schoeps m221b transformer

[grantlack]

long time no melt!  been too busy moonlighting as a mechanic lately.  :
anyway, i was considering some bench time with the schoeps m221b circuit but have found absolutely nothing in regard to transformer data.
i'm not familiar with the way the transformer is connected here.  should it not bother me that there's a dc-coupled path to HT via the plate resistor?  not the first clue on determining a reasonable primary inductance either.  i could just drop stuff in for a bit of play and pray, but most of my microphone transformer iron is rather good and i'd like to keep it that way.

 

[CJ]

R6 is 3.3 meg, so not a lot of current thru the  pri,

 

[dmp]

R5 & R6 form a voltage divider to bias the tube it looks like. The cathode is connected directly to ground. C10 is still a coupling cap for the output.
I would think a transformer suited for a AC701k would be fine. AMI has transformers for the AC701 tube with ratios of 6:1, 10:1, and 12:1.

 

[grantlack]

i was concerned with the r4 junction and had sorted out the current-squashing at r6 from the get-go, it was the other primary connection that was throwing me... but the primary is continuous winding, and as such the current limited at one side of the primary is limiting it through the entire thing.  duh grant.  still dusting out the cobwebs  :-[

i'll give a bv.107 a shot and see what shakes out.  thanks for the kickstart!

 

[PRR]

Cathode is actually connected to heater + supply.

If grid were just ground-referenced, that would probably be too much for the tube.

The 3.9Meg+200K hold the grid up a few volts. The difference of this and heater voltage is, presumably/certainly, a proper bias for the tube.

The connection to tube plate (via transformer??) makes it semi-self-correcting.

> there's a dc-coupled path to HT via the plate resistor?

Sure. It is resistance-coupled, just like a guitar-amp preamp stage.

> clue on determining a reasonable primary

Plate resistor 200K. Not knowing the tube, I'd guess the combination of tube||R gives output impedance 1/3 to 1/4 of the resistor, 66K-50K.

Coupling cap 0.5uFd. For -3dB@20Hz response this can drive 20K. We probably want better, a higher load.

Both darts converge to like 50K for a max-transfer load. (Higher for a max-voltage load, but 50K is already as high as we want to go.)

Inductance is critical because L against 0.5uFd will resonate. If the reactance isn't swamped by resistance, it will ring-up, which is probably unwanted.

Assume 10Hz resonance. 0.5uFd at 10Hz is 32K. If swamped with ~~50K of resistance, resonant Q is about 32/50 or near 0.6. That's a nice Bessel-like slump, no drama, no subsonic bump.

0.5uFd resonates at 10Hz with 400H.

Double or half that is no big deal.

 

[grantlack]

spot on as usual with the circuit liner notes, i had most of that sorted out but am glad to have some gaps filled in- and the transformer stuff is a real gem.  i haven't done rlc resonance stuff in forever, i had to re-read it a few times.  the bv.107 seems fairly appropriate here though.

anyway i got it fired up, tons o' hums but i think that's more a function of wiring and a dirty chinese L7806 supply than a noisy tube.  no 60Hz, but harmonics galore.  when i kill power, noise floor cleans up nicely until the inevitable fade to black.  i'll try with a stiffer heater and maybe shield the lead, can't really reroute much in an mxl 991.  i used a 5703 since it was at the top of the drawer and a 1uF solen for coupling.  the mxl capsule is visibly unhappy with the polarization voltage, the diaphragm isn't completely sucked to the backplate but it's close. it sounds grainy as it sits, but i think there's potential (haha) if i can find a more tolerant capsule or wuss out and tweak the voltage.

 

[PRR]

> mxl capsule is visibly unhappy with the polarization

Isn't that a 48V capsule? Putting 120V on it may be unwise.

 

[grantlack]

> mxl capsule is visibly unhappy with the polarization

Isn't that a 48V capsule? Putting 120V on it may be unwise.

;D yep, that's a gentle way to put it.  i'm at a loss trying to think of a suitable non-schoeps capsule without modifying the circuit though.  maybe hack an electret?  pairing an omni diaphragm with cardioid chambering may help too, given that omnis often have a higher membrane tension.

edit: i just jerry-rigged it with your average 32mm k67 knockoff and (aside from the crap -65dB noisefloor) it behaved a bit better with no real protest from the capsule.  i'd really rather keep this an sdc though, so ideas welcome.  shielding the A+ lead accomplished nothing so hopefully an improved supply does the trick.