Tube circuit design as art

Tube talk

Tube circuit design as art

Postby Doug Coulter » Sat Jul 17, 2010 12:04 pm

I couldn't resist putting this link up, which was provided by Joe Sousa. Now, this is art!
Some EE's really appreciate elegance in design, so here's some good examples.

http://electronbunker.sasktelwebsite.ne ... Super.html

And here is another more-complete one with a similar thrust.

http://www.radiomuseum.org/forum/russia ... tubes.html

I very much like the style there -- enough content to really cover a topic insofar as the author can, then responses that tend to be shorter.
If you have something like this -- start a thread/topic on it, and you can back edit your own post, wiki style as you like to make it definitive.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Re: Tube circuit design as art

Postby William A Washburn » Sun Mar 20, 2011 3:16 pm

Thanks Doug for including this section,

This is where I got my start when I was 10 years old. We were poor and a few of my relatives
who had some money had radios (the old ones of wood that looked like little churches) and if
their radio died and the local repair chop couldn't fix it it would be given to me. My father
let me borrow a soldering IRON he had at the retail paint shop he ran and that is how I would
remove the parts that were still good and built little projects. A little one-tube amplifier here,
a superregen receiver there and, of course 120VAC driven power supplies. I also learned
respect for electricity this way. I guess the fact that I'm still alive may have something to
do with Darwin.

Again, thanks.
That was the way to learn, Bill
User avatar
William A Washburn
 
Posts: 93
Joined: Fri Oct 15, 2010 9:12 am

Re: Tube circuit design as art

Postby Doug Coulter » Mon Mar 21, 2011 10:22 am

I'm only sorry we don't have more posting on this one. I was making superhets at age 8 or so (1961) and thermins, under my dad's tutelage. It was only later that he managed to get me a CK-722 to learn with. But the early Ge transistors were....well, early Ge transistors. Very interesting, but not so hot, so I kept messing with tubes myself. Now that I'm on solar power (and semiconductors are so much better), not so much. However, there's still a lot to learn from tubes, and things like virtual cathodes in beam tubes have some relevance to fusors and such. After all, a charged particle is a charged particle (mostly) except for scale. Only electrons flow in a normal (semi) conductor of course -- that whole hole thing is kind of a convenient fiction.

I will be adding to this at some point, because there are still some things that only tubes do best, and some of them are planned in my fusion project. The two uses that come to mine right off are in-tank preamps with small pencil electrometer tubes, and RF/HV sorts of electrode drives. If you work things out, it's pretty hard to put a modern op-amp into a vacuum tank with a fusor and have it live for more than seconds. Neglecting that the entire tank sometimes gets to hot it would fry a semi even if glued to the wall, it turns out that any decent opamp can't really get rid of its own heat well enough. In a situation where the measurement would be affected by the leakage and noise pickup implied by sending a probe signal through a feedthrough, a little preamp is needed...but no semi's need apply there.

For doing this set of experiments where I'm perturbing the fields the fusor operates in to see the effects, I need a broadband high voltage source. Due to the nature of the beast, it's more than possible for an arc or at least some large conduction between the real HV and any electrode put in there to affect the field. So a solid state RF amp would be toast at the first accident unless some very heroic measures are taken to protect it -- might be impossible to make it really robust. So I plan an RF driver that will be tube based at least for the output stages. For that I'll be using some old, large, horizontal sweep tubes, 6kd6, which were kind of the last gasp for consumer power tubes for non-transmitting use. Hams use these to make linear amplifiers, and the tube specs are quite amazing actually. They can stand off 7kv or more, while also being able to pull over an amp of plate current (not both at once). And like all tubes, transients just don't bother them unless they last long enough for something large to physically melt. For that job, I can afford to ignore the relatively large filament power they need (about 20w each).

Should be a fun design, in that it will probably have a solid state switcher for the main B+, then driving tubes. I remember a day when those roles would have been switched more or less --
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Re: Tube circuit design as art

Postby vmike » Tue Mar 22, 2011 10:02 am

Sweep tube amps can be nasty little beggars to make drive a capacitive load. I have a couple of commercially built RF Plasma generators I plan to use for similar experiments, like driving my cyclotron. One uses a 4-400A, the other uses a 450TH (a little older than the first). The sweep tubes will work if you have spares, but they are not rugged, at all. I had a radio shop in the seventies and was about the only guy in town that would work on those types of amplifiers, mostly because the run of the mill sweep tube amp was junk! A good transmitting tetrode will probably do you better.

I also grew up with tubes and love them. I usually tell about the first tube amp I built at 9 years (1962 by the way), then tell about the one that worked properly built at 12. I built a lot of one and two tube radios as a kid, always from salvaged parts.

I've built a number of tube amps since and several tube type tesla coils and plasma speakers. Even some ham transmitters and amplifiers. I'll take some pictures and post them to the thread, I'll also take some of a few of the neat looking tubes I've collected.

Fun topic,

mike
I tried to contain myself, but I escaped.
User avatar
vmike
 
Posts: 30
Joined: Wed Sep 15, 2010 5:20 pm
Location: Stillwater, Oklahoma

Re: Tube circuit design as art

Postby Doug Coulter » Tue Mar 22, 2011 10:56 am

Good to hear from you again, Mike.

I've had pretty good luck with this particular sweep tube, not so much the others. Compared to your basic 6bq6 or similar, a 6kd6 is a monster (3x the specs) - they were only made at the very end of the big screen tube TV/CRT era. You just have to work within it's limitations (like anything). My first big use of them was a tube tesla coil thing that used 6 in parallel, no special issues. No question a modern tetrode could be better (but higher knee on the plate current curve, almost every tube has a higher knee than this one), and would probably take more drive effort. The 6kd6 (depending on the screen volts) needs under 100v grid drive for full power and full cutoff at HV on the plate, and I'm planning to drive it with some semiconductor VGA cathode driver chips (but drive the control grid) I found that do 100v at some ma. That gives me a broadband driver, no tuned circuits at all there. The problem no matter what device I use is really the magnetics. Until I have things nailed down to inside an octave (and I may need non-sines here) that's the big issue. Since I'll probably only have about 3kv pk-pk plate swing and need some stepup (I think, this is uncharted turf) that's a real problem. Broadband magnetics are one of the blacker arts. Funny how audio guys easily do 10 octaves, where the RF boys get all proud with a fraction of one.

Of course, one "advantage" of the glass bulb on this is you can see when you're getting the plates too hot. We picked up a 1.5kv switcher supply originally meant for cap charging for lasers, which looks nice for this. I have the heavy old iron as well, but if I can make this efficient, why not? I plan to use 4 tubes here (after a one tube prototype) as that switcher has a 28v 8a aux output to run things like filaments. The 6kd6 is quite a power hog there.

Could it be that most of the sweep tube amps you saw, being homebrew by cheapskates, simply weren't built that well? I've beaten these puppies like a rented mule and they don't seem to have any real bad issues. Easy to neutralize in the bargain, and few parasitics. You can mostly pretend those issues don't exist if you have the old 47 ohm R and coil in the plate and make the circuit "tight" re RF and grounding issues.

Note the dual pinouts on the grids to help with parasitics. I'm guessing the designer had a little glint in his eyes when he did that -- no need for that for the intended use!

6KD6.pdf
6kd6 datasheet
(383.37 KiB) Downloaded 209 times
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA

Re: Tube circuit design as art

Postby vmike » Wed Mar 23, 2011 9:21 am

I have never seen a 6KD6 amplifier so they truel7y may be better. The amps I worked on in the middle 70s were the commercially made CB amps, marketed of course as "ham" amps. They were built as cheaply as possible and sold for more than they were worth. I have ham transceivers that use sweep tubes without problem, drake, swan, wrl, heathkit, etc.. I'm sure I have some 6KD6s, I'll dig one out and make an oscillator out of it and see if I can burn a hole in the plate. Just kidding, :) It was the common failure mode for sweep tubes, even in televisions.

mike
I tried to contain myself, but I escaped.
User avatar
vmike
 
Posts: 30
Joined: Wed Sep 15, 2010 5:20 pm
Location: Stillwater, Oklahoma

Re: Tube circuit design as art

Postby Doug Coulter » Wed Mar 23, 2011 1:56 pm

Most of the (truly ham) linears I've seen based on them use them with all the grids tied together and grounded, with cathode drive input. In that mode, they are triodes that don't need much bias even with a kV on the plates, and are pretty stable, from low to about 6 meters. My first try was with that config, but I just made it oscillate at the resonance of the HV coil I was driving, no particular problems with that. For that I used 6 in parallel, like the ham design.

Here I will be attempting to use them as beam power pentodes, a different story altogether, and controlling the amount of grid drive volts I need by controlling the screen volts. More power gain since the driver will be this little TO-220 chip from VGA monitors and really can't put out much power (about 1 watt). I've built and tested the driver -- it's phat to 75 mhz or so, and so suits the driving signal source, an arbitrary waveform generator, a DDS-3X25, which by the way is a pretty nifty toy itself.

In windows (or an emulator) you can just draw waveforms on the screen and have them output at any frequency, all in realtime. Neat!

Like all things CRT related (yeah, I know I've put out this rant before) the drive chips are going the way of the dodo -- so I bought 100 of them for future needs. Lm-2445 is the number, and here's what a 3 ch DC-75 mhz driver looks like. Takes about 3v in for 80 or so volts pk-pk out.
VidDriver.jpg
Vid Driver prototype

LM2445.pdf
Data sheet for lm-2445
(755.58 KiB) Downloaded 203 times

PA-1KW.pdf
Typical ham sweep tube amp
(567.18 KiB) Downloaded 187 times
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 2964
Joined: Wed Jul 14, 2010 8:05 pm
Location: Floyd county, VA, USA


Return to Legacy (Tubes!)

Who is online

Users browsing this forum: No registered users and 2 guests