Neumann U67 Clone : D-U67 Tube Microphone Build Thread.

[Delta Sigma]

Make a quick dummy load for testing. Try something like 210k ohms (any wattage) for testing B+, 33 ohms 2W for H+. Those values are just using the currents written on the schematic. It won't be bang on but it'll get you close.

 

[FarisElek]

I did the test you suggested just now. B+ with 230kohms was almost exactly 220vDC and
the H- with 36 ohms  and it was about -16.0DC and slowly rose to -16.55DC where it kind of stayed at. Even with the resistors attached AC 210 in was still 270AC.

After the PSU stayed on for about 5 minutes unloaded, I began to smell a little bit of electrical burn. Nothing appears damaged and the readings didn't change. However, R2 is very hot to the touch when I turned it off. Not any hotter than a cup of coffee, I could hold my finger against it without it causing damage to my finger, but it is hot for sure. Zener Diode is hot but not as hot as R2.

Ryan

 

[TillM]

Did you put the 33ohm resistor between ground and the pin for the heater ?
When you turn the potentiometer, what is the lowest voltage, with resistor between ground and h+ ?

Never ever let the psu turn on too long without load !

 

[Purplenoise]

Has anyone experimented with different kinds of coupling caps? I have a few on order but I’d like to hear what you guys think. Thanks.

 

[Purplenoise]

So here is the first coupling cap test. Two wet tantalums ( 1uf 150v) in series ( sum is 0.5uf 300v). The high end is smoother and not as outlined/upfront as with the Vishay mkp. The mic feels weightier in a way

 

[Gus]

So here is the first coupling cap test. Two wet tantalums ( 1uf 150v) in series ( sum is 0.5uf 300v). The high end is smoother and not as outlined/upfront as with the Vishay mkp. The mic feels weightier in a way

You need to be careful when using caps in series for voltage rating. Look up how they voltage divide.  Then think about the cap values.

 

[Purplenoise]

You need to be careful when using caps in series for voltage rating. Look up how they voltage divide.  Then think about the cap values.

In series voltage is the sum of all caps no? Like 2 150v caps would give you 300v. Am I wrong?

 

[Gus]

If they are perfectly matched they will divide the voltage equally.

Calculate what happens if the values are different.

Often when caps are used in series(look at some tube guitar amp schematics in the power supply section)  "balancing" resistors are used

You can not use  balancing resistors in this application.

 

[Purplenoise]

If they are perfectly matched they will divide the voltage equally.

Calculate what happens if the values are different.

Often when caps are used in series(look at some tube guitar amp schematics in the power supply section)  "balancing" resistors are used

You can not use  balancing resistors in this application.

Capacitance will divide equally yes. But from what I read voltage is the sum of all caps in series. So in this example I have two caps of exactly the same values, 1 uf, 150V. Those two in series should give me 0.5uf 300v. Correct?

 

[Gus]

It depends on how they divide the voltage in real life.

If the two 150V rated caps are mismatched in value and have different leakages and you apply 300V, one cap might have greater than 150V across it.

In the U67 case the B+ is about 210VDC

What happens at the output cap(s) at B+ turn on?

Something to think about

 

[Purplenoise]

It depends on how they divide the voltage in real life.

If the two 150V rated caps are mismatched in value and have different leakages and you apply 300V, one cap might have greater than 150V across it.

In the U67 case the B+ is about 210VDC

What happens at the output cap(s) at B+ turn on?

Something to think about

Oh now I understand, you are talking about the actual real life tolerance of the components. Thanks, yes that’s a very good point. Well...I hope that especially with wet tantalums since they are made for very critical applications there is not so much difference between the two. Even at start up I don’t believe the B+ reaches more than 230v ...hopefully. 
One think I’ve noticed though is that I have “noise” in the sub frequency spectrum ( 20, 30, 40hz) about -50db max. It is not constant in level and since I can’t hear it on my setup I don’t know if it’s noise or the mic is now pickin up low freq info from environmental noise. When I record it seems dead silent though ( I can see it on the analyzer). Could that mean something?
I hope that works though because those caps sound fantastic in this mic, it smoothed out the high end, high mids, and gave a kind of “diffused” high end that places the sound a little back...not so direct. Buttery and really pleasant.
Thanks so much Gus.

 

[mickdundee63]

My mic in figure 8 and omni is a little unbalanced - i.e. one side is fuller and louder. Not enormously but easily noticeable. I tested some voltages and they don't seem correct. My B+ is ok at 210. On the R13/R15/R16 junction I was getting around 95 - 100v and at the R12/R13 divider junction for the polarisation voltage, getting only 47v. I pulled r12 and r13 and they are measuring fine. R15 was juuust out of tolerance at 3.5m so I swapped 3.3m in. This got me up to almost 110v at the R13/R15/R16 junction but still only 50v at the R12/R13 junction.

Any clues on troubleshooting this? I note also that the R13/R15/R16 voltage is drifting a lot - it oscillates up and down within a range of about 7 - 8v. Is this normal?? My B+ is rock solid.

 

[Delta Sigma]

You cannot measure voltage at either of those juntions. Your DMM will have an input impedance of 1M or 10M.

If your components test fine and your B+ is good, you'll have to trust your voltage divider math.

If both Fig 8 and omni, maybe you have some dirt, oil or flux loading the rear diaphragm. Check your Hi-Z section; clean it up if you can. Make sure no leads are touching wire and no components are touching. R1 is common to both fig 8 and omni.

 

[FarisElek]

I may be wrong. But isn't it normal for the backside to be out of phase with the front? You may want to flip the phase on your mic pre and see if that makes it sound normal/makes the front sound less full.

 

[Delta Sigma]

I may be wrong. But isn't it normal for the backside to be out of phase with the front? You may want to flip the phase on your mic pre and see if that makes it sound normal/makes the front sound less full.

For fig 8, yes. For omni no. Effect on performer depends on isolation of headphones.

Recorded audio should sound the same on front/back in fig 8 though.

 

[Tim Campbell]

What capsule are you using? Perhaps there is a loose connection or the capsule halves are porly matched.

 

[mickdundee63]

Ok thanks for the help. The Hi-Z section can be cleaned up a bit, I will try that first. Tim, it's a Neumann kk87 so I guess should be well matched halves. Obviously I've connected the backplates, not sure if that connection, if it's poor, could be having an effect. Ok that's a relief I'm not measuring incorrect voltage, but in researching this issue I saw discussion of the 59v polarisation voltage so is there a spot it can be measured?? I use it mostly in Cardiod so perhaps I'll just clean it up and get on with recording... Right off to research impedance (again)

 

[Delta Sigma]

I think Tim may be referring to the connection at the capsule. Probably safest to check everything else first (it's an expensive capsule to make a mistake on). Good to understand that it may be a possibility though.

 

[mickdundee63]

To what extent can the capsule connection be compromised? The backplates wires are soldered together, the multimeter indicates strong continuity. Could there still be issues?

 

[Purplenoise]

To what extent can the capsule connection be compromised? The backplates wires are soldered together, the multimeter indicates strong continuity. Could there still be issues?

So you have the kk87 with the 4 wires? For the old 87? Or the kk870 with the 3 wires? ( current production 87ai and original/new u67)