why do many tube microphones use triode connected pentodes instead of triodes?

[ELS122]

I see a lot of tube mics that use triode connected pentodes instead of just triodes, for example the legendary Neumann U47,
the capacitance between the control grid and screen + plate is around the same as that of triodes..., so is there a reason for this?

[TimFox]

Often, the choice of a triode-connected pentode instead of a regular triode is based on availability of a particular set of parameters.
Triodes are usually preferred for low-noise applications because they do not have "partition noise" that occurs in tetrodes and pentodes due to the cathode current randomly dividing between screen and plate.
Usually, the only downside of a triode-connected pentode is that the screen-cathode voltage rating is typically lower than the plate-cathode voltage rating of the pentode, and thus possibly lower than that of a true triode of similar dimensions.
With a pentode, in either connection, the g_m referred to cathode current depends almost entirely only on the screen voltage, where the dependence is that of a triode.
The original Neumann U47 selected individual Telefunken VF-14s for noise, grid current, and its good microphonic behavior (only one microphone in the package, please!), but that tube went out of production.  Neumann operated it at lower than normal plate and heater voltages to extend its lifetime.

[ELS122]

Often, the choice of a triode-connected pentode instead of a regular triode is based on availability of a particular set of parameters.
Triodes are usually preferred for low-noise applications because they do not have "partition noise" that occurs in tetrodes and pentodes due to the cathode current randomly dividing between screen and plate.
Usually, the only downside of a triode-connected pentode is that the screen-cathode voltage rating is typically lower than the plate-cathode voltage rating of the pentode, and thus possibly lower than that of a true triode of similar dimensions.
With a pentode, in either connection, the g_m referred to cathode current depends almost entirely only on the screen voltage, where the dependence is that of a triode.
The original Neumann U47 selected individual Telefunken VF-14s for noise, grid current, and its good microphonic behavior (only one microphone in the package, please!), but that tube went out of production.  Neumann operated it at lower than normal plate and heater voltages to extend its lifetime.

but didnt telefunken/nuemann order the VF-14's to be to be built in the first place?
looking at the datasheet it looks like the gain in triode mode is around 30, and transconductance of around 7mA/V, why not use something like a 12AV7, which has gain of 41, and gm of 8.5mA/V, also more linear and much less inter-electrode capacitance... they could've paralleled the 2 sections and still have less capacitance than the VF-14, and could drive the coupling transformer harder.
well there's one difference tho, the heater voltage for the VF14 was 60 volts, and they ran it off the B+ supply... but I've seen other neumann mic's have a seperate heater supply so idk. what were they thinking? 

well yeah looking into it, the 12AV7 came around much later than  the mic, but I doubt there werent triodes with similar specs at the time 

[TimFox]

Neumann and Telefunken were different companies when the U47 was introduced.  The VF14 predated the microphone, and was used for "normal" things during WWII.  There are interesting websites dedicated to it that point out the strange voltage environment back then for radio uses.
The U47 was distributed in the US by Telefunken early on.
A good reason to use the VF14 instead of a nominal 12.6 V tube such as the 12AV7 was the high-voltage, low-current heater, a good match for 48 V DC "phantom power" (reduced voltage to increase life).  With a triode mu of about 30, the tube could also operate with a plate supply of 48 V.  (The cut-off grid voltage of an ideal triode is minus the plate voltage divided by mu, and grid voltages less negative than, say, -1 V result in grid current.)
Another important reason is microphonics:  look at the construction of the VF14.  9-pin miniature tubes such as the 12xx7 series have a more microphonic construction.
They went to a 13CW4 Nuvistor (13.5 V at 60 mA heater) later when the VF14s went to Valhalla to play with the other obsolete tubes, but with controversial results.
Linearity may not have been of primary importance due compared with the non-linearity and relatively low AC output voltage of the capsule.

[ELS122]

Neumann and Telefunken were different companies when the U47 was introduced.  The VF14 predated the microphone, and was used for "normal" things during WWII.  There are interesting websites dedicated to it that point out the strange voltage environment back then for radio uses.
The U47 was distributed in the US by Telefunken early on.
A good reason to use the VF14 instead of a nominal 12.6 V tube such as the 12AV7 was the high-voltage, low-current heater, a good match for 48 V DC "phantom power" (reduced voltage to increase life).  With a triode mu of about 30, the tube could also operate with a plate supply of 48 V.  (The cut-off grid voltage of an ideal triode is minus the plate voltage divided by mu, and grid voltages less negative than, say, -1 V result in grid current.)
Another important reason is microphonics:  look at the construction of the VF14.  9-pin miniature tubes such as the 12xx7 series have a more microphonic construction.
They went to a 13CW4 Nuvistor (13.5 V at 60 mA heater) later when the VF14s went to Valhalla to play with the other obsolete tubes, but with controversial results.
Linearity may not have been of primary importance due compared with the non-linearity and relatively low AC output voltage of the capsule.

but they didnt run the mic's from phantom power, I dont think phantom power was even a thing at the time.
I doubt a pentode would be less microphonic than a 12av7 triode, or almost any triode I can think of...

[langwadt]

Neumann and Telefunken were different companies when the U47 was introduced.  The VF14 predated the microphone, and was used for "normal" things during WWII.  There are interesting websites dedicated to it that point out the strange voltage environment back then for radio uses.
The U47 was distributed in the US by Telefunken early on.
A good reason to use the VF14 instead of a nominal 12.6 V tube such as the 12AV7 was the high-voltage, low-current heater, a good match for 48 V DC "phantom power" (reduced voltage to increase life).  With a triode mu of about 30, the tube could also operate with a plate supply of 48 V.  (The cut-off grid voltage of an ideal triode is minus the plate voltage divided by mu, and grid voltages less negative than, say, -1 V result in grid current.)
Another important reason is microphonics:  look at the construction of the VF14.  9-pin miniature tubes such as the 12xx7 series have a more microphonic construction.
They went to a 13CW4 Nuvistor (13.5 V at 60 mA heater) later when the VF14s went to Valhalla to play with the other obsolete tubes, but with controversial results.
Linearity may not have been of primary importance due compared with the non-linearity and relatively low AC output voltage of the capsule.

but they didnt run the mic's from phantom power, I dont think phantom power was even a thing at the time.

google to the rescue, https://www.telefunken-elektroakustik.com/product/m-940h-tube-microphone-power-supply-u47-u48

wiki claims 48V phantom power came with the transistorized KM84

48V through two 6k8 resistors wouldn't power much of a tube  ..

[TimFox]

Maybe not phantom power, but still the same DC voltage for plate and heater supply.
There is no reason why a triode and a triode-connected pentode (all other things being equal) would be different in microphonics.
Microphonics in tubes depends on the construction, and physically smaller devices may be more prone to vibration due to less space for bracing, etc.
There is anecdotal evidence (which I have no reason to doubt, but no evidence to cite) that octal tubes, with their heavy plastic bases wired to a glass stem (such as the 6SL7) are less microphonic (again, all other things being equal) than 9-pin miniature tubes (such as the 12AX7) which have rigid pins through the glass base going directly inside to support the guts.
This webpage, from a manufacturer of a replacement module for the VF14, has an interesting essay on the history of the original:
http://www.phaedrus-audio.com/VF14M_story.htm

[ELS122]

Maybe not phantom power, but still the same DC voltage for plate and heater supply.
There is no reason why a triode and a triode-connected pentode (all other things being equal) would be different in microphonics.
Microphonics in tubes depends on the construction, and physically smaller devices may be more prone to vibration due to less space for bracing, etc.
There is anecdotal evidence (which I have no reason to doubt, but no evidence to cite) that octal tubes, with their heavy plastic bases wired to a glass stem (such as the 6SL7) are less microphonic (again, all other things being equal) than 9-pin miniature tubes (such as the 12AX7) which have rigid pins through the glass base going directly inside to support the guts.
This webpage, from a manufacturer of a replacement module for the VF14, has an interesting essay on the history of the original:
http://www.phaedrus-audio.com/VF14M_story.htm

> There is no reason why a triode and a triode-connected pentode
yes there is, since more grids = more moving parts.
TFK made incredible E88CC (there was some other letters in the model number iirc) tubes, frame grid, which were not microphonic at all, having everything going for low microphonics.
sure the E88CC tube was produced way later, but why did they still use pentodes in most of their mics when there were tubes like their own E88CC's...

it just doesnt seem like the right answer--"for the specific specs", there must be a specific reason that pentodes were used. maybe pentodes have some advantage when a super high value of grid leak resistor is used, idk. I guess only way to find out is to try contact some engineer from back in the day.

[TimFox]

There is no E field between the screen and plate (in triode connection) so the extra moving parts are not so important.
The frame grid tubes (including the Nuvistors) come much later.
An excellent frame-grid pentode (in 9-pin miniature) was the E180F, which was used in triode connection for lower-noise applications as well.
Back in the WWII American radar development, triode-connected 6AK5s were popular for low-noise.
There were a lot of tube models over time, but I don't know of any 50 V heater signal-level triodes.  Higher voltage heaters (with correspondingly lower current) were easier to power with DC (for low hum) before modern semiconductor voltage regulators. 
If you need a high grid-leak resistor, obviously you want low grid current, which is a very interesting topic that I have looked at intermittently over the years.  It's easier to obtain with a lower (or medium) mu triode, since that uses a more negative grid bias for a given plate voltage and current.  The thing not to do is use a high-mu triode (like half a 12AX7) at low plate voltage, since that requires a less-negative grid with correspondingly higher grid current.
Anyway, the article I referenced gives some interesting history about the VF14 tube before the Neumann microphone use.

[CaptDon]

I haven't looked at any specific characteristic curves lately but I thought the deal on triode connected tetrodes or pentodes had the downside of less gain but the upside of more linear transfer across a wider range of plate and grid operating voltages and the reduced stage gain showing a better signal to microphonic noise ratio and possibly less thermal noise contribution?

[S57UUU]

If I remember right, ECC88 was a high frequency double triode designed for TV tuners (which of course doesn't mean it didn't work at audio frequencies). EF180 (a succesor to EF80) was mainly for TV IF amplifiers. Because of their TV roots, these were very common tubes 50 years ago.
The dedicated audio pentode was EF86, in preamp stages often mounted on rubber vibration isolation.
E88CC was just a long life variant of ECC88, etc.   Just my 2 cents...

P.S.
I read the Phaedrus article, can really recommend it. The guys know their stuff.

[TimFox]

Many RF signal tubes of that era are "remote-cutoff" or "variable mu", designed for good AGC characteristics, and are not as linear as one would like for audio purposes.  Allegedly, in the post-vacuum-tube era, some 6ES8/ECC189 dual triodes, pin-compatible with the 6DJ8/E88CC, were re-labeled 6DJ8, but the 6ES8 is a "semi-remote-cutoff" construction designed for use in RF cascode input amplifiers.