>>>
Hey guys, how about an actual answer to the question?
It's 1/Gm
>> Is it an actual answer? 
> Gm = 5000
That's probably not the actual answer. Gm varies with current (Gm obviously falls to zero at zero current). 5,000uMho is measured at 10mA. That's a heck of a lot of heat in a microphone body; you more likely run 0.5 to 3mA. You can read Gm at your actual operating current by squinting at the 5840 graphs. Without knowing your design current or doing your squinting for you, I'll guess closer to 1,000 or 1,500uMho for typical mike head amp current, and 1,000-500 ohms cathode impedance.
However the real problem, as
**guest** asserts, is sometimes about driving long cable capacitance (or RF filters in solid-state board inputs). As an extreme, you may be putting 3V peak in 10,000pFd at 17KHz, or around 3mA peak. So, as
**guest** asserts, a simple cathode follower at low current will turn into a rectifier and splatt the highs. More realistic conditions don't actually splatt, just give rising distortion at high frequency and level.
A 1,000 ohm output impedance through 2:1 would appear to be 250 ohms. We know the 280 ohm SM-58 dynamic mike works, so this is fine, right? No. The dynamic is 280 ohms up to explosive sound pressure. The transformed tube is 250 ohms for small signals, swing 200 to 300 ohms for large signals, and will go to many K ohms when signal current exceeds tube current. We normally design condenser mikes for low output Z, ample signal current. After all, it isn't like condensers have low voltage output and we have to optimize voltage to overwhelm board input hiss.
You will generally be wanting 4:1 or 7:1 OT unless your mike-tube current is unusually large.
>
does the Gm of a pentode change when it is used triode-strapped?
Hardly. Essentially it rises from Ip to Ip+Is. If Ip/Is is ~10, as in many pentodes, Gm rises 10% which is less than our overall accuracy. The 5840 has Ip/Is more like 3, so Gm rises about 33%. If we knew what we were doing, this would matter, slightly. If we just leap to the first Gm spec on the show-off datasheet and assume it applies to our actual operating point, ignoring this 33% correction is fine.
