Retarding fields in tubes

Tube talk

Retarding fields in tubes

Postby Joe Sousa » Sun Nov 07, 2010 6:37 pm

In my travels in the tube world, I have been particularly fascinated by the effects of retarding fields. A retarding field is formed when a negative potential slows down the path of electrons. In the context of a tube, a retarding field exists when electrons pass near an electrode that is biased at a lower potential than the previous electrode, or when electrons approach a pre-existing space charge, as might have been accumulated by the beam plates of a beam tetrode, or right at the cathode, when thermionic emission is thwarted by the accumulated space charge of electrons around the cathode.

The effects of retarding fields on tube operation are quite varied: There is the beam plate action in a beam tetrode, there is the effect of plate potential when it is lower than the electrode adjacent to it, the third and last grid in a pentode which is usually designed with a sparse pitch to produce little retardation in electrons passing it, may have a more pronounced effect if the potential is sufficiently negative. Even an incandescent light bulb with a hard vacuum and a few hundred volts driving the filament may exhibit spontaneous VHF oscillations as electrons are accelerated and whip-lashed back as they travel from the negative side of the filament to the positive side of the filament.

I have explored some of this phenomena from the point of view of an antique radio hobbyist at the following links.

http://www.radiomuseum.org/forum/dual_c ... todes.html
This link deals with the effect that the third grid (suppressor) in a pentode has on plate and second grid (screen) currents. Follow some of the links within this page for a further treatment of the topic.

http://www.radiomuseum.org/forum/rustik ... ments.html
This link deals with VHF oscillations in an incandescent light bulb, and may have an incidentally closer connection to space charge behaviour in a fusor.

I would love to hear about any insights that come from the perspective other than that of of an antique radio-hobbyist.

Regards,
-Joe
Joe Sousa
 
Posts: 5
Joined: Wed Jul 14, 2010 10:51 pm

Re: Retarding fields in tubes

Postby Doug Coulter » Mon Nov 08, 2010 4:15 pm

Actually, Joe might be well ahead of us here, in a least one sense. A lot of "how a fusor is supposed to act" when you think about it, can't happen without some explanation similar to his oscillating light bulbs. Else we would simply get to a static equilibrium real quick. A lot of people have talked about recirculation and oscillation but without anything more than the classic:

Step two: a miracle occurs.

A lot of this is because the configurations and effects we know about that can directly change DC to AC in a vacuum tube don't seem to have much if any applicability to our situation. It's not a magnetron, klystron, TWT and so on. It is what it is. And so is Joe's bulb. Perhaps there is a relation.

In our tests we haven't seen RF coming out of fusors here, but we have seen some oscillatory modes of different sorts than what Joe describes, though his insight may be of use. Our trouble is that the soup we operate in has two (at least) flavors of particles, with two polarities and two different (at least) charge to mass ratios, and both see each other as part of the net field in there, which on measurements here has not as much to do with the applied field as many assume. At some point I'll put up some data on that, when my measurements are worth posting about, but it's not like what you would have with a grid in a vacuum, or even close to that.

We have found some pulsed modes here, but they are more relaxation oscillator-like than something that makes sine waves, and I kind of doubt the causes are the same as even a normal gas relaxation oscillator as it acts so differently though it produces non sines. Instead of ramps or sawtooth waveforms, we see pulses more like an avalanche in our rig, when all the other parameters are right for this, which include the proper series impedance in the drives, or appropriate waveforms introduced on other electrodes to create the situation.

Thus, rather than us helping him explain his oscillating light bulbs (though I made the attempt) it may be Joe who helps us understand the somewhat different, but surely related effects we see!

We also see what sort of has to be a dynamic equilibrium in normal, stable fusor operation. Whether any actual recirculation of ions happens is not answered, however. It seems to me that electrons emitted merely are accelerated to the tank walls, and often ionize some neutral gas on the way, which keeps the thing "lit" and certain geometries accomplish this better than others. I also think I have a lot to learn about that, and am taking what measurements I can to find out more.

I have often thought of a standard fusor as an inside out triode, with the grid being sort of a screen grid. The plate is the middle focus, and the emitter is the tank walls in this model, though it's not that great a fit. But given the direction the ions travel towards a collision at focus (if that's even where it happens mostly) it would be a decent fit. Which implies that tons of work with normal vacuum tubes (including beam tubes) might be very applicable here. I would personally hate to see nearly a century of work with charged particles get tossed in the trash just because here we are working with ones of the other polarity and more mass. To be honest, if my Faraday probes are telling me the truth (likely) then we are working with far more electrons than we are D+ ions, and for that matter, usually more D2+ ions than D+.

Bussard started some discussion about virtual electrodes and the effects of space charge, though I have some real issues with the rest of his analysis, and the truth is, no Bussard machine has done as well as we have, or some other approaches. But I feel that looking at things like space charges and virtual electrodes due to that and particle inertia are the way forward.

Here's the classic cartoon from Physics Today for reference.
Miracle.gif
Classic cartoon
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: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Retarding fields in tubes

Postby Joe Sousa » Sun Nov 14, 2010 11:32 pm

The light bulb oscillations only have electrons as the carrier of electric charge. However, there are two distinct modes of behaviour. One is the counter rotating oscillations first described by Holmann, the other is the charge accumulation on the wall, which serves as a floating plate to couple in my experimental FM modulation, and also reinforces the retarding field beyond the anode end of the filament.

I just posted my Tomato-Soup HI-FI AM transmitter at
http://www.radiomuseum.org/forum/hi_fi_ ... itter.html
This transmitter design draws heavily from lessons learned at the RadioMusuem.org forum, and uses the suppressor grid G3 retarding field to effect a very linear two quadrant multiplication between Audio applied to it, and the oscillating current from the cathode. The resulting product appears differentially between the plate and screen.

Regards,
-Joe
Joe Sousa
 
Posts: 5
Joined: Wed Jul 14, 2010 10:51 pm

Re: Retarding fields in tubes

Postby Doug Coulter » Mon Nov 15, 2010 10:24 am

I personally love this kind of thinking, using tubes in ways mostly not anticipated by the designer. With clouds of charge moving around it's not as fixed function a thing as a transistor, and there's more possibilities to play with for someone who thinks a little sideways.

I consider all this work relevant to what we are trying to do in fusors -- we are just working with the other polarity and more mass/charge ratio, but the same laws apply given various scale factors. That's one large body of work in tubes we shouldn't be ignoring.

For that matter, I miss UJT's, dual gate mosfets etc, where some of that same kind of fun could be had. At least the chip guys can have multiple emitters and collectors, but us poor schmucks don't get to play with those.
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: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Retarding fields in tubes

Postby ModeratorNuke » Tue Nov 16, 2010 10:28 am

I hate doing this, but had to delete some off topic posts here. Try and keep it more on topic. Joe did something very cool here (though it'd be more obvious with a pic and schematic here than just a link to elsewhere) and deserves better, as does the topic.

I know of no other instance, in all of EE, of being able to get a single active device to be a high quality, stable, am transmitter (and this one can do DSB/reduced carrier too). In fact, there's quite a few words in the ARRL handbooks saying it can't be done -- and a lot of people have tried, and failed, even with crystal control.

We very much don't buy into any thought that you don't have to know the fundamentals at least before moving on here. If that's your cup of tea, there are many other boards where people who don't know opine endlessly about things they know little of. Since this is all meant to be a teaching document set, things that are known to be wrong will be removed before they mis-teach someone else, unless they represent a good excuse to teach what is correct.
ModeratorNuke
 


Return to Legacy (Tubes!)

Who is online

Users browsing this forum: No registered users and 2 guests

cron