stablised screen grid supply

[Tubetec]

Thanks for looking over that Abbey ,
I needed to bounce out what was going round my head just to be sure I was on the right track.
I can see when I calculate the reactances at different frequencies for the choke it will make almost no difference at low frequencies and gradually has more effect as frequency rises .
I can see it would make a difference to the feedback  depending on if the choke or resistor is closest to anode , but which scheme is better or the merits of either I dont fully understand . In this instance where a simple single coil choke is being used I can see how having the resistor onto the the anode might insulate it from capacitance and because the feedback path is resistive keep that more linear also,  if the choke comes first it is encompased directly within  the feeback loop etc

I suppose I should try and test it ,but Im not really sure what to test for , I can see triode, UL  and  pentode (resistive or regulated bias) are possibilities with or without additional choke loading ,but what I cant decide is what would be a meaningfull way of measuring or comparing  the performance differences other than listening tests with actual musical content.

 

[Tubetec]

I found a good  page on Turner audio about choke  loading ,it also explains the choke/resistor load in signal stages.

http://turneraudio.com.au/audiofilterchokes-page3.html

 

[guavatone]

I have to say I like setting g2 screen with a separate filtered supply with a zener clamp bypassed with a 100n cap. The nice thing is that it sounds like you may save space with dropping/decoupling resistor/cap in your PS box feeding  a resister to feed 2-3 mA to the Zener.  In your main box the bypass cap doesn’t need to be full HT but enough to cover the Zener voltage so it doesn’t need to be a huge 600V film.

 

[Tubetec]

Thanks for the reply Guavatone ,
I found the old Bax post and decided to continue the discussion under that topic instead.

The screen by pass is 3x2.2uf 250vac.
I just posted a more complete schematic in the 'Baxadall puzzle' page.