHz.)
I suppose something else is wrong.
The 2N3819 is fine, Neumann used this type in the U87.

 

[mjrippe]

Let me ask the obvious - Is the capsule original?  If it was changed at some point (to a clone perhaps) it may have a different response.