Output transformer placement in tube mics

[ghiatorino]

Hi folks,
i am building a tube mic using a paralleled ecc88 tube, and because of small avaliable space into mic body i want to place the output transformer inside the PSU,  like some older designs i saw (Sony C37, AKG C61 etc), but..
those mikes take output from tube's cathode, thus an already low impedance line.
My plate output impedance is around 2k, and i need to run this signal into a 5 meters (good shielded) multicore cable before to get to the output trafo, so together with B+ and Heater DC voltage lines.
Is it a bad idea or may i possibly get a noise-free signal?
regards
t.

 

[ruffrecords]

My main concern would be possible loss of HF due to cable capacitance. At 100pF/metre. 5 metres could easily be 500pF. This has a 3dB point at about 160KHz with a 2K source. Seems  superficially OK.

Cheers

Ian

 

[gyraf]

Try it, and see what happens. Maybe it works for you.

The reason why this (probably) wasn't done in traditional tube mics, but only by cathode coupling, is probably because ECC88 and related low-rp frame-grid tubes really dosen't work well for microphone head-amps.

Yes, I know several companies use them like this, but that does not mean that it is a good idea.

Jakob E.

 

[ghiatorino]

Thanks for answering,
i will try this output trafo setup and post results.
btw, Jakob, why do you think that ECC88 doesn't work well with mic head amp? I have some built with this tube in parallel operation and they work very well to my taste.
let's discuss it if you like, my thoughts about:

Pro's
1) Low rp allows a lower ratio output trafo (so lower tube gain, lower noise, cheaper parts..)
2) Hi mu if needed
3) paralleling halves outZ and noise
4) cheap tested tubes avaliable
5) good noise and microphonics behaviour

Con's
1) paralleling doubles Cin
2) ECC88 Cin is double comparing to classic head amp tubes (i.e. 6072)

cheers,
t.

 

[CJ]

go for the CCa version,  ???

http://www.audiotubes.com/6dj8.htm

 

[gyraf]

Requires quite low grid leak resistor. Opposite of what we want for mic front end.

Jakob E.

 

[ghiatorino]

Tested the setup as written and no problems at all in running the output unbalanced until the output trafo (placed inside the psu).
Regarding Jakob's comment on the low value grid leak resistor required by ECC88 i made a lot of thinking about, but think not to have enough knowledge on tubes to understand it properly.
Is it related to the frame grid design or to the low Rp value? or are this 2 characteristic related (i.e. a frame grid tube has always a low Rp value). And, more important, what's the reason why it calls for a low value grid leak resistor? What's the theoretical downside of using the usual 1G to ground? I am using ecc88 with usual 1G to ground with no problems in terms of output level/noise/freq resp
cheers
t.

 

[abbey road d enfer]

Pro's
1) Low rp allows a lower ratio output trafo (so lower tube gain, lower noise, cheaper parts..)
2) Hi mu if needed
3) paralleling halves outZ and noise
4) cheap tested tubes avaliable
5) good noise and microphonics behaviour

Agreed.

Con's
1) paralleling doubles Cin
2) ECC88 Cin is double comparing to classic head amp tubes (i.e. 6072)

Not such a big deal in a cath-follower configuration since Cgk is bootstrapped almost 100%. The resulting Cin is typically lower than any common-cath stage, where Miller effect dominates.

 

[abbey road d enfer]

Requires quite low grid leak resistor. Opposite of what we want for mic front end.

This needs an explanation. The only one I see is that paralleling tubes doubles grid current, so for a given tube, there would be an increase of noise due to halving the grid resistor (CkT noise) of  3dB (this is assuming the bias resistor is much higher than the grid resistor), an increase of 3dB of flicker noise due to doubled grid current and a decrease of 3dB of the 0.4/gm noise. In all likeliness that would translate as a slight increase  of the head amp noise, that may be (should be?) negligible against the acoustic impedance  noise of the capsule.