High-value resistors for tubes vs JFETs (1G-100M)

[Khron]

I promise i'm not trolling, but i've always wondered, how come 1Gohm is the norm for JFET gates in microphones, while for tubes it seems to be 1-200Meg to maybe 500Meg?

 

[Ruud]

I suppose tubes generally don't allow such high values, because of the grid current.
FETs don't have this problem, you can go to 10 G.ohm (or higher) if you want.

 

[analogguru]

I promise i'm not trolling, but i've always wondered, how come 1Gohm is the norm for JFET gates in microphones, while for tubes it seems to be 1-200Meg to maybe 500Meg?

The value for FETs with built in gate-resistor (e.g. found in electret capsules) is only about 30 MOhm.

I suppose tubes generally don't allow such high values, because of the grid current.

+1

 

[zephyrmic]

With the mics I have built with JFETs I have used 5G SMD resistors to which I have attached flying leads, in parallel with a 33pF cap to go directly from the centre pin on say a TSB2555 mic capsule to the gate of the JFET and then to the output of the OCCAM 3B charge amplifier, and they have been very successful in that arrangement.

 

[Khron]

So that's the answer then, grid current?

 

[gyraf]

yes

 

[Khron]

In that case, pardon my ignorance, but what happens if the grid resistor is too great a value?

Excessive noise? More-than-usual distortion?

 

[gyraf]

bias drifts off in negative direction

which imo is why tubes from the same batch will sound wildly different in a high-z-input mic (like the G7 that I did with 1G grid resistors, not knowing this) grid leakage is unpredictable (but known to average lower at lower cathode temperatures - so underheating is often nice, if the tube can take it)

/Jakob E.