Screen resistors on ultralinear tap (Also bias question)

[RJ Guitars]

Hey guys,

I found this old post through a somewhat convoluted path, but it looks like you have some good understanding of UL tranny use for a bass amp. Would you mind taking a look at this forum post and add any relevant comments please?

http://diyguitaramps.prophpbb.com/forum28.html

Greatly appreciate anything you can add. Specifically I am looking at the need for adding screen resistors to my circuit so that I can keep a pair of KT-90's from dying an early death as well as just make a better amp.

thanks

rj

[gingertube]

RJ,
I re-read that thread, particular the stuff I posted to it.
Would stick by what I said there.

With Ultra Linear you have larger Miller capacitance at the output tube grid which probably will not make any difference for a Bass Amp but might for a HIFI Amp.

Would also recommend MOSFET source followers direct coupled to output tube grids with bias applied at the MOSFET gate (or tube cathode followers if you want to be a purist). The lower drive impedance from these would better drive the higher Miller capacitance but I would do it mainly to allow use of low Rg1 resistors for output tube bias stability without adding excessive load to the phase splitter.

Some tubes have more fragile screens that others, and pure pentodes like the EL34 tend to have higher screen current than beam power tetrodes and so get a reputation of having a more fragile screen. Therefore EL34 will normally need higher screen resistors in the UL connection. That's why you see 1K minimum on EL34 UL Application Notes.

On a 4 x KT88 into a Hammond 1650T, UL connected Bass Amp, I went 1/2 way. I used 6SN7 cathode followers off the phase splitter (just one for each side) then separate coupling caps and low value Rg1 resistors for each of the parallel output tubes on each side.

That was 10 years or more ago. If doing it again today I would use a MOSFET Source Follower direct coupled to the output tube grid for each of the 4 tubes.

Remember that the screen grid has gain and therefore REQUIRES a grid stop to supress parasitic oscillation. So you should always fit screen resistors of say at least 100 Ohms even if you don't need them for screen protection. Just like grid 1 stopper resistors the resistor body should be as close to the tube socket pin as practical.

Also UL seems to require higher stopper resistors than triode or pentode mode for parasitic oscillation suppression. AND counter intuitively it seems the better the output tranny the higher the screen grid stopper required.

That 4 x KT88 amp of mine had a HIFI prototype heritage in which I ran the Hammond 1650T and also tried a Plitron PAT4006 Torroidalal output tranny.
The PLITRON needed bigger screen grid stop resistors.
Cheers,
Ian

PS While I'm random raving on UL the other thing to take into account is that the Output Impedance of the amp is much less (say 1/3) of the value with Pentode connection. You will also want to drop any global feedback to 1/3 to do a fair UL / Pentode comparison.

[RJ Guitars]

Thanks! I am greatly appreciative. I went with 1K 5W screen resistors and I still have a few other things to sort out but this was a great help and probably saved me from a shortened tube life and other associated problems.

If you don't mind taking a look at the attached photo, I've made a little more progress and should be firing it up by the weekend. Lemme know if you see anything else regarding the use of a UL output section that might lead to disaster.

thanks again,

rj

[gingertube]

Looks good so far. Nice to see bias test points on the rear panel.
Cheers
Ian