Neumann M49 Clone : Build Thread Puck Style (TLM49 Conversion To M49 b-c)

[lkipod]

So after burning in the tube all night, now the frequency of this strange noise is 1.6Hz.
I can't get my head around why it's there in the first place.

 

[Delta Sigma]

Is audio still cutting in and out? What rate is it doing it at? Does it match the frequency your oscilloscope sees? Can you monitor your B+ at the power supply and see if it is changing (dropping) when audio cuts out?

 

[lkipod]

Is audio still cutting in and out? What rate is it doing it at? Does it match the frequency your oscilloscope sees? Can you monitor your B+ at the power supply and see if it is changing (dropping) when audio cuts out?

Hey Delta, I am extremely greateful for your tips!

So after listening in to the audio it seems that it doesn't cut out, rather as if it changes phase in time with the frequency I'm seeing at the oscilloscope.
I'll try to record a bit where you can hear it.

 

[lkipod]

Some more tests, when switching patterns, at first this wave doesn't occur but after even a little noise in the room it starts to amplify itself, as if it is some sort of bad feedback, again just changing phase at a rate of now 2.2Hz.

I tried to remove C4 but the feedback persists. 

Which part of the circuit would cause such behavior? ???

I'm totally puzzled by this.

Best
Moshe

 

[ln76d]

Hello Moshe,

i didn't read all your posts, so if you already tested mentioned below parts of the circuit, then just skip it
I'm not sure which tube you are using for your build.

- try it with "true cardioid" only - if you used S2 for that purpose, then use it and just setup pattern selector to omni
Otherwise disconnect C1 from the front capsule side and set pattern switch also to omni

- try to change capsule for any other and observe is there the same behavior or swap the capsules sides if you don't have any spare

- check C1 at the first place in case of passive components

- for different substitute tubes there's worth to test C4 values from 3.3pF-10pF range

- if anything from above list didn't help, then disconnect R3 from C5 as also C2 from R2 - then test again

 

[lkipod]

Hey In76d!
I really appreciate you chiming in.

Hello Moshe,

i didn't read all your posts, so if you already tested mentioned below parts of the circuit, then just skip it
I'm not sure which tube you are using for your build.

I'm using 5840W, but also tried 6s6b-V.

- try it with "true cardioid" only - if you used S2 for that purpose, then use it and just setup pattern selector to omni
Otherwise disconnect C1 from the front capsule side and set pattern switch also to omni

I did disconnect C1 from the capsule side, but didn't set the selector to Omni, so I probably need to try that again. When I disconnected C1 I had less off that strange waveform.

- try to change capsule for any other and observe is there the same behavior or swap the capsules sides if you don't have any spare

I don't have a spare capsule at this moment, but I put two 51pf caps instead of capsule for testing, to the same effect.

- check C1 at the first place in case of passive components

I didn't understand what you mean by that. Anyway I swapped out C1 to another cap I had lying around.

- for different substitute tubes there's worth to test C4 values from 3.3pF-10pF range

When I put a 10pf cap in series with the 8pf the problem got worse, when I put it in parallel it got a bit better. 

- if anything from above list didn't help, then disconnect R3 from C5 as also C2 from R2 - then test again

I will try that if it comes to that.

Thanks a lot!
Now I have something to do

Best Moshe

 

[ln76d]

Hey In76d!
I really appreciate you chiming in.
I'm using 5840W, but also tried 6s6b-V.
I did disconnect C1 from the capsule side, but didn't set the selector to Omni, so I probably need to try that again. When I disconnected C1 I had less off that strange waveform.
I don't have a spare capsule at this moment, but I put two 51pf caps instead of capsule for testing, to the same effect.
I didn't understand what you mean by that. Anyway I swapped out C1 to another cap I had lying around.
When I put a 10pf cap in series with the 8pf the problem got worse, when I put it in parallel it got a bit better. 
I will try that if it comes to that.

Thanks a lot!
Now I have something to do

Best Moshe

Why you resigned from 6S6B?
Selector to omni is just to prevent  applying voltage to the back diaphragm.
What kind of capacitor and what's it voltage rate -  for C1?
Great - i forgot to write about replacing capsule with capacitor and testing - if the behavior is the same then, i think we can exclude the capsule - last question about it - are the diaphragms clean?
10pF and 8pF in series are ca. 4.5pF, paralleli 18pF (exluding capacitors tolerance here).
With lower value, you will get more gain so the "side effect" is just boosted and opposite with higher value.
What's voltage rating for your C4? If these are typicall ceramic capacitors with 50V ratings then maybe try some different with higher voltage rating.
I didn't had such issues with 5840, but some folks here had issues with oscilations using solid state replacements.
From what i remember it wasn't proper substitute to work with feedback which provides fixed HPF.
It's mentioned earlier whole node which starting from R3.

 

[lkipod]

Why you resigned from 6S6B?

I didn't, just tried different tubes to see if the problem goes away.

Selector to omni is just to prevent  applying voltage to the back diaphragm.
What kind of capacitor and what's it voltage rate -  for C1?

Polystyrene, 1000pf 600V, I thought at first that maybe I damaged it so I replaced it with polystyrene 250V that I had.

Great - i forgot to write about replacing capsule with capacitor and testing - if the behavior is the same then, i think we can exclude the capsule - last question about it - are the diaphragms clean?

Yes, capsule is brand new Thiersch red line.

10pF and 8pF in series are ca. 4.5pF, paralleli 18pF (exluding capacitors tolerance here).
With lower value, you will get more gain so the "side effect" is just boosted and opposite with higher value.

OK, so that explains what I found.

What's voltage rating for your C4? If these are typicall ceramic capacitors with 50V ratings then maybe try some different with higher voltage rating.

C4 is silver mica (first 8pf, now 18pf) rated 300V, would you recommend switching to ceramic cap?

I didn't had such issues with 5840, but some folks here had issues with oscilations using solid state replacements.
From what i remember it wasn't proper substitute to work with feedback which provides fixed HPF.
It's mentioned earlier whole node which starting from R3.

Ok i will search for it and read up.

Thanks a million!

Moshe

 

[ln76d]

I didn't, just tried different tubes to see if the problem goes away.

Polystyrene, 1000pf 600V, I thought at first that maybe I damaged it so I replaced it with polystyrene 250V that I had.

Yes, capsule is brand new Thiersch red line.

OK, so that explains what I found.

C4 is silver mica (first 8pf, now 18pf) rated 300V, would you recommend switching to ceramic cap?

Ok i will search for it and read up.

Thanks a million!

Moshe

250V-600V are ultra fine here - polystyrene are fragile for heat during soldering and cleaning with alcohol.
So the capsule is definately fine
You can try different type of capacitor - silver mica voltage ratinge is totally great, but urban legends says that these can be microphonic - maybe this is your extra sound source?

 

[Delta Sigma]

Super happy for you that ln76d has chimed in. I definitely feel too inexperienced to be giving you advice.

I like the idea of changing the capacitor type for troubleshooting. Don't worry about your end result, just find your problem. If you throw in a cap you have you're not happy with (correct value and voltage rating of course) but your problem goes away you can always reorder a polystyrene or C0G/NP0.

I once suspected a polystyrene cap (wasn't the cap) so I placed a Mouser order for replacement C0G/NP0 caps. Some people say polystyrene caps can be microphonic. I've never noticed this but I do like the peace of mind in troubleshooting with less sensitive (to IPA and heat) caps. When I ordered the few caps I needed I went to Dany's site and viewed all of his projects to order values that I may need in the future.

The side benefit of replacing components is the rewiring job you're giving your mic could end up fixing the problem.

Also, please re-confirm your DC voltages are fine (correct drop across R10, R6 & R7) and are not fluctuating with the phase issue. If replacing the capsule with fixed caps leave the problem still audible this narrows your issue down quite a bit. C2, C3, C4 (C4 can be removed), R5 and R3 (R3 can be removed).

This is frustrating stuff but you can learn a lot here, especially with this circuit. This may be one of the best circuits for hobbyists to modify and experiment with.

 

[lkipod]

Thank you Delta and IN76d.

I removed R3 and the problem is gone. ;D ;D
If I understood right it has a function of a HPF? Does R3 act in conjunction with C3 as an EQ?

I checked all voltages again, voltage drop across R10 is 68V, seems in the right direction.
Did some short tests on my voice and it sounds really good.

Actually I didn't find it frustrating, I liked the learning experience.
Most of the build it was a lot

Now I can try the Aluminium Oil output cap that I have here

I think I will try also to change C4 to 2-4pf C0G/NP0.

Cheers
Moshe

 

[ln76d]

Hello Moshe,

great that you found the problem

Yes, it's fixed HPF implemented to feedback circuit.
You should also get higher output. Without feedback it is similar to M49B/3 revision.
http://repforums.prosoundweb.com/index.php?action=dlattach;topic=37067.0;attach=2643;image
As you can see on schematic plate resistor is twice higher 200kohm - without feedback with this value you should get similar output level to the real M49, if we assume that with substitute tube it's correct.

If you don't want to use that feedback at all - then you can remove R3, C3, R4, R12, C8 and connect C2 direct to ground.
It will also remove measurement input, but i think it is useless for your needs.

For the grid-plate feedback try mentioned earlier range of capacitance. Don't let trick yourself with higher output, it usual can give impression of better response - rather do more tests and find the best sound for your needs

Alu-oil? can be good - my first choice for the output always is paper cap, second a good MKT.
Try 0.5uF-1uF range if you like testing. You can be surprised with lower values, but all depends, if your mike with 5840 and bias schematic values have lower low end, then rather go to 1uF.

 

[Delta Sigma]

I removed R3 and the problem is gone. ;D ;D

Awesome!!! It's quite a good feeling when you finally nail a circuit you've been struggling with.

If you have a few minutes, can you review your posts and edit them so both the original symptom is bolded and the solution to the problem. It makes it easy for those troubleshooting a similar problem to see your problem and solution.

I wonder if there was a problem with your 5M resistor or if maybe it has some oils from handling. Did you wear gloves? Next Mouser order, you should pick up another 5M resistor, maybe even a something bigger (for less feedback). You may not like the mic better with R3 back in but it's good to know.

Same with C4. Don't rip it out and be happy with no C4 vs C4 of 8pF. Try a few different values and record your results. I ended up deciding on 2pF then eventually removing it altogether. C6 is a low pass filter that can be reduced in value or removed. Many people remove C6 but I have mine in for now.

This is a very fun circuit to play with!

 

[lkipod]

Thanks!

I edited the posts, good idea.

In76d, do you think I should switch the plate resistor to 200k? Now I'm getting voltage drop of around 68v witch should be ok, or did I get that wrong.
Anyway it sounds pretty good.

I think I'm gonna try to put a 15MOhm resistor I have in place of R3.

I wonder if there was a problem with your 5M resistor or if maybe it has some oils from handling. Did you wear gloves? Next Mouser order, you should pick up another 5M resistor, maybe even a something bigger (for less feedback). You may not like the mic better with R3 back in but it's good to know.

Same with C4. Don't rip it out and be happy with no C4 vs C4 of 8pF. Try a few different values and record your results. I ended up deciding on 2pF then eventually removing it altogether. C6 is a low pass filter that can be reduced in value or removed. Many people remove C6 but I have mine in for now.

This is a very fun circuit to play with!

I do wear nitrile gloves when I work on mics. I don't think that the resistor was faulty, I measured it after I removed it.

For now I left C4 8pf.

Again thanks a million!

 

[Delta Sigma]

In76d, do you think I should switch the plate resistor to 200k? Now I'm getting voltage drop of around 68v witch should be ok, or did I get that wrong.

Just in case ln76d doesn't chime in, 68V gives you 0.68mA. The Neumann schematics (M49b) show 0.73mA. Not only are you close to the schematic but you're very close to the 0.69mA that my M49b with 6S6B is biased at right now.

I believe the M49 version that used 200k had a different transformer which made a 200k load resistor make sense.

I am curious if ln76d has thoughts about a different load for the 6S6B or 5840. There are ways to keep the amp biased the same and change the load sort of like in a U47.

 

[ln76d]

Just in case ln76d doesn't chime in, 68V gives you 0.68mA. The Neumann schematics (M49b) show 0.73mA. Not only are you close to the schematic but you're very close to the 0.69mA that my M49b with 6S6B is biased at right now.

I believe the M49 version that used 200k had a different transformer which made a 200k load resistor make sense.

I am curious if ln76d has thoughts about a different load for the 6S6B or 5840. There are ways to keep the amp biased the same and change the load sort of like in a U47.

Yes, but this is voltage and current reading, did you tried to trace both tube characteristics to compare operation points?
Not without a reason 5840 in exact M49C circuit have low end rollof  (without feedback HPF aplying).
If you change the bias (voltage and current), which change the point, then you will get flat response in low end, but it will change output level etc.
Except AC761, rather there's no other tube which fit exactly AC701 specs.
Even Oliver made for this occasion a version of T49 with "extended"low end.
I can't speak for M49B/3 transformer, since i didn't had it in my hands.
You can change the operation point in both ways - changing plate resistance and changing biasing resistance.
Best way is to mix both options and tune it for the needs.
Of course change of plate resistance will change output impedance but are you sure that 5840 or 6s6b will give you 200ohm output?
There were different transformers from what i have found. Even for BV11 some sources stand for  10:1 ratio, like some of transformer equivalents,  but IRT schematic shows 7:1.

 

[Delta Sigma]

I used Oliver's BV11r (extended low end and low mid) in my build. After replacing a 5840 with a 6S6B and getting a fuller low end I wondered if I made a mistake buying the BV11r. Interestingly, though, the AMI transformers page lists the BV11a for AC701 and the BV11r for 5840 but I found this on the TLM49 conversion page:

For my AC701 build I used the AMI BV11a (the very early transformer original used for the MSC2 tube) and for the 5840 I used the AMI BV11R (the one that was used for the C version).

http://www.tab-funkenwerk.com/id84.html

 

[ln76d]

Hard to guess what Oliver (R.I.P) had in mind and what were his sources of knowledge for the original transformers. He often wrote about Braunbuch.  Anyway this text from the link can be understood in double way.
There's probability that original BV11 of M49C had higher ratio. If you will look for  B and C datasheets, sensitivity of B is 7mV/Pa and the C is 6mv/Pa. I'm not sure which B version this datasheet is. If standard B, then both should be 8pF plate to grid feedback.
With the same ratio transformer C should be higher sensitivity than B. Of course still there's question about the output impedance of both!?!?
Still AMI BV11 standard also is 10:1 and it datasheet for the source shows exactly Braunbuch mentioned earlier
It's funny because i had both in my hands (never at the same time) but at the time i didn't even think about ratio measurement, it wasn't any point of the work, at the time i even didn't think that for few years i will start to build these circuits
Anyone can borrow time maschine?? Tardis will be enough i think - anyone have number to that new Dr. Who? 

 

[r2d2]

Neumann M49 is around ten thousand dollars these days. A TLM49 conversion at a fraction of the price sounds like a great deal for essentially the same mic.

Does anyone know of any similarly sized/shaped donor bodies for the D-M49 (hopefully cheaper than a Neumann)?

-James-

just "as well"  , it must be said that:

The sound of the original U47 come from  the VF14 (Helmet) original Tube ,
many have tried various different types of tube as replacement,
but no success about ,
and beware : there are many VF14 Fakes for sale , overall online…..
some info about here (as well):
http://www.phaedrus-audio.com/VF14M_story.htm

 

[lkipod]

Wow Delta and In76d, loving these posts...
It just takes me time to understand them :
That's the reason why I got hooked up with microphones, every detail is a whole world of knowledge... One day I'll understand

In another thread (about negative feedback) there was mentioned that the Phaedrus tube had this 3Hz oscillation that I experienced. Someone said there that it is because the electronic tube has a 180° phase shift at 10 Hz. If that's so why did it also occur to me with 6s6b-V and 5840 tubes? That doesn't make sense to me.
Or could it be that the tubes experience the same sort of phase shift? (Not very likely?..)
For me the problem is gone but it still intrigues me.

Best
Moshe