Neumann CMV5/B Build Thread.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

somorastik

Well-known member
Joined
Dec 12, 2008
Messages
47
Hello,

I built  CMV5/B  according to this schematic
http://elektroakustika.cz/cmv5b.html

And I am having some problems.
Both the Anode and heater supply are regulated by a linear regulator from a transformer.
I am using a capsule from a cheap single sided chinese microphone. The output cap is 500nF (Tesla box 250V), the tube is measured and 100% (Tesla EF12). Transformerr 1:10 I guess, it is something very interesting one guy suggested to me. It is a transformer from old microphone adapter(sometimes even build into a body of a dynamic mic) that was used to adapat the dynamic to a Hi-Z input aka an input for Piezo mic, typical tube preamplifier.
You can find them really cheap, hidden in these things.

The microphone has some hum and low sensitivity. I think the problem is connected to the fact that it should use a negative heater supply. I have possitive. I read that just now here on the forums that this circuit uses heater voltage to bias the EF12.


Is there a way to use a positive heater supply?

IMG_20150612_124904.jpg

IMG_20150612_124913.jpg

IMG_20150612_131553.jpg

IMG_20150612_131559.jpg

IMG_20150612_131606.jpg

IMG_20150612_131611.jpg


Thanks

Raz

P.S.: If anyone is interested in the supply I used I can post it here. It is a simple mosfet regulator for the 100V (set by trimmer), and an LM317 for the heater.
 
Hello,
Just a few Thoughts,
Do You use Cathode Bias ?
With whitch Resistor Value ?
If You Do so the Mic will Sound different than the Original because Cathode  (Self ) Bias  sounds and reacts different  than Fixed Bias , like it s used in this Mic. A Lm 317 works for Cathode Bias, for Fixed Bias not so good because it s too noisy, all Noise and Hum from the Heater Powersupply will be injected directly into the Cathode, no good.
I would stay with the original passive Powersupplydesign.
Have a  look here,
http://elektroakustika.cz/cmv5bpower.html ( the modern Part One )
or here
http://elektroakustika.cz/images/cmv/cmv5/cmv5ef12.jpg (Neumann  Original )
The 2 Cokes are 500 mH  each. ( I measured with my old handheld Meter at 200 Hz , at the relevant 50/60 Hz  it s probably 1,5/2 Henry ) mind the required 200 mA !!
The Funny Thing in the Cathode Circuit is a mechanical " On " Indicator , can probably be replaced with a 6,5 Ohm Resistor. Remenber that the Grid Resistor 50M , the Bias Resistor and the Condenser Capsule represents a Highpassfilter at 100Hz, so don t aspect too much Bass ,or increase the Grid and Load  Resistors  to 100 M.
For more Information  read here,
http://groupdiy.com/index.php?topic=50972.msg647054#msg647054
Good Luck and please keep us informed  how it s going  on
Greatings
Lothar
 
Ideally you need a negative heater supply and a very smooth one for this kind of biasing.

If you don't want to change your PSU, you can instead place a resistor between the cathode and the ground and bypass it with an electrolytic capacitor - 47uF to 100uF, doesn't have to be high voltage. The resistor should be chosen to put the DC at the cathode somewhere between 1.1 and 1.5V, I'm guessing. If you get the triode curves for the EF12, you could draw a load line and figure out where best to bias it. Currently, if you had negative heaters, the bias voltage would be the heater voltage divided by 9, according to your schematic and heater voltage divided by 3 in the Neumann original schematic.

You could bypass the bypass capacitor with a smaller film cap too, if you like.

There are other ways to do it, but this is quite simple. You're changing it from fixed bias to self-bias.
 
Hello,
Another Argument to use Self Bias in Your Case is that the EF 12 Tube  was not designed  for a Microphone Amp. Neumann used selected  EF 12 k ( klingarm, low Noise,  low  Microphonics )  for there Mics.  It can  be that You need to check 10 to 20  Tubes to find One that is working with low Noise for the  Fixed Bias Design. A Friend of me got  almost  crazy to find a quiet EF 12 Tube for his CMV 5.  Self Bias is not so  Tube sensitiv  and a not so  perfect clean Cathode  will work without too much Noise. Gridcurrent what  results in Distortion is also  a not so  big  Issue with Self Bias , I think.
Greatings
Lothar
 
Hi and thank you all for such a quick and supportive replies!

So first of all I decided to change it to self bias,  since I dont want to take the supply apart. But I decided to make a few changes in the pcb for the future, so it can be populated for negative heater supply as well.

I will chnage the cathode bias as such, and report back
http://www.offsetguitars.com/forums/viewtopic.php?f=27&t=42991&start=15

Because I didnt have a 50M resistor (and didnt want to hook five 10M in series) I used a 150M, does this mean my Highpass filter will go below 100Hz?

Thanks!
 
Hello,
If You increase the Grid and the Capsule Load (Bias ) Resistor, the Highpassfilter will have a lower cut off Frequency. The Funktion of the HP is  1/ (2 *Pi *R*C )
R is the Seriesresistance of the Grid and Biasresistor in Ohm, the C is the Capsulecapitance (~ 70 pF ) in Farad.
Fcut  (-3 dB Point ) = 1/(2 * 3,14 * 100000000* 0,000000000070.)
  If You use 150 M  for both R , it is Ok, for the EF 12,  if You make it much bigger You might get a higher Gridcurrent.
Greatings
Lothar

Edit, ....., sorry for my wrong Information.
R is the parallel Resistance of Bias and Grid Resistor,    just was reading this , posted  by Jonathan  Knight  http://www.neumann.com/forums/view.php?bn=neumann_archive&key=993770179&v=f

" Therefore, the resistor used to apply bias voltage to the capsule and the resistor used to bias the grid of the tube will appear in parallel as a load on the capsule. If each resistor were 150 Megohms, then the load seen by the capsule would be 75 Megohms. There is one more factor which will affect the low-frequency cutoff point (with respect to the grid of the impedance converter). Whether the capsule is direct- or capacitor-coupled to the grid, the input capacitance of the tube (grid-cathode plus the Miller capacitance: product of the grid-plate capacitance and the gain of the impedance converter) will also load the capsule. This will cause some attenuation due to the voltage divider effect. But since the equivalent input-referred capacitance will appear in parallel with the grid- and capsule-bias resistors, the effective RC time constant will be increased. In other words the low frequency cutoff point will be reduced. This effect would be quite small with cathode follower mics, but with Neumanns (common cathode amplifiers) it is not insignificant. It should be taken into consideration when selecting the grid resistor(s) to achieve a desired cutoff point. To put it in the form of an equation:
Fc = 1/(2*pi*Rg*(Ccap+Ggk+(1+Av*Cgp)),z
Where Fc = cutoff frequency Rg = bias resistor or combination of paralleled resistors Ccap = capsule capacitance Ggk, Ggp = grid-cathode and grid-plate capacitance, respectively Av = gain of impedance converter measured from grid to Plate "

Greatings
Lothar


 
OK. The circuit from the other forum is clever in that, much like the U47, it uses the heater supply to provide most of the current in the cathod self-bias resistor - so you can avoid using a bypass capacitor as the negative feedback is tiny (the cathode resistor can be kept very small, so changes in anode current generate very little voltage change across it).

However - this still means you're going to need a very clean heater supply. Hum and noise on it will be amplified almost as well as with the direct grid bias approach (actually, possibly more so, looking at the two circuits).

Take care to note the other differences in the circuits - the anode resistors, for example.
 
Hi Matt, thank you for the remark about the other schemtic.

Ok lets say I wont make changes to the original circuit I used, I will only connect the heater dircetly to my supply (+6V - GND) and connect a 1k (bias at around 1,6V anode current at about 1,6mA, if I am not mistaken) resistor to cathode and g3, parallel I will connect the bypass capacitor as you mentioned 47uF-100uF.

Would that work ok? I am not quite sure about the operating point
How should I modify the circuit connected to anode?

Anyway, it starts to occur to me that it would be easier to use negative voltage for the heaters... :)
 
Hello,
Now we are talking about 3 quiet different Mics.
If You want to stay with the CMV 5 b Design You need a negative ,very clean Heatersupply,. The easy Way is to use a passive Choke and R C  filtered  Design. The hard Way is a dedicated discret aktive Design , but not with a LM 317/337 (the negative Brother ) it s too Noisy. You also need probably more than one EF12 Tube  because the  Possibility that Your EF12 is low Noise is not very high.
If You want to change to Self Bias , try different Resistor Values from 1 to 1,6 k Ohm (from Cathode to Ground! ) and have a listen , it s better and more easy than every Calculation, I think. You may have to try different Grounding Concepts for a Humfree Operation.
If You want to go for the alternative U 47 style Design, You have the same Requierements  like the CMV 5 b Design.
So , the hard Part is to get a Powersupply for a fixed Bias Design , that  works with low Noise and a Tube that has a very clean , noiseless Cathode, not easy to find.
I  hope Your Outputtranny has enough Henries ( at least 150 or 200) to get Your EF 12 driven correctly., to have enough Bass.
Don t get confused, try the Self Bias first , than You can build another Powersupply and get You more Tubes to try for the Fixed Bias.
Just my 2 Cents........
Good Luck and happy Building !
Lothar
 
Back
Top