Bias supply and Rg1

[hylaphone]

Hi all, I'm having a hard time finding a definitive answer to this - does the bias supply resistance itself factor into the maximum Rg1 rating of a power tube? Or is this only referencing the grid leak + stopper value?
I'm working on something with KT120 which has a max Rg1 rating of 51K, but the PT I'm using has no bias tap. Deriving bias supply from the HT secondary will mean a big series resistance and I wonder if I'm better off adding a dedicated transformer for this.
Thanks--

[Helmholtz]

- does the bias supply resistance itself factor into the maximum Rg1 rating of a power tube? Or is this only referencing the grid leak + stopper value?

The way I understand it, the limit applies to the total DCR between grid and cathode (or ground).
So it is the sum of grid leak (aka bias feed resistor), grid stopper and bias supply shunt resistor.
51k means a heavy load for the PI, lowering PI gain and drive capability.
So make sure to use a strong PI (not a 12AX7 with a 100k plate resistor).
Best would be CF drive.

This said, most designs exceed the Rg1 limit of the power tubes.
Making Rg1 too large bears the risk of thermal runaway caused by the grid building up a positive charge.

[hylaphone]

The way I understand it, the limit applies to the total DCR between grid and cathode (or ground).
So it is the sum of grid leak (aka bias feed resistor), grid stopper and bias supply shunt resistor.
51k means a heavy load for the PI, lowering PI gain and drive capability.
So make sure to use a strong PI (not a 12AX7 with a 100k plate resistor).
Best would be CF drive.

This said, most designs exceed the Rg1 limit of the power tubes.
Making Rg1 too large bears the risk of thermal runaway caused by the grid building up a positive charge.

Great, I thought so. Any drawbacks, theoretical or practical, I should be aware of when getting bias voltage from HT secondary?

[Helmholtz]

Well if the bias supply shunt resistor (that's the resistor across the output of the bias supply) already needs to be around 50k, then you're having a problem.

[pdf64]

My understanding is that the g1-k resistance limit applies at full dissipation.
As dissipation reduces, the g1-k grid circuit resistance can be increased.
So provided your design isn't attempting to squeeze the KT120s hard, you may be ok to eg double the total effective grid circuit resistance.

[catalin gramada]

My understanding is that the g1-k resistance limit applies at full dissipation.
As dissipation reduces, the g1-k grid circuit resistance can be increased.
So provided your design isn't attempting to squeeze the KT120s hard, you may be ok to eg double the total effective grid circuit resistance.

Yes, bias till max 70% power dissipation and it may double it. Some datasheets published the ratings for different power dissipation.

[Helmholtz]

My understanding is that the g1-k resistance limit applies at full dissipation.
As dissipation reduces, the g1-k grid circuit resistance can be increased.
So provided your design isn't attempting to squeeze the KT120s hard, you may be ok to eg double the total effective grid circuit resistance.

That makes sense as thermal runaway must depend on tube temperature.

[Helmholtz]

Yes, bias till max 70% power dissipation and it may double it. Some datasheets published the ratings for different power dissipation.

Can you post a datasheet example?

[catalin gramada]

Can you post a datasheet example?

Here you go

[catalin gramada]

Hi all, I'm having a hard time finding a definitive answer to this - does the bias supply resistance itself factor into the maximum Rg1 rating of a power tube? Or is this only referencing the grid leak + stopper value?
I'm working on something with KT120 which has a max Rg1 rating of 51K, but the PT I'm using has no bias tap. Deriving bias supply from the HT secondary will mean a big series resistance and I wonder if I'm better off adding a dedicated transformer for this.
Thanks--

Do a service for yourself and try to make a bias winding. 40-50 turns is not so hard to do if you have 1-2 mm room to insert the wire between iron and the coil. You may use any pvc isolated wire, not necessarily enameled one. If you can get enameled one, better. 0.3mm dia is plenty. Say awg26 hookup wire. Make the winding in 2-3 pieces to not be forced to fight workihg with the whole length you need..Be sure you make the winding tight ( meant the isolation have no chance to touch the iron) because a coil is vibrating. You don't want a short caused by polished isolation in time. Use a bridge rectifier.

[Chuck H]

Hi all, I'm having a hard time finding a definitive answer to this - does the bias supply resistance itself factor into the maximum Rg1 rating of a power tube? Or is this only referencing the grid leak + stopper value?
I'm working on something with KT120 which has a max Rg1 rating of 51K, but the PT I'm using has no bias tap. Deriving bias supply from the HT secondary will mean a big series resistance and I wonder if I'm better off adding a dedicated transformer for this.
Thanks--

The series resistor won't be the consideration here. Helmholtz already indicated this but I wanted reitterate that it's the grounded shunt resistor value in question. I think it's probable to calculate backwards for the circuit. As in, use a 12k shunt resistance and then calculate your series resistance to supply the required voltage. The series resistance will almost surely be sufficiently high so as to not load the HV supply significantly.

[catalin gramada]

The series resistor from hv tap will limit drastically the current. The charging time will be huge for a decent value cap to keep the ripple low enough. You want a negative bias voltage to be present before anything else...Otherwise a stb switch in screens circuit acting after certain time may help.(not to useful for accidentaly power stop/start which may take the hv fusing protection with it...)

[Chuck H]

Current is not typically a consideration for bias supplies in audio amplifiers.

Unless there is some supposed need to supply grid current I don't even know why this was stated.

And I would expect a compromise could be reached between circuit impedance (cap charge time) and loading on the HV secondary. These are normal considerations for any design.

[catalin gramada]

Yes, no significant current involved in circuit as bias voltage reference.

[hylaphone]

Thanks for the responses. How does this look?