Designing a single 9 pin tube consneder mic circuit.

[ELS122]

Trick being:
1. without a transformer
2. with 2 phase split signals on the output
3. with gain

at first I had a design that was simply a cathode follower feeding a gain stage, the CF biases the condenser capsule and keeps a high AC input impedance even with low Rg values.
but this design just outputs 1 out of phase signal with a large output impedance too.

then I had a design that was a cathodyne phase inverter with the plate feeding a CF directly, simple and elegant, has low output impedance, biases the capsule, high input impedance, and 2 phase split signals. But it has no gain.

so now I want to design a circuit with high input impedance, some gain, phase split signal outputs.

at first I thought of using an LTP phase inverter but that only gets me an input impedance twice Rg so that's not all that great for high input impedance, yet it's probably the best I'll be able to do.

I thought of using that uncommon single pentode phase inverter which runs the screen really hot, biases the suppressor grid really negative, and then gets 2 out of phase signals from the screen and plate.
but I don't know of any triode-pentode tubes with the suppressor grid internally connected to the triode's cathode, not surprising 
since then I could have the triode bias the capsule, then feed into the pentode phase inverter.

maybe you have some ideas?

[TimFox]

As you know, there are no 9-pin triode-pentode tubes that fit your circuit.  With three triode pins, five pentode pins, and two heater pins something has to be internally connected to fit the 9-pin socket, and no one before you has thought of connecting the triode cathode to the pentode suppressor.  For your pentode phase splitter, the 6AS6 (7-pin) is probably the best choice:  I believe it is essentially a 6AK5 with a different suppressor grid pitch brought out separately to give a useful suppressor control.
Of course, you don't want to drive the pentode phase splitter directly from the condensor mike, since there would be a substantial Miller capacitance from the screen grid.
For 9-pin tubes, what about using a variation on the 6AN8 triode-pentode, using the pentode with cathode degeneration first to minimize the Miller capacitance load, followed by the usual triode phase splitter directly coupled to the plate (as used to drive output tubes in many power amplifiers).
One could use a bit of negative feedback from the triode plate to linearize the amplifier, and trim the triode's cathode resistor for balance.

[TimFox]

If you relax your requirement for nine pins, there are several possibilities in twelve-pin "Compactron" tubes.
For example, the 6AL9  https://frank.pocnet.net/sheets/084/6/6AL9.pdf
I don't know about the suppressor characteristics, but all the pentode terminals are available, and the screen has a reasonable dissipation rating (1.5 W).
Amusingly, the data sheet doesn't give the plate dissipation ratings.
However, the recommended operation data for 250 V on the pentode plate works out to 7 W plate dissipation.

[ELS122]

If you relax your requirement for nine pins, there are several possibilities in twelve-pin "Compactron" tubes.
For example, the 6AL9  https://frank.pocnet.net/sheets/084/6/6AL9.pdf
I don't know about the suppressor characteristics, but all the pentode terminals are available, and the screen has a reasonable dissipation rating (1.5 W).
Amusingly, the data sheet doesn't give the plate dissipation ratings.
However, the recommended operation data for 250 V on the pentode plate works out to 7 W plate dissipation.

well compactrons wouldnt fit in the 28mm? ID copper tube Im using as the case, for the time I built an LTP circuit with 10Meg grid resistors, it does slightly lack low end response, should probably use like 27Meg at least.
Im using a 6N3P (6CC42) for the tube. thermal noise is terrible, but I want to use up those tubes since I donr want to put them in anything else cause of the pinout.
the quietest tubes I have are suprisingly some Tesla E83CC frame grids.

I compared an LTP to a cathodyne with equal part values and same type NOS tubes, gain equalized.
and to my surprise the LTP sounded better, the cathodyne seemed to lack prescence and didnt have the 'body' to the sound as the LTP had.
wasnt too big of a difference tho.

btw I found out that the 6N3P would easilly arc from grid to cathode in a DC coupled cathode follower before the cathode heats up, so I needed put a neon between the grid and cathode, which is a good idea on any cathode follower anyway.

[TimFox]

Maximum outer diameter spec (Sylvania) on the 6AL9 T-9 envelope is 1.188 in = 30.2 mm.  Close, but no cigar.

[ELS122]

Maximum outer diameter spec (Sylvania) on the 6AL9 T-9 envelope is 1.188 in = 30.2 mm.  Close, but no cigar.

oh, that's much smaller than I had imagined.

I can probably just get thin wall tubing for the case with an ID of like 32mm and it'd still have a really similar form factor.
and I could fit a larger condenser capsule in it too...

so for the 9 pin tubes it seems the best I can do is using an LTP phase inverter for the requirements, that would give me an input impedance of roughly twice Rg.

but what specifically influences the input impedance? my LTP circuit has the tail resistor more than half of each plate resistor to give me a cathode voltage half of B+. so would this uncommonly high tail resistor value give me a higher input impedance? I get how cathodyne and cathode follower circuits get their high input impedance but with LTP the influence of the 2nd stage confuses things...

[TimFox]

For your triode circuits, the real part (input conductance) of the tube itself should be negligible, and the input impedance is essentially capacitive.
The high cathode resistance basically "bootstraps" the grid-cathode capacitance, leaving the grid-plate capacitance.
If your circuit also bootstraps the grid resistance (required for DC voltage, and limited by the requirement not to let the small grid current run away with the tube), that helps as well.
You can simulate this with Spice (.AC analysis), using a simple model for the triode (with only the voltage generator, plate resistance, and capacitors) and the external resistors.
Have you considered the pentode-triode circuit with a 6AN8-type tube?

[ELS122]

For your triode circuits, the real part (input conductance) of the tube itself should be negligible, and the input impedance is essentially capacitive.
The high cathode resistance basically "bootstraps" the grid-cathode capacitance, leaving the grid-plate capacitance.
If your circuit also bootstraps the grid resistance (required for DC voltage, and limited by the requirement not to let the small grid current run away with the tube), that helps as well.
You can simulate this with Spice (.AC analysis), using a simple model for the triode (with only the voltage generator, plate resistance, and capacitors) and the external resistors.
Have you considered the pentode-triode circuit with a 6AN8-type tube?

Spice sims have often given me false readings from poorly detailed tube models, so I'd like to calculate impedances myself from values included in datasheets.

what do you mean by pentode-triode circuit? using a pentode grounded grid amp to drive a phase inverter?

[TimFox]

See my "reply no. 1" above for my pentode-triode suggestion.
It looks something like the 6AN8/7199 stage in a Dynaco power amplifier.

Note that the relevant parameters (mu, rp, and capacitances) are the only things you can get from a datasheet, and should suffice for small-signal Spice calculations.
There are often problems with .TRAN analysis of non-linear circuits, but .AC analysis is essentially the algebra you could do explicitly and is as accurate as the parameters you have.
If you need large-signal performance, there are Spice models available that simulate the characteristic curves for some tubes.  https://www.duncanamps.com/spice.html
The input conductance at audio frequencies is never specified, but some data sheets for tubes like the EF183 and EF184 pentodes give conductance at RF, which depends on the unbypassed cathode resistance.
However, most tube data sheets specify the maximum grid resistance (constrained by grid current) for safe operation.