Measuring Tube Gm

[mogliaa]

thanks, will check sim when I get back home.

1) yes. Problem is that I have a -90V negative supply and multiturn pot to set bias. Also a multipole switch to measure anode, screen and grid bias. When measure grid, the 1meg (or so) input resistance of panel meter affects the bias point (unless I reduce the grid leak resistor and replace input cap for a bigger one?). What option do you suggest instead?

2) my AC panel meter is 200mV (199.9mV actually), so could replace the 10 ohm for a 100ohm but limit GM read to 20mA/V?

 

[mogliaa]

Could change bias method with a cathode CCS bypassed by a cap and leave a grid leak input resistor?

 

[euro21]

Like this?
I think ccs not too happy with 3V remainder voltage.

 

[mogliaa]

Benefit of having a -90V supply at hand is that can provide the voltage drop the CCS needs. May not be better a bipolar CCS here instead?
Any additional advantage in using this bias method instead of fixed grid bias?

 

[euro21]

IMHO this measuring layout OK.

 

[mogliaa]

Tried yesterday same bias circuit but with 220K and the panel voltmeter shunted the bias point. Haven't tried with 100k, should be better but enough to compensate the meter impedance impact?

 

[euro21]

Try 220K grid leak resistor and 1Meg serial resistor from potmeter center point. These with 1Meg panel meter resistance OK about -14V grid voltage (-90V grid supply). Static grid setting point -9V.

Use 1uF shunt cap between pot center point and ground.

 

[mogliaa]

Try 220K grid leak resistor and 1Meg serial resistor from potmeter center point. These with 1Meg panel meter resistance OK about -14V grid voltage (-90V grid supply). Static grid setting point -9V.

Use 1uF shunt cap between pot center point and ground.

Not sure if I got this right: 1 meg resistor from pot Center to the 220K grid leak?

 

[euro21]

Voilá!

 

[mogliaa]

Not sure if I follow you here, when meter is connected then 1meg impedance is in parallel with grid leak presenting a resistance of about 180K that forms a divider with R5. If then the meter is switched to measure anode voltage for example, then the 220K is only forming the divider with R5. Surely this will move the bias point set before?

 

[euro21]

Oh, I understand!
To use only one panel meter with limited (1Meg) resistance to metering bias and another voltage, not an easy task.
Switching panel meter (on/off) surely will modify bias point, unless driving resistance not enough low.

 

[euro21]

Theoretically if you use (circa) 100mA current source (from 90V supply) and 500R potmeter, then moderately (high) voltage divider, 1Meg switching on/off only affect operating point below 1% (0.66).

 

[mogliaa]

Theoretically if you use (circa) 100mA current source (from 90V supply) and 500R potmeter, then moderately (high) voltage divider, 1Meg switching on/off only affect operating point below 1% (0.66).

Right, so we are back close to my original gyrator proposal from previous posts. A PMOS gyrator can sustain grid bias point when meter is switched on/off and also will offer high AC impedance, therefore will not load the source and provide good BW with a small Input capacitor, so why not using it?

Thanks for your help, I'm learning a lot with this!
Cheers,
ale