Neumann U67 Repair - Low End Loss - Help needed

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Whoops

Well-known member
GDIY Supporter
Joined
May 9, 2008
Messages
8,186
Location
Portugal
Hi,
I'm working on a Neumann U67 that was sent to repair because it had low end loss.
The Studio has two U67 mics, one of them is fine but this one at some point lost low end.
The Mic works and although the sound is not ultra thin or radio-like there's a noticeable low end loss.
The owner had 2 telefunken EF86 tubes as spares in case needed.

I started the repair by checking the PSU,
The Filter electrolytic capacitors were originals, took them off and there's some bubbles on the underside.
Those were all replaced by new Panasonic Caps.

The PSU power transformer doenst have an 240V input tap so power is hard wired to the 220v tap, a previous engineer revised some resistor values so the voltages would drop to the schematic values while being fed by 230/235 VAC mains power.
I tested all the PSU voltages with the mic plugged in with tube in place and after tweeking the value of one of the resistors the voltages were spot on , 210V and -6,4V.

Next I tested the tubes,
tesla EF806S and two telefunken EF86. The 3 tubes tested fine in the Hickok 6000A tube tester.
Then tested one by one for sound in the U67, all of them sounded the same, low end loss was present regardless of the tube used.

I visually checked the capsule and gave a gentle clean, although it didn't look dirty or dusty at all.

Next with the EF806S tube installed I followed the schematic and measured the voltages at the mic itself.

here is the schematic:

U67%20Repair%20-%20Voltages.png


The first voltage measurements are marked on the schematic with a carbon pencil.

I had 140v at the junction of R12 and R13 instead of 59V, also at the capsule Pin 1 I had 30v instead of 59V.

Even with those wrong voltages the mic was working, and was just noticeable because it lacks a little bit of low end.

Started measuring resistors and lifting legs.
Found that R12 was open, resistor gave no reading on the DMM.

Replaced R12 and measured again, those new voltages with R12 replaced are marked on the schematic with a Red marker pen.
Now HT rail droped from 210v to 206/207V (this doesnt seem problematic as Neumann states +-2% tolerance on the HT rail)
Heater voltage drop from -6.4 to -6.22, seems to be good enough

At the R12 and R13 junction the voltage is now 48V , better than the 140V reading with the dead R12 but still 11 volts less than the schematic value of 59v.
Strangely at capsule pin1 there's only 8V now and at Pin 5 only 3V instead of 120V.

Even with these strange voltages the Mic still works, and to be honest it seems to sound the same as before R12 replacement.

I'm really tired now and need some help before moving on.

Do you know by looking to the schematic which are the components I should suspect next?

Thanks
 
Last edited:
You cannot measure voltages at the capsule unless you have a high input impedance voltmeter. A standard meter will load the points down too much.
The U67 has a test input where you can inject a sine wave and check the preamp frequency response. Pin 6 on the Tuchel.
 
radardoug said:
You cannot measure voltages at the capsule unless you have a high input impedance voltmeter. A standard meter will load the points down too much.
The U67 has a test input where you can inject a sine wave and check the preamp frequency response. Pin 6 on the Tuchel.

So you're saying It's not possible to have  correct readings at the sockets end of Pin's 1 to 6?
(the sockets were the Pins from the capsule and grille apparatus plug in)

the DMM I'm using is a Fluke 179 dont know if it has high input impedance or not, it's this one:
http://en-us.fluke.com/products/digital-multimeters/fluke-179-digital-multimeter.html

How about the 48V between R12 and R13 where it should be 59v, who do you think could be the suspect culprits?

R11, R12, R13, R15 and R16 all measure the correct resistive values.

 
It's obvious that you have no experience in this area, so please be careful with the mike worth 10k. You can easily lower its value for significant amount.

You can find input impedance of your DMM in the manual, or you can measure it using second DMM, or you can calculate it using voltages you get on resistors R12, R13 and R15 before and after replacing faulty R12, just ask Mr. Ohm for help.

The reading of 49V on R12 is fine.

From my experience in repairing U67,  you should check C9 for capacitance, the output transformer's windings continuity and all switches for  oxidation.
 
Fluke is around 10M so any polarisation voltage you need measure before hi-ohm resistors. Result should be -5V from correct voltage. Check C1, C2, C3, C9. For proper low end test on cardioid.
 
ln76d said:
Fluke is around 10M so any polarisation voltage you need measure before hi-ohm resistors. Result should be -5V from correct voltage. Check C1, C2, C3, C9. For proper low end test on cardioid.

Thank you so much, so I would still be able to read the voltages there although I should expect the readind to be aroun -5v less than the schematic values is tgat correct?

How do you advise on testing the capacitors for the low end loss? Should I carefully replace one by one while listening each one after change and check it that solves the issue?
And then if no improvmemts are made I put the original caps back in

I can measure Capacitor values out of tge circuit with the Fluke, I have also some cheap capacitor ESR meter, although i think the caps can have good readings on both and still impact low end loss so its better to do by listening, does this seems right?

Thank you
 
Whoops said:
Thank you so much, so I would still be able to read the voltages there although I should expect the readind to be aroun -5v less than the schematic values is tgat correct?

Something like that, it all depends on meter - there's possibility that you will read proper value, but it can also go lower around -5V.

How do you advise on testing the capacitors for the low end loss? Should I carefully replace one by one while listening each one after change and check it that solves the issue?
And then if no improvmemts are made I put the original caps back in
I can measure Capacitor values out of tge circuit with the Fluke, I have also some cheap capacitor ESR meter, although i think the caps can have good readings on both and still impact low end loss so its better to do by listening, does this seems right?

Start from the back - first change  C9 and do the listening test - if working the same, change C3, if the same - another.


 
ln76d said:
Whoops said:
Thank you so much, so I would still be able to read the voltages there although I should expect the readind to be aroun -5v less than the schematic values is tgat correct?

Something like that, it all depends on meter - there's possibility that you will read proper value, but it can also go lower around -5V.

Ok thanks, so it seems the readings I'm having at Pin1 and Pin5 are so far away from the schematic reference values that it's not due to the lower impedance of the multimeter but show that something is really wrong in the mic circuit.

How do you advise on testing the capacitors for the low end loss? Should I carefully replace one by one while listening each one after change and check it that solves the issue?
And then if no improvmemts are made I put the original caps back in
I can measure Capacitor values out of tge circuit with the Fluke, I have also some cheap capacitor ESR meter, although i think the caps can have good readings on both and still impact low end loss so its better to do by listening, does this seems right?

Start from the back - first change  C9 and do the listening test - if working the same, change C3, if the same - another.
[/quote]

Will do that and report.
thank you

I will also have the other Neumann U67 from the studio in a few days, that one is working properly and sounds amazing, at the moment it was being used in a session.
I can check if the Low end loss is a capsule problem by replacing the capsule plugin module of the bad mic by the one in the good mic, and vice versa.

 
measure the voltage where C9 joins the transformer primary,

this is to see if a leaky cap is allowing DC into the transformer which would drop the bass response due to a massive drop in transformer inductance,

even with a 10 meg voltmeter, you have a 400 meg resistor, which forms a voltage divider with the voltmeter on the order of 10/400=1/40 , so a reading of 3 v would really be 3 * 40 = 120 in the real world,
 
For pin 1 you should measure between R12 and R13 for pin 5 between R13 and R15.
I would check whole network - R10,R11,R12,R13,R15.
Even with good resistors but let say 10% your voltages can change, becuase there's a few dividers.
You can also use 1G resistor, connect in series to the multimeter lead and measure at high ohm resistors (this is only for check it wouldn't be so accurate). Then you will need to use 2V range for DC.
 
Hi mate .  Sorry because I comment so late. As I can see guys covered almost eveything. I can add that you should clean all hi Z parts like R8. Also if the c9 is leaky check the primaries of the tx.  From red to green you should measure approx 1kohm in both  windings. Be carefull, some of the leads are "under the hood". Make detailed sketch and color code (make some photos too) so u can assemble it again. Also don't be lazy and clean the tube socket ;)
 
CJ said:
measure the voltage where C9 joins the transformer primary,

this is to see if a leaky cap is allowing DC into the transformer which would drop the bass response due to a massive drop in transformer inductance,

Thanks will measure C9 for leakage.

CJ said:
even with a 10 meg voltmeter, you have a 400 meg resistor, which forms a voltage divider with the voltmeter on the order of 10/400=1/40 , so a reading of 3 v would really be 3 * 40 = 120 in the real world,

I see,
so for Pin 1, it would be something like 1/6, so 8V*6*=48V
that makes sense


 
ln76d said:
For pin 1 you should measure between R12 and R13 for pin 5 between R13 and R15.

So for it means that R1/R16 and R2 will not drop voltage?
Voltage before them will be roughly the same as after?


ln76d said:
I would check whole network - R10,R11,R12,R13,R15.
All those resistors measure the correct Resistance value.
R12 was open so I changed that


ln76d said:
Even with good resistors but let say 10% your voltages can change, becuase there's a few dividers.
You can also use 1G resistor, connect in series to the multimeter lead and measure at high ohm resistors (this is only for check it wouldn't be so accurate). Then you will need to use 2V range for DC.

thats a nice trick, but I didnt get the part of "you will need to use 2V range for DC" ,
could you elaborate a bit more please?
Thanks
 
Last edited:
Moby said:
Hi mate .  Sorry because I comment so late. As I can see guys covered almost eveything. I can add that you should clean all hi Z parts like R8. Also if the c9 is leaky check the primaries of the tx.  From red to green you should measure approx 1kohm in both  windings. Be carefull, some of the leads are "under the hood". Make detailed sketch and color code (make some photos too) so u can assemble it again. Also don't be lazy and clean the tube socket ;)

No worries my friend, wish all is going good in Belgrade!

What do you mean by "clean all the HiZ parts like R8"?
I cleaned the components with a soft brush, although everything was quite clean.
The Mic physical condition I would say is Excellent.

I will measure C9 for leakage and also measure the Transformer,
thanks for the DCR of the primaries.
Do you know also with aprox DCR should I measure at the secondaries?

I already cleaned the Tube socket with Deoxit and varios tube insertions



 
Use clean isopropyl to clean soldering points contacts etc.
I didn't use Deoxit, so i don't know it, but be carefull with all these contact cleaners - i had really bad experience with some.
Result was completely opposite.
 
Whoops said:
Moby said:
Hi mate .  Sorry because I comment so late. As I can see guys covered almost eveything. I can add that you should clean all hi Z parts like R8. Also if the c9 is leaky check the primaries of the tx.  From red to green you should measure approx 1kohm in both  windings. Be carefull, some of the leads are "under the hood". Make detailed sketch and color code (make some photos too) so u can assemble it again. Also don't be lazy and clean the tube socket ;)

No worries my friend, wish all is going good in Belgrade!

What do you mean by "clean all the HiZ parts like R8"?
I cleaned the components with a soft brush, although everything was quite clean.
The Mic physical condition I would say is Excellent.

I will measure C9 for leakage and also measure the Transformer,
thanks for the DCR of the primaries.
Do you know also with aprox DCR should I measure at the secondaries?

I already cleaned the Tube socket with Deoxit and varios tube insertions
yeah, everything is fine and cloudy-cold here ;)
Cleaning hi Z is usuall procedure for all mic's. During the years dirt from atmos can lower the resistance of the parts. Even if it sounds silly, do it it cant harm.  Clean Isopropyle is way to go. sec. DCR should be low. 10-20 ohm. Let me check mine transformers :)
 
Sec on one of mine U67 tx is: Blue-yellow,  green-white,  19.7ohm.  White-brown 36.8ohm
Hope this helps ;)
 
I always have my doubts if it is a good idea to measure the resistance of a (microphone) transformer winding with a multimeter...
(You might magnetize the core material.)
 
RuudNL said:
I always have my doubts if it is a good idea to measure the resistance of a (microphone) transformer winding with a multimeter...
(You might magnetize the core material.)
Meterials used in mic tx's are said to be magnetically "soft" because they can absorb magnetic energy without retaining it. Magnetically "hard" materials, on the other hand, make good magnets because they can retain a strong magnetic field even after the source of the magnetic energy has been removed. When a magnetic field is removed from a magnetic shield alloy, there is no remaining (residual) magnetic field. The only problem I see with measuring guys like this is possible super thin wire but here i's not critical
 
Moby said:
RuudNL said:
I always have my doubts if it is a good idea to measure the resistance of a (microphone) transformer winding with a multimeter...
(You might magnetize the core material.)
Meterials used in mic tx's are said to be magnetically "soft" because they can absorb magnetic energy without retaining it. Magnetically "hard" materials, on the other hand, make good magnets because they can retain a strong magnetic field even after the source of the magnetic energy has been removed. When a magnetic field is removed from a magnetic shield alloy, there is no remaining (residual) magnetic field. The only problem I see with measuring guys like this is possible super thin wire but here i's not critical

Thank you so much
 
Back
Top