My first coil : 572B VTTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

My first coil : 572B VTTC

previous 1 2 3 4

Move Thread

LAN_403

PhilGood

Sun Jun 30 2013, 10:19PM

Registered Member #3806 Joined: Sat Apr 02 2011, 09:20PM
Location: France
Posts: 259

Hi Jan,

That's really interesting if your formulas verifies on most of your VTTC's with very different setups ! By any chance, do you have some kind of summary of all you VTTC's on a blog or personnal website ?

For a k of 0.32 and L of 122ÂµH I need 67 turns of Ã˜ 1mm enameled copper wire on a Ã˜ 50mm form (calculated with JavaTC)

With the +/- 40% for tuning (73ÂµH to 170ÂµH) it gives a total of 88 turns for a total primary height of 88mm, with taps from turn 45 tu 88

I will try this setup and let you know how it performs ...(I should receive my vaccuum variable capacitor next week and this might help a lot in fine tuning)

There is another point we didn't discuss yet: grid leak circuit. The actual values where determined by experiment for the best output and less heating (before I added the voltage doubler). But I recently read somewhere that the grid leak circuit time constant should be "more than a RF cycle, but not too long". with my actual values, I have a time constant of around 22ÂµS, which is around 20 RF cycles at 2MHz. Could that be a problem ?

I was also thinking to something: isn't there a way that would allow the tube to see the correct loaded impedance and still have primary and secondary circuit tuned to the same frequency ? If this is possible, wouldn't that give a better output ? (Maybe by connecting the tube anode and tank cap to a different primary tap ? If I remember well I have seen such setup for a GU81M VTTC)

Thanks for your kind help

Dr. Dark Current

Mon Jul 01 2013, 08:35AM

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

Hi
unfortunately no blog or anything...

That sounds like a lot of turns, but it should work. For such high frequencies, the formulas tend to have a bigger error so dont be disappointed if it doesnt perform as expected (but this is to be expected when experimenting

)

As for the grid leak circuit - if there is no squegging (random dropping of oscillation, characterised by crackling or squealing sound, often accompanied by rectifier and/or transformer death) then a too long RC constant doesn't matter. For the high frequencies it is often not practical to do a shorter constant. On the other hand, if it is too short (less than 2 RF cycles) then the coil loses efficiency.

wrote ...

I was also thinking to something: isn't there a way that would allow the tube to see the correct loaded impedance and still have primary and secondary circuit tuned to the same frequency ? If this is possible, wouldn't that give a better output ?

I'm afraid there is no simple way. The coupled resonant circuits will never oscillate at the same frequency anyway, see the pole splitting. If you were to design the correct loaded impedance with both circuits tuned to the same frequency, you would need to know the proper primary tank circuit Q-factor which you can never know beforehand, however you can also alter the impedance by changing the coupling, or different plate/cap taps like you said. Anyway the coil would probably tend to switch poles (unexpected output drop), because of the low total Q-factor the oscillator would be very prone to squegging and/or low oscillator efficiency, all of which would make things worse not better.
There is a different approach which is feeding the secondary base - but this is a different topic anyway...

-Jan

PhilGood

Tue Jul 02 2013, 10:10AM

Registered Member #3806 Joined: Sat Apr 02 2011, 09:20PM
Location: France
Posts: 259

Hi Jan,

No worries, I won't blame you if things do not work as expected. In fact my instinct would tell me to choose some "less extreme" values, like using half the L and twice the C suggested by your calculations. But I will stick to your formulas, I'm really curious about the result :)

Thanks for the detailed explanation about grid leak RC constant, this is another very valuable info I added to my own VTTC documentation.

Dr. Dark Current wrote ...

I'm afraid there is no simple way. The coupled resonant circuits will never oscillate at the same frequency anyway, see the pole splitting. If you were to design the correct loaded impedance with both circuits tuned to the same frequency, you would need to know the proper primary tank circuit Q-factor which you can never know beforehand, however you can also alter the impedance by changing the coupling, or different plate/cap taps like you said. Anyway the coil would probably tend to switch poles (unexpected output drop), because of the low total Q-factor the oscillator would be very prone to squegging and/or low oscillator efficiency, all of which would make things worse not better.

About pole splitting, we could detune the primary to the lower frequency pole. And according to what you say impedance matching seems to be possible under certain conditions...(including luck ?) I wonder if the DRVTTC might be an interesting area to explore ^^

I also wonder if, let's say VTTC's were something useful built at an industrial scale, if engineers with their actual knowledge and tools, would be able to design a VTTC and get the exact result expected.

Dr. Dark Current wrote ...

There is a different approach which is feeding the secondary base - but this is a different topic anyway...

I have an ongoing project on this topic with a GK71 tube... wish days would be 48h

I also fixed something on my secondary: I noticed I had way too much resitance (more than 100Î©) betweent the bottom of my secondary and my top load. I found out it was due to a piece of adhesive copper foil that was between the top of my secondary and the top load. I fixed that and I now have 9Î© between the bottom of secondary and top load. I don't know if this was a problem, the HV was probably arcing through the copper foil's glue, I will see if this fix improves things.

This changed my secondary resonant frequency, it is now 2.430 MHz (weird, I thought it would rather lower the frequency)

I made a small Javascript app to make calculations faster, the new value I get are C=80pF and L=109ÂµH

I really appreciate that you spend time on this topic, going that deep into understanding VTTC operation is fascinating :)

Philippe

Dr. Dark Current

Tue Jul 02 2013, 05:39PM

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

Hi

No problem, but please keep your calculator private. I don't want everyone to have access to my knowledge and experience, which took years to gather, and some kid just inputs the data and builds a great VTTC. I don't want that, I like that I can use my knowledge to have some advantage before others. Of course I do help, but only people which show effort and have some previous experience from experiments they did themselves.

Cheers

TwirlyWhirly555

Tue Jul 02 2013, 07:50PM

Registered Member #4104 Joined: Fri Sept 23 2011, 06:54PM
Location: Uk .
Posts: 122

Interesting Read this thread has become :)

I have tried myself to calculate and make a VTTC from the data . though each time I have found that in the end manual tuning of the primary was needed to get the best results / feedback resistor/cap .

I never really played with the feedback turns just used a ballpark figure , usually only ever touched primary and grid cap / resistor .

PhilGood

Tue Jul 02 2013, 11:57PM

Registered Member #3806 Joined: Sat Apr 02 2011, 09:20PM
Location: France
Posts: 259

Hi Jan,

I understand, no worries about the calculator, I made it for my own use, it saves me a lot of time
I would never make someone else's work public without their prior authorisation

Thanks again for your kind help :)

Philippe

previous 1 2 3 4

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