U87ai sounding 'thin'

[RuudNL]

I have a couple of Neumann U87ai microphones, that (compared to other condenser microphones) sound a bit 'thin' to my ears.
I measured the frequency response through the "messeingang" and found a severe LF rolloff.
(Of course with the LF switch in the "flat" position!)
My first idea was to replace the 47 nF capacitor (C 104) in the filter of the microphone with a 100 or 150 nF one.
But an international famous "microphone guru" states that this will only influence the "messeingang".
(I won't mention his name...)
IMHO the filter is part of the feedback circuit, and changes to the filter will influence the signal from the capsule as well.
Any opinions?

 

[ln76d]

This "microphone guru" don't know the shit
Deemphasis feedback affect both ends.
If i build U87 circuits in chinese donors with switch on the PCB i always make switchable in parallel 33nF/330nF caps in feedback node (33nF for old U87 topology). Also i mod few original U87Ai with 470nF instead 47nF for full low end response.
Anyway, comparing old U87 to Ai the low end cut sounds worse in Ai, probably because there's higher polarisation voltage which cause little bit different capsule response as also more high frequencies from lower value lpf cap.

 

[RuudNL]

Yes, indeed, that's what I thought!
Which proves that even microphone guru's (the ones that charge $1800 for a microphone modification) can be wrong from time to time...

 

[salomonander]

well there is also severe high end roll off...  the amp has lots of filtering in order to compensate for the capsules response.  personally i would not mess with it. neumann knew what they were doing (until sennheiser bought them . if the ais sound thin its most likely due to other issues.
from my talks to some techs most likely  too much voltage on the capsule and a too weak of a fet to handle it without distortion.  maybe give mr grosser a call.

 

[granger.frederic]

how old are they ?
first i would do a few tests
replace c108 and c109 first and then c107 and c110 after, if they are old or intensively used
check also polarization voltage at c6 and c9
clean all the hiZ parts especially r105
inject before c103 without the capsule (replaced with 80pf) to make a frequency plot...
you can tune and compensate with the feedback but i would first try with initial values, to eliminate the major leak

regards

 

[Tim Campbell]

They can sound as you describe if they receive less than real 48v phantom power. Some AI's have a piggy-backed pcb in them that can develop poor connections and might need to be re-seated. I have an 87ai and it is not bass shy.

 

[ln76d]

I  don't think that Rudolf have a need to call Andreas
Personally i don't like U87 response, especially Ai, but i'm aware why so many singers like it.
From my point of view there's a lot to improve in this circuit.
Properly working U87 is not bass shy, anyway low end response can be improved.
Neumann did low end filtration to minimalize proximity effect.
There's few different versions of both U87 topologies, there are different values of low end rolloff caps 10nF, 33nF, 47nF etc.
I never tested any U87 with weak phantom voltage but Tim can be right, worth to check.

 

[granger.frederic]

another way to extend the low end and soften the high end is to solder a zener 50v (bzv85) on top of D7, this  lowers the polarisation voltage like in the u87 vintage (around 48v)
you'll have a bit less sensitivity and lower SNR but i prefer the sound of the k67/k87 at 48v in a u87
i usually adjust the feeback after that modification and that really brings the mic to life ...

 

[RuudNL]

The U87ai microphones are only a couple of months old, so ageing can't be a problem.
They also sound identical, so it is not an exemplary fault.
A fact is that the LF rolloff is even more than specified by Neumann, with a signal injected into the "messeingang".
According to Neumann: -3dB @ 40 Hz.
I will try to increase the value of C104 first.
If this works, I maybe will add a switch to add a parallel capacitor to C104.

 

[Tim Campbell]

RuudNL you don't mention the phantom power. Most sound cards and many modern mixers put out 24v or less than 48 with many channels in use. The U87 needs a lot of clean phantom power to function correctly.

 

[RuudNL]

Phantom power is 48V!  ;D
(I'm using a pretty 'serious' mixer...)

 

[granger.frederic]

ok RuudNL,

if they're recent they must have the smd board in the HiZ section and i understand better why they sound thin !
i've modded two U87ai like that , least year, for a Jazz Studio in France...
the schematics doesn't match...
if i recall correctly , i've changed the coupling capacitor after the capsule (unsoldered smd) for a 1nF styroflex, changed a couple of capacitors for audiophile ones ( i remember c110 was a poor Jamicon !!! incredible for a 2600€ mic ) ,  done my polarisation mod , removed output RF filters, changed the feedback a bit and then rebiased the jfet
this brought back the mics to life thus it was not the capsules at all ( if you see what i mean ...  )

 

[RuudNL]

I am lucky, this is not the version with the (infamous!) SMD board! It says 'Circuit diagram 06'.
(I once had a version with a SMD board here, but I think Neumann wasn't very happy with it, because they went back to the 'old' version, with the 2N3819.

 

[granger.frederic]

ok then  my advice is to lower the polarisation , the caps mods, rf removal, feedback adjust, re- bias jfet
this should work
if not then the capsules are bad ones (but i doubt that...) if they are not contaminated

 

[micaddict]

Gotta love this forum.

Mr. Sennheiser, you there?

 

[granger.frederic]

you could try at last to unscrew the center screws then rescrew them to try to have a better diaphragm tension uniformity
(sometimes the tension is not uniform because of a violent shipping or a mic fall ...)

 

[RuudNL]

I think it would be very exceptional if two microphones would have the same 'damage'!
(Because they sound exactly the same!)
Anyway, I am not going to mess with the capsule itself...
From GroupDIY member Gus (thanks!) I received a Spice simulation for different values of C 104.
Values displayed are: 47 nF, 100 nF, 220 nF and 330 nF.
This shows a big improvement in the LF response, so I am going to try this first.

 

[granger.frederic]

ok but this doesn't eliminate a potential issue with your mics
this will do a "compensation" and the result can be sonically weird
a working U87ai like yours should not be "thin" or low end/low mids shy, unless the capsule is like that for some obscure reasons
however changing c104 impacts definitely the HPF 

 

[Recording Engineer]

Wait, a couple months old? Why not talk to Neumann about it even if you bought them used? Obviously they're not right even if they are matched. If Neumann won't fix a couple-months-old mics for maybe at the cost of shipping, even from a 2nd-owner, they are one of the lamest companies on earth now! On the other hand, if Neumann says there's nothing wrong with them, they really have gone kaput!; I don't care what anyone says!

 

[granger.frederic]

if they were not "kaput" , we wouldn't be here (and in many other topics) to discuss how to mod and improve some 2600 euros microphones ...
i add that i resold their top of range M149 tube as it was my worst sounding microphone in my studio (ok i have some jewels but ...)
the same for my tlm103 and my km184
i've  also modded a lot of M147 wich are the worst LDC for the price IMHO (imagine: they put a generic plastic encapsulated switching unit inside the metal  power unit !!! )
fortunatly they have redone the U47fet  and their capsules are still wonderfull  ( most their mics are amazing when the headamps are modded ... )