A few questions about base fed 'vttcs'. / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

A few questions about base fed 'vttcs'.

1 2 next

Move Thread

LAN_403

Arcstarter

Mon Jul 04 2011, 01:25AM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Well, these so called 'plasma tweeters' may not really be Tesla coils considering there is no primary or magnetic coupling, but it does still utilize resonant rise, so i guess this is the right forum.

The basic schematic i am referring to in this thread is:

Basically a resonator tied to the plate of a vacuum tube and some sort of capacitive feedback. I am using a GU-50 (40w dissipation, 1kv plate voltage at 6.5m wavelength), and 700vac fullwave rectified and smoothed. Here is a video of it running with some resonator i whipped up.

Anyway, i have made many different plasma tweeters, with both triodes and pentodes, but i have noticed that the resonator's dimensions are hugely important. For example, i have a small resonator about .65 inches in diameter and 1 3/4 inch tall wound with 24 gauge that ran around 30MHz. This gave me about 3/4 inch flame, which i was not very satisfied with.

So, i made a larger resonator that was 2.25 inches in diameter and an inch of 24 gauge resonating around 10MHz. I hoped the lower frequency would give me a spark with somewhat less current density, but more length. Instead, the spark was less of a candle-like spark and more of a brush discharge, and even shorter length.

I know there are a lot of factors, such as feedback, impedance matching, spark dynamics at different frequencies, and too many others to count. What i am wondering is how can you figure which dimensions will give the best spark length? I use the same schematic as this guy, and the same method of feedback, and i have made resonators just like his, but i cannot come close to his coil:

I know he puts sodium on his breakout point which helps make the spark longer, but that is still one hell of a plasma flame considering the tube used.

If i am using the same tube, the same basic schematic, approximately the same feedback, and a similar resonator, why is the output so much different? I know you can use some of the various inductance calculators to calculate a resonators impedance at resonance and match it to the tube's plate impedance, but will that actually make that dramatic of a difference? Also, what about the RF inductor? I guess you would want to get that as close to the resonator's impedance at whatever frequency it is running as possible? The problem is that there is not a whole hell of a lot of in-depth info on these curious little gadgets.

Any ideas or suggestions are much appreciated!

Dr. Dark Current

Tue Jul 05 2011, 02:22PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Hi,
I don't like the capacitive feedback because you don't really know how much voltage the grid is getting.

This is the circuit I was planning to use with my plasma flame oscillator, it uses just a single resonator coil (no other coils or tuning capacitors). The topload may be important for stability.
(sorry for schematic quality)

edit, circuit fixed

Arcstarter

Tue Jul 05 2011, 07:47PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

I like that idea. I have used an isolated feedback coil:

This definitely works better than capacitive feedback. Spark length and plate heating are both very dependant on the capacitance of the feedback, that is why there are alot of people with big tubes with nearly molten plates and barely achieve breakout.

Not to mention, it is incredibly hard to get the right amount of feedback for good performance without getting arcovers. This problem is even worse with more power, because even though the output is larger, it still needs a lot of capacitance.

Yet another factor is noise. If there is not enough feedback, there is a lot of 60hz hum . If i move my 'air capacitor' plates further from each other, the spark has more 60Hz hum. This is probably because without enough feedback the tube does not turn off 'fully', and the load causes more ripple. I only had 11uf of smoothing at 800-1000v, but when the spark was longest, the ripple was mostly gone. I will use more capacitance when i am past the testing stage. It may be some 60Hz being coupled into the feedback though, i don't really know.

Just yesterday i found out my GU-81M works (you've probably seen the thread), so i think i will try your idea instead of the typical capacitive coupling. Either that or an isolated feedback coil, but i really like the idea of using the resonator as the HF and feedback coil. I will hopefully experiment with the two, and i look forward to seeing any progress you make with your idea.

However, i still wonder how to calculate the HF inductor. I know what purpose it serves, but im unsure if it's impedance should be the same as the resonators at whatever frequency is resonance. That is how most people make theirs, but i usually wind some arbitrary coil and it still works with fair performance.

One last thing. If you wind the whole resonator/inductor/feedback coil as a single solenoidal coil, would the coupling be correct? That is definitely something that would have to be experimented with to get perfect feedback, but if anyone has any opinions i would like to here them. It's better to have an educated guess than to blindly kludge things together.

Dr. Dark Current

Tue Jul 05 2011, 08:38PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

I was definitely going to wind the whole coil on the same diameter with the same wire, the best coil for this would be a rf coil for ham radios (or whatever), there are some of them on Ebay but not very cheap.

The topload should ensure a quite linear voltage gradient across the resonator, with almost none phase shifts.

As for the series inductor, it just should have a big enough inductance - it serves as a current source.
Edit: If I think about it now, that may not be 100% correct (about the inductor as a current source).

Arcstarter

Tue Jul 05 2011, 09:31PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

I just assumed the inductor's only duty was to provide a high impedance at the high frequency oscillations. Without this, you would effectively short the DC supply through the tube, which i do believe is bad

. Wait, i think i just explained a current source...

About the topload, i think you are onto something there. I never used toploads for these kinds of coils, to make sure the resonator was as high frequency as possible (i just love how high frequency sparks look). But, i think there is more to the fact that i got longer sparks with some topload than just because the frequency was lower.

Edit: I made a feedback coil inside of my resonator. It is so much more convenient, and the performance is about the same, perhaps a little worse. Here is a video of capacitive feedback:

And here is one of the magnetic feedback:

There was too much feedback, as you can see. It was about 200v.

Another problem i am running into is that i doubled the voltage into the MOT, but the spark still looks about the same length! It is around 1500v on the plate now, and the spark is completely unsatisfying. I still have 300-350v on the screen, i will increase that to around 500-600v at some point. I used only one turn for feedback this time, so the voltage on the control grid sticks around + - 120v peak. That still doesn't explain the performance though. Maybe the primary voltage on the MOT is sinking somewhat, but surely not enough to make performance *this* depressing.

Arcstarter

Fri Jul 08 2011, 12:04AM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Well, the spark length vs input has puzzled me to the point i felt it necessary to take yet another video. I show the difference between 40, 80, and 110v input. As i say in the video, i have made the plate start to change color. Don't kill me guys, i did not even think the step-down transformer from a 120w stereo could supply enough power to do that, especially since it usually make crackling noises if i overload it! Poor epoxy stuff must be cracked as hell

I think this project has raised more questions than any other project of mine... Including drsstcs.

Dr. Dark Current

Fri Jul 08 2011, 09:23AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

So what circuit are you using?
On this tube, it is OK when the plate is dull red color.

Arcstarter

Sat Jul 09 2011, 01:32AM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Dr. Kilovolt wrote ...

So what circuit are you using?
On this tube, it is OK when the plate is dull red color.

I keep forgetting to say things, i forgot to show the circuit. I will attach a schematic. It isn't the best but it will give you the general idea... Also, the cap and resistor values are approximately what i used, but the values are where chosen arbitrarily, i just used what i had that had the right voltage rating for the grid cap.

And i didn't think the plate on this tube was something you wanted to turn red, like the 833C's plate. What is the plate material for a GU-81M anyway? I seem to remember reading it was graphite, though im sure it was an unreliable source.

Edit: oops, tried to upload the bmp instead of the jpeg, and then the pic is too big to insert, so here is a link:

Nah

Sat Jul 09 2011, 03:40AM

Registered Member #3567 Joined: Mon Jan 03 2011, 10:49PM
Location: USA, 1960s
Posts: 260

I think it was a metal plate (an alloy of alunimun, if I remember), but I'm not too sure.

If graphite, it will be black, a metal plate is more silver is color.

Not counting "black/grey plate tubes"

Dr. Dark Current

Sat Jul 09 2011, 08:58AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

The plate material of this tube is indeed graphite, the silvery material on it is a some kind of coating with getter material.
Some guys have told me when this tube was used in a transmitter, they tuned the output power for dull red plate = an indication of that the tube is running at the rated dissipation.

As for your circuit, it looks pretty much ok... Try to increase the grid drive, put 2 turns on the winding and maybe a smaller grid resistor. If that doesn't help, try using a topload.

Have you tried reversing the feedback winding?

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob