UltraLinear Screen Drive

[TheGimp]

Most tubes respond well to 40% screen taps on ultralinear transformers. However, there are tubes suuch as the 4164B that require greatly reduced screen voltage (250 max) wrt plate/anode voltage (700).

Acrosound made a transformer specifically to address this (TO-350) with the ultraLinear taps as a totally isolated winding which could be biased at any voltage necessary.

See p11 of the acrosound transformer catalog for a schematic of a 100W 4164B amp:

http://www.clarisonus.com/Archives/Trans/Acro55.pdf

Would it be feasable to use an output transformer with the standard 40% screen taps, and capacitivly couple to the screens with screen bias supplied from a seperate DC supply via screen resistors?

Has anyone ever seen this done?

Tks.

Steven

 

[gingertube]

Check this other thread

Gimp,
See this thread for a discussion of what you are talking about.
http://www.diyaudio.com/forums/showthread.php?t=97384
A simple capacitor coupling won't work but a MOSFET circuit can be implemented to separate the AC and DC conditions on the screen connection.
Cheers,
Ian

 

[SpreadSpectrum]

I've never heard of it being done, but don't see why it wouldn't work. The question is, will it work well? Bias network would have to be low enough impedance to take care of screen current. That might be the killer.

I think plate to grid feedback is a much more elegant solution for beam tubes. I just made some curves for KT88 with 20% plate to grid feedback. I think mu was about 5, Rp was like 400 Ohms and it looked to put a 300B to shame as far as linearity goes. I'm just wondering if it will be stable if I build it.

 

[Tubelab_com]

I did get the circuit that I simulated in the thread mentioned in post #2 working. I was experimenting with cheap 6L6GC's in case something blew up. I don't think that I got it to sound as good as a real UL teansformer, but I didn't have much time to work on it before a string of unfortunate events curtailed most of my tube experiments for almost 2 years. I was thinking along similar lines, except I was thinking sweep tubes, big sweep tubes. The kind with a 1000 volt plate rating, but a 250 volt screen rating.

I had all sorts of new ideas during the 2 years of no experiments, so I started in on some of the fresh ideas and have since learned how to extract over 100 watts from a pair of 6L6GC without any glow (AB2) so I never got back to the "adjustable split load" experiments.

I recently read a thread on another forum where someone proposed a rather novel circuit using a 12AU7 to feed the screens.

http://geek.scorpiorising.ca/GeeK_ZonE/index.php?topic=4239.0

I think plate to grid feedback is a much more elegant solution for beam tubes.

Use a driver with a fairly high rP and then simply add a resistor from the plate of the output tube to the plate of the driver (2 resistors for P-P). Start with about 100K and tweak from there.

 

[Michael Koster]

Gimp,
See this thread for a discussion of what you are talking about.
http://www.diyaudio.com/forums/showthread.php?t=97384
A simple capacitor coupling won't work but a MOSFET circuit can be implemented to separate the AC and DC conditions on the screen connection.
Cheers,
Ian

The adjustable distributed load thread... All the promises to build it, yet no one
built it, at least no one reported back... Maybe now it will get built!

I think plate-grid feedback is much more useful and easier to design to a specific
impedance and mu goal, so that's what I'm building these days.

Edit - Sorry tubelab, posts crossed. So there was a working prototype!

 

[Tubelab_com]

Sorry tubelab, posts crossed. So there was a working prototype!

There was a furball of clip leads that made sound. Not sure if that qualifies as a prototype. It's still around here somewhere, but I got a big amp that needs to get out of my head and onto a chassis. Soon the Simple P-P's will be here, to take over my workshop.