Chimneys for Your 813's - only $4.95 each

[WA1GFZ]

Plate Z will be pretty low. A series inductor will raise the Z ahead of the pi network. A simple Pi network would have a small inductos since the Plate Z may be under 500 ohms. Layout will be critical. I wonder if it would be better to configure it as a traveling wave amplifier.

[K1JJ]

Plate Z will be pretty low. A series inductor will raise the Z ahead of the pi network. A simple Pi network would have a small inductos since the Plate Z may be under 500 ohms. Layout will be critical. I wonder if it would be better to configure it as a traveling wave amplifier.

I'm open to design ideas, Frank.

I was initially planning to  put four tubes across and three deep in a big rectangle. Then run some wide brass or copper stock down the lines to bolt off the grids and screens directly to ground/chassis.   

For the bifilar choke I wud use a long string of ferrite cores with the two #4 wires run through. That worked well in my other linears.

I know your 4cx-250 1kw test amp is a traveling wave. Tell me the advantages and basic physical layout for it to work.

I think the traveling wave is a broadband amp maintaining 50 ohms along it's path?  How broad is "broad" when centered on 3.8 mhz?

Are we talking a class A1, grid driven config now?  If so, that would solve the gain problem, but make it more complex and harder to tame.


T

[KD6VXI]

I doubt they made lanterns big enuff for 4X1's but that wud be a good find too... :-)

T

I've used Coleman glass in practice on a single 4CX5000 as well as a 4-1000.

--Shane

[WA1GFZ]

Tom,
I don't know of any NB application for a TWA but each tube is isolated from each other by a series inductor setting up a delay line with the tube C.
Just don't know you might surf and see what you find. They also make solid state TWA amps. Not very efficient in BB class A. I think we run about 6 to 7 KW to get 1 out.

[K1JJ]

Tom,
I don't know of any NB application for a TWA but each tube is isolated from each other by a series inductor setting up a delay line with the tube C.
Just don't know you might surf and see what you find. They also make solid state TWA amps. Not very efficient in BB class A. I think we run about 6 to 7 KW to get 1 out.

That's what I thought - poor efficiency with little to gain in ham service.  I'll stick with the GG config and go from there.

T

[w4bfs]

hi Tom ... 12 - 813 in parallel would be a visual marvel but VERY difficult to get each tube to equally share the load ... at the very least you would need to use some resistance in the grids to ground ( or individual bias adjustments) to equalize the individual tube gains ... unfortunately this would tend to lower the gain even more ... i'm not sure what b+ you were planing to run ... getting the parasites under control could be a huge challenge ... maybe using 2 / 4 to drive 8 / 10 as a 2 stage amp with a different gg topology (the control grid tied back to filament or some impedance between cathode  and ground - see Bill Orr WCHB 14th or so for a 4-400 gg discussion) .... I think gg may be too difficult to do in one stage and if one stage is what you want, then consider grid driven (especially a variant of G2DAF) ... most interesting ...73...John

[K1JJ]

hi Tom ... 12 - 813 in parallel would be a visual marvel but VERY difficult to get each tube to equally share the load ... at the very least you would need to use some resistance in the grids to ground ( or individual bias adjustments) to equalize the individual tube gains ... unfortunately this would tend to lower the gain even more ... i'm not sure what b+ you were planing to run ... getting the parasites under control could be a huge challenge ... maybe using 2 / 4 to drive 8 / 10 as a 2 stage amp with a different gg topology (the control grid tied back to filament or some impedance between cathode  and ground - see Bill Orr WCHB 14th or so for a 4-400 gg discussion) .... I think gg may be too difficult to do in one stage and if one stage is what you want, then consider grid driven (especially a variant of G2DAF) ... most interesting ...73...John

Some good points there, John.

Yes, I wondered myself about the difficulty of setting all the tubes to similar idle AND pull current equally under load. In GG for idle, that wud be tough cuz they are all fed with a common CT fil xfmr with diode biasing. To add individual diode bias wud mean separate fil xfmrs  or diodes in the fil leads which wud then upset the fil volatge... sigh.  New tubes would help the balance compared to random used ones, so that's why I'm looking at the Russian tubes.

Individual bias pots wud be the way to go in grid driven service. They wud be easier to drive with higher gain too.  I never really considered anything but GG to this point.


I didn't want to go grid driven cuz of the good chance of taking off - and it wud be a cleaner amp running in GG.  They could be neutralized, grid driven - wouldn't that be something.

I want to run 4KV to get the plate impedance up as high as possible and of course give it some kick.

I dunno. To see them under Plexiglas with a mirror behind them wud be cool. I also like the heat and quietness for Winter. But the technical problems make me wonder if it will be a waste of time.

I've always advocated one of two tubes max in a linear.  This wud be breaking all the rules... :-)

T

[w4bfs]

yup, this agrees with my experiences .... most folks ignore the G2DAF topology since the importance of the screen voltage filter time constant is poorly understood ... a lot of drive power is converted to screen voltage ... typical drive power of an G2DAF 813 is also about 30W / tube  with about 25 or so of those W as converted screen V ...The inportant thing is that the Varney circuit also has a 6 db or so distortion improvement over straight grid driven and approaches gg distortion performance ... the screen voltage could be generated by another method ... if you are interested, I will gladly share ...73 ...John