UL Screen Tap Question...

[Ritchie200]

Reading up on UL operation and I have a question.  All the articles I read state that the 6L6/KT88/KT66/etc family "likes" about a 42% plate v tap.  The 6V6 family "likes" about a 23% plate v tap.  Ok, I'll take their word for it.  However, nothing that I have read explains WHY!  Why that percentage and why the difference in tube families? Other than it is figured on a percentage of the impedance - there is no explanation.  If optimal is 40%, what happens if it is 60%? Or 20%?  Can anyone "esplain" to me in plain english what the heck is going on?

Thank you in advance!
Jim

[spacelabstudio]

Reading up on UL operation and I have a question.  All the articles I read state that the 6L6/KT88/KT66/etc family "likes" about a 42% plate v tap.  The 6V6 family "likes" about a 23% plate v tap.  Ok, I'll take their word for it.  However, nothing that I have read explains WHY!  Why that percentage and why the difference in tube families? Other than it is figured on a percentage of the impedance - there is no explanation.  If optimal is 40%, what happens if it is 60%? Or 20%?  Can anyone "esplain" to me in plain english what the heck is going on?

Thank you in advance!
Jim

http://en.wikipedia.org/wiki/Ultra-Linear

Apparently gives you characteristics of pentode and triode operation.  As far as how they arrived at the numbers, I'm guessing trial and error.  I wouldn't be surprised if no one has managed to analyze it physically.

Chris

[RicharD]

As I understand it (and I'm usually wrong), it's got a lot to do with the physical construction of the tube.  Beam Power tubes tend to have a more durable screen and therefore can handle higher voltages better.  True pentodes and tetrodes are happier with a lower screen voltage and therefore the UL tap should be further away from the plate taps.  Tubes such as the 6CD6 call for a screen voltage to be roughly 1/2 of the plate voltage.  They might not even work in a UL configuration.  Hopefully someone in the know will answer this question properly.

[jjasilli]

Without burning the mid-night oil, that's what I'm thinking.  Take the 6550: http://www.mif.pg.gda.pl/homepages/frank/sheets/135/6/6550A.pdf

Especially good for hi-fi.  In Class AB1 you get only 0.6% Harmonic Distortion with 400V plate; 270V screen at idle.  UL is a good way to get there.  And for mass production, UL also eliminates the cost of filter circuitry, including a choke, for the screens.  May as well put all that iron in one package.

So I'm thinking that design choice in selection of: a) the power tube; and b) the class of bias -- will determine what %-age of plate voltage the UL tap should be.  The goal seems to minimize Harmonic Distortion, and to fine-tune quiescent plate/screen voltage to do so.  But this will not apply if Harmonic Distortion is not an issue, or if it is desirable; then screen & plate voltages can be equal at idle.

However, I would consult the literature.

[DummyLoad]

try the link below. i have not read the whole article... it may or may not contain the information you seek, however, broskie does explain things well.

http://www.tubecad.com/january2000/page3.html

[PRR]

> WHY!  Why that percentage

It's BS. The percent isn't that fussy.

There's the two extreme "UL" conditions:

0% == Pure Pentode, high power output, no damping
100% == Pure Triode, lower power output, good damping

In-between you have a range of in-between possibilities.

There is a study, somewhere, showing that for specific tubes and goals, 25%-60% is in-between triode and pentode, with a slight preference for "43%". But it isn't that fussy. And if you pick different conditions and goals, you'd come up with some other "optimum", still not critical.

6V6 at traditional voltages (250V-300V) is marginal for current. Heavy UL probably trims power a lot. A lesser tap might do better than you'd pick for a fat tube like EL34.

Since the only OTs you can actually buy are 40% or 43%, and that's as good as any other value, it isn't worth asking "WHY?"

----------

> Tubes such as the 6CD6 call for a screen voltage to be roughly 1/2 of the plate voltage.  They might not even work in a UL configuration.

If you must do it, you use an additional winding, wound about 40% of the plate winding and fed the lower G2 voltage.

[Ritchie200]

Why?  Because it's there!   Thanks PRR, that makes sense.  Yeah, all the iron I've seen is 40% so it's a moot point if I was building - I was just trying to figure out how they got there.  And thanks to the rest of you guys for weighing in - I've been to all those sites!

Jim

[Merlin]

I read (though I can't remember where) that the UL tap that results in the greatest power/distortion ratio is closely related to the distance between screen and cathode; valves that have the screen closer to the anode naturally want a UL tap closer to the anode, and vice versa.
Plausible, but I don't know whether to believe it.

[kagliostro]

If you must do it, you use an additional winding, wound about 40% of the plate winding and fed the lower G2 voltage.

PRR do you mean like this ?

Kagliostro

[PRR]

> related to the distance between screen and cathode; ...Plausible, but I don't know whether to believe it.

The form of the problem suggests a ratio, not a pure "distance".

It's late. I'm not thinking. But a triode has good (certain assumptions) balance of power and gain when Mu is 4 or 5. 6L6 EL34 and kin have Mu of 10. 42% of 10 is 4.2, a good Mu. But I can't convince myself that 42% of Mu is the right answer.

[Merlin]

> related to the distance between screen and cathode; ...Plausible, but I don't know whether to believe it.
The form of the problem suggests a ratio, not a pure "distance".

You know perfectly well what I mean. If you want to be pedantic, "100% times the screen-cathode distance / anode-cathode distance".