M49C circuit details / questions

[Khron]

I've started dabbling recently with some blasphemous twists on tube circuits for mic amps (impedance converters), and just today i've more or less "stumbled upon" the M49 circuits. I was also intrigued by the application of feedback, as well as the charge amplifier configuration (with the pF-range cap from plate to grid).

I'm slightly confused about some of the connections between the calibration / test signal input (pin 8), the heater supply (pin 4) and the pattern adjust voltage (pin 6).

Since i doubt i would or will ever use the calibration input, would the 2meg R12 still need to exist, and be connected to the pattern adjust voltage?
And how or why is C8 connected between the calibration input, and the heater supply, of all things?
Would removing these two components affect the workings of the circuit (significantly)?

I took the liberty to re-draw the M49C circuitry so as to be able to label the B+ and ground signals and not have to have long winding "wires" connecting them, to improve readability for myself.

Looking forward to responses from you much-wiser-people-than-I

Later edit - corrected component designators in redrawn schematic

 

[RuudNL]

Forget C8. Connect the 'bottom' of C3 directly to ground if you don't need the calibration input.
C4 ("abgleichen") is used to adjust the output level (Higher value of C4 = lower output. It is a form of feedback.
I suppose you don't need a fixed mV/Pa. ratio. In that case simply omit C4.

 

[Khron]

Wouldn't connecting the bottom of C3 to ground defeat the 5meg feedback resistor from the transformer primary, though?

 

[RuudNL]

No. R3/C3 form a low-pass filter. (More feedback for low frequencies.)
The presence of R4 (200 ohm) in series can be neglected compared to the 5M.
C2 is only there to isolate the AC feedback from the polarisation voltage.

 

[abbey road d enfer]

Wouldn't connecting the bottom of C3 to ground defeat the 5meg feedback resistor from the transformer primary, though?

C3 is already grounded by the 200r (R4). Compared with 5Meg it's a short.

 

[Khron]

Oh, right, of course - i see i mislabeled C2/3 in my redrawn schematic (now corrected), and that threw me off Thanks for the insight!

 

[Khron]

PS: Just how critical are the exact values of the 100-150Meg resistors?

 

[abbey road d enfer]

PS: Just how critical are the exact values of the 100-150Meg resistors?

R1 and R5 are the ones that define the spectrum of noise level due to thermal agitation (brownian noise) in the resistors. Actually they define the LPF cut-off requency of that noise. The lower the frequency (higher resistance) the better. But R5 cannot be increased that much, because grid current moves the bias point.
R2 is less critical, but it acts as a LPF with capacitor C2 to filter whatever noise could exist on the bias voltage.

 

[Khron]

While possibly sub-optimal, would it be "the end of the world" if the values were dropped down to, say, 33meg?

I see the ELA M251 uses a 30meg grid resistor, same with the M50.

 

[RuudNL]

It is a simple C-R filter calculation; say the capsule is 60 pF, this means with 33M the low -3 dB point will be around 80 Hz.
http://sim.okawa-denshi.jp/en/CRtool.php

 

[Gus]

If you search you can find that the spec for the 49c is -3dB +- 1dB at 40Hz

With some microphones it is good to have the PS schematic and manual if you can find them to figure out what is going on

 

[OneRoomStudio]

I think it is interesting to look at the choices that Oliver made in his version:


He omits C8/R4/R12 (from the original designations) since there is no need for the calibration circuit. He also omits the negative feedback (R3 in the original) and level-setting capacitor (C4). Switch S2 is permanently closed, and he uses 100M for R1 (R2 on his schematic) instead of 150M. All of these choices make sense, but I wonder what the cumulative effect is - especially omitting the feedback and using 100M rather than 150M for R1.

 

[abbey road d enfer]

He also omits the negative feedback (R3 in the original)

Which cancels the slight attenuation in the 10-100Hz region.

Switch S2 is permanently closed,

It means that the cardiod position is not as good as the front-capsule only cardioid. LF directivity would tend to be more omni.

and he uses 100M for R1 (R2 on his schematic) instead of 150M.

It makes tube selection a little easier (less reject) at the cost of slightly increasing the head amp noise.

 

[OneRoomStudio]

Which cancels the slight attenuation in the 10-100Hz region.

Do you know what the original intention of this was? Do you think Neumann & co intended for this to be somewhat of a built-in high-pass filter? Was it to counter-act proximity effect as maybe the U47 was seen as "too bassy," and the M49 was supposed to be more of a "neutral" sounding microphone?

 

[abbey road d enfer]

Do you know what the original intention of this was? Do you think Neumann & co intended for this to be somewhat of a built-in high-pass filter? Was it to counter-act proximity effect as maybe the U47 was seen as "too bassy," and the M49 was supposed to be more of a "neutral" sounding microphone?

I don't know for sure, but I would think the latter makes more sense than the former. The attenuation is mild (less than 6dB at 30Hz) and there's no difference below 10Hz.
Caveat: This is my conclusion resulting of simulation of both circuitsn, not the result of experimentation.

 

[RuudNL]

I must have measurements of the M49 circuit with and without the filter active.
But I suppose they are on my old computer.
You can simply disable the low-cut filter by lifting one end of the 5M.ohm resistor.

EDIT: found it! Blue is the fixed low-cut active.

 

[OneRoomStudio]

I was thinking about this circuit more and playing around with some variables, I thought others might find it interesting. Simulating the original M49c circuit gets you a curve that looks something like this:

Basically, down ~1.5dB at 60Hz compared to 1kHz, and down around 0.4dB at 20kHz compared to 1kHz. The slight treble attenuation is due to C6 (which I missed in my earlier analysis above).

The "Oliver" circuit looks something like this:

It's about 2.5dB hotter overall, due to the lack of feedback and lack of the "level-setting" cap (C4 in the original). It's also a bit flatter - down a little less than 1dB at 60Hz, and flat above that (technically, UP 0.004dB at 20kHz compared to 1kHz).

Not a huge difference overall though. I wonder if the treble cap (C6) was intended to have an audible effect, or if it was more about RF.

 

[elskardio]

or if it was more about RF

I remember reading somewhere that the M49 was intended as a broadcast microphone.

 

[bremusound]

I remember reading somewhere that the M49 was intended as a broadcast microphone.

The M49 was invented by the NWDR, a (North-West) German broadcasting company, in the 1940s.

 

[tomas.borgstrom]

Not a huge difference overall though. I wonder if the treble cap (C6) was intended to have an audible effect, or if it was more about RF.

The "treble cap" is indeed audible. The LPF in your graph already starts between 2-3k.

This cap works well with AC701 since it has a very detailed and excited top end. Other tubes can sound very dull and dead with it.