anybody got a link to a data sheet for that tube???? understand that it's designed for lower voltage but not sure that disqualifies it as starved... hard to get those electrons boilling off at 20V it seems to me... and does it use a heater???
No, it's really easy to get electrons to boil off at 20V!
But first, a few points: first, it isn't voltage that gets electrons to boil off, it's heat, or in electrical terms, watts. Thus a 12A_7 tube can have the heaters run in series or parallel; the voltage changes, the current changes, but the watts remain the same.
Also we are talking about electrons boiling off the cathode, which warmed by the heater (in 6418 though, the heater and cathode are not separate parts, which I imagine increases the efficiency of the tube at the cost of some performance parameter). They flow towards the plate because the plate is positively charged!
Yes, electrons actually flow the opposite direction that some might conceive. If you have a positive DC voltage, it attracts electrons, it doesn't emit them (of course, AC flows both ways). This is also why for a FET, the "output" is called the source and the "input" is called the drain. That sounds backwards relative to where voltage is applied, but it's not from to point of view of the flow of electrons.
OK then, since we know that electrons flow from cathode to plate, we know that in any tube, you just have to get the cathode hot enough for electrons to jump to the plate. That is a function of temperature, as well as distance from cathode to plate, and plate charge, and probably some things I don't understand like materials.
"Starved" is not a term found in any datasheet I've ever read. Loosely, it is used to describe a tube operating below its specifications. Thus, 12AX7, for example, would be starved plate at less than 100V plate, or starved filament at less than 1.9W. You will see some people say tubes need the maximum rated voltages to not be starved (30V for 6418, and 300V for 12AX7), but I see no basis for that in the datasheets, nor in my admittedly limited experience designing tube circuits.
As for the AT3060 and the MSH-4, actually although they use the same tube, the design is somewhat different. AT uses an output transformer, and I used another FET to buffer the output. 6418 can actually only supply a pitiful amount of plate current, so something is required. AT also set the 3060 to run in a much more linear range of the 6418 than I used. That is partially because I wanted the harmonic distortion, but also because a byproduct of a more linear operating point is much more gain from the tube. That's exactly what you would want feeding an output transformer, but I was feeding a FET which would simply overload much earlier with a hotter output from the tube. And before it overloaded, the mic would have had a sensitivity of around -30dBV/Pa or more, which I felt was excessive.
I don't have any MSH-4s left, even for a picture.
Why use a tube in the mic and not the preamp? Well, condenser mics need a high-impedance buffer, either FET or tube. If you decide the tube is OK in the preamp, then you'll need a FET as the first stage in the mic. Lots of people feel that first-stage buffer is critical to the sound of the mic, due to the very high impedance at that point of the circuit, and its interaction with the tube. It would be somewhat analogous to a tube amp for guitar; running straight in vs. using a Tube Screamer in front of it.
Also, the tubes used in mics and preamps are often different. Preamps seem to use triodes quite a lot; you see those in mics sometimes, but more with inexpensive models. Many tube mics use pentodes instead. You'd have to ask someone more knowledgeable about tube design why that is; I used 6418 (a pentode, but quite different than the usual externally-powered mic tube) because it's the only tube I know of that can be phantom-powered and is readily available. It has a similar twin, 6419, but I haven't found any of those. Another tube that may be interesting is 6088; that could be run off of phantom power for the plate + an AA battery for the filament.
Hmm, 6418 datasheet has gotten hard to find for some reason . . . I'll host it on my site later tonight.
http://frank.pocnet.net/sheets/138/6/6088.pdf