Re: Rebuilding VTTC
RoV IW3IPD, Mon Dec 28 2020, 03:40PM

Hi Terry, my thoughts:
1. you will overload a bit the transformer, hopefully getting a bit more power. Shouldn't be a great problem
2, 3, 4: all those caps are necessary to bypass RF, but the voltage rating for C4,C5,C6 can be lowered. Since 1 nF has a reactance of 447 ohm at 356 kHz, I would suggest a higher capacitance for C4/C5, like 10 nF, but a 250 V rating is more than adequate for them: they are required, because the cathode RF current flows on both heater wires and must be bypassed to GND. I would also at least double the capacitance of C2, but in that case the voltage rating is ok.
I would try a smaller number of turns for the grid winding, after all you need a peak grid voltage around 1/10th of the plate voltage. Besides, I would wind them in the bottom side of the primary. You could try with a separate temporary winding with - say - 5 turns and see what happens.
Changing the grid coil will require adjustment also to the grid resistor, where I would expect a lower dissipation.
Nice job, thanks for sharing!
Roberto

Re: Rebuilding VTTC
Terry Fritz, Wed Dec 30 2020, 12:14AM

Hi Roberto,

Many thanks for your input!

I will leave all the caps in. I won't change any of the values yet since it works right now and I don't want to break it. :)

I now see what the caps are doing. The same things those NST filters were doing to filter the RF out of the outer windings of NSTs.



I should have remembered what "that guy" was doing there! 0:D






I have the ability to directly measure currents and voltages so I might play around seeing what all is there.

I bought this coil like 20 years ago on Ebay and it has always worked very well. But I really don't know how it works so I thought I would fix it up and figure it out now.

I actually have a computer simulation of it going now.





It's very preliminary but it sort of runs and allows me to fiddle with things while I learn. After messing with the model I refined the values and it should actually be running at 407kHz.

Antonio's INCA program was super useful!





I got the secondary measurements all wrong since it was installed in the coil and very hard to measure in place. It is actually 800 turns of #27 wire and the INCA program found this error right away!

So I think I am off to a good start figuring this coil out. :)

Best Regards,
Terry

Re: Rebuilding VTTC
RoV IW3IPD, Wed Dec 30 2020, 09:21AM

Hi Terry,
I think the simulation model lacks a diode with a resistor in series from grid to GND, in order to model grid rectification on the positive side of sinusoids.
I have read your notes on NSTs... When I was young (I suppose it was 1979), for my first TC I used an ignition transformer coming from an old centralized boiler. I made the mistake of connecting the secondary in parallel to the cap, with the spark in series with the coil and in this way I burned one of the two secondaries. I had to drill through the resin to remove the damaged winding and save the remaining half transformer... fortunately it was still usable.

Re: Rebuilding VTTC
RoV IW3IPD, Wed Dec 30 2020, 09:43AM

I have found one photo with my first TC , from a demo of a few years ago. It still works.
It's the second from the left and the missing secondary is still visible. There is also a paper-oil capacitor to limit the primary current and avoid a resistor on the HV.
1st and 2nd from the left are mine, the others belong to a friend.

photo

Re: Rebuilding VTTC
Terry Fritz, Thu Dec 31 2020, 06:40AM

Hi Roberto,

Thanks for the tip about the grid diode! I never thought about that before.

The models are starting to look sort of real now.




I am sort of guessing about the grid and anode resistors. Any idea as to if that is close or not?

I tried to make real tube models using current sources and all to match the real tube specs. But I got overwhelmed and it would not run yet. So I bailed out back to using just the simple relay for the moment now.

I just got all the parts in to fix the coil so I will work on getting it all running nice.

Once it is running I can use voltage and current probes all over to get real actual readings and data to study! !:)
Then I can work on real tuning and all that.

This is great fun! China still makes tubes and can supply them quick (10 days over Christmas direct from Shenzhen) and not super expensive ($70 with shipping). Amazon sells all kinds of microwave parts very cheap!
Neon sigh transformers are getting very hard to find and super expensive. Not sure what the solid state guys are up to these days but tube coils are very practical right now.

Thanks so much for your help!

Terry

Re: Rebuilding VTTC
RoV IW3IPD, Thu Dec 31 2020, 10:05AM

Terry, the grid diode should be reversed.
You can find tube spice models around, specially in diyaudio forum. There are models
ready for the LTSpice free simulator, for which there are also builtin tube drawing symbols

Re: Rebuilding VTTC
Terry Fritz, Thu Dec 31 2020, 10:46AM

"Terry, the grid diode should be reversed."

OH!!!!! Shows how much I know about tubes :DD


I have run Linux for the past 20 years now so Window's LTspice is odd... QUCS was the next thing on google that runs everywhere...... :)

Qucs as been in the beta software development stage now for 16 years so that is a giant plus! :D

Of course, all the simulators are just the same. Same s***, different decade... :D

What really matters is getting the real readings to match the simulators. That's what really counts.

Simulation are fun time wasters while waiting for parts to arrive, but the parts are here now!!!

Stay tuned :)

Terry

Re: Rebuilding VTTC
Terry Fritz, Sun Jan 03 2021, 03:35AM

Hi,

I have been playing with my toob coil things.

Mostly working on the primary circuits.

This is the latest schematic model.



I went through 4 circuit simulators now……... and LTspice is working the best. All the other suffered ‘bazaar’ failures… :p
I run in Linux Windows simulators, so yeah...

Working on the C2 1nF original filter cap thing. Without any cap there is like 100Vac riding on the rail (blue trace, ignore the yellow stuff). The microwave transformer seems to have about 170uH of reactance.



So I simply stuck a 10nF 2000V MMC cap on it and the noise is vastly reduced!



The inductance of these transformers makes them worthless for draining RF and also makes then susceptible to damage from RF….

MMC type caps have like a rating of 2.5 Arms for decades and even though I am doing low power testing the current is use trivial for poly caps.



Just some pics for reference and for those that think I just make this stuff up. :)




So I have figured out the C2 rail cap thing… I think the caps across the plate transformer will be super easy to figure now too. ;)

So I looked at the grid voltages in the low power case….



That looks like a mess in over voltage and the computer models for the tube (so far) do not track it at all….
The model suggest that 2.5nF might be a far better tuning for the Primary cap…..
Of course this coils likes to blow things up so adjustments are expected. :D

The computer model track things like the Fo frequency perfectly (357kHz)! Even with all those mutual inductances. But other things need a lot of work…

I run at very low power right now with old tubes in case it all messes up. =:D

T.

Re: Rebuilding VTTC
RoV IW3IPD, Sun Jan 03 2021, 10:26PM

nice!
But the last figure (the grid voltage) doesn't seem taken on the grid: it seems taken on Cfeed, where it would make sense. Very high voltage, according to me due to an excessive number of turns on the feedback coil.

Re: Rebuilding VTTC
Terry Fritz, Mon Jan 04 2021, 07:00AM

Yes you are correct. That voltage was on Cfeed and not the grid voltage.

I have cleaned up all the high voltage and other basic circuits. The filament transformer now has 18nF caps across it that can take 3 amps each. Way too big but I had them on hand. I have added grounding to all the parts and cases. Also just cleaned up some other wiring. I know have two working 811 and two new 572B tubes that all basically test the same. The 811s are pretty beat up but good for experimental stuff. I’m working on a new primary capacitor since the old one was not that greatly made.





I have found the the primary coil form is fairly easily removable or replaceable so changing the primary inductance and the location of the feedback and primary coil is not too difficult.

Do you think the feedback coil is ok above the primary coil or should it be moved under it?

Does the inductance of the feedback coil matter or is it just a voltage thing to get to 1/10 of the anode voltage?

I can also change Cpri and Lpri if there is any advantage?

Many thanks for all your help!

Terry

Re: Rebuilding VTTC
Terry Fritz, Thu Jan 07 2021, 10:17AM

Many many thanks to all that have helped me!

The coil is all updated now.




Just running with the old 811 tubes at very low power but it is purring like a kitten now!


The tube plates are cold despite how that pic looks...

I just need too check the anode voltage limits now and the grid currents and voltages.

Then plug in the normal running 572B tubes >:D

It no longer feels like it is going to blow up in my face :D

It happily busts out corona right of the minimal variac twist now!

I have learned with this coil that the primary "Q" is super low (0.3) so precise primary tuning is a don't care unlike spark gap things...
These coils pretty much tune on the fly at every cycle!

A lot has to go into the care and feeding of the tubes though....

Terry

Re: Rebuilding VTTC
RoV IW3IPD, Sat Jan 09 2021, 10:53PM

wow... very good!
Regarding the feedback coil, yes it is most a voltage thing related to turns ratio. I would place it on the bottom to avoid risks to the grids in case of sparks and to avoid electric rather than magnetic coupling (electric field is quite high above the primary).

Re: Rebuilding VTTC
Terry Fritz, Sun Jan 10 2021, 11:27AM

So this is the latest schematic.



I have not figured out a really good model equivalent for the tubes yet… But going with the computer modeling army I got. LTspice is doing very well :)

Just updated a few values slightly for the new C2 and Cpri caps and Rfeed measures slightly lower. But nothing significant.

I was working on the unloaded case of Vrail traveling up and down 150V over several minutes…
No model predicted that and when actually scoped it seems to be just switching noise off D1 or prevailing wind noise charging C1 with a DC offset. Pretty darn weird actually…
I was worried it was an LC resonate thing on the MOT and C1 but no. It has always done it. I thought replacing C1 And D1 would get ride of it – but no… Very weird, but very real…..
That could use more investigation to track down since it really could damage something……

So I pulled out some HV probes out of my very dusty toy box…



I have not used some of these in 20 year so I had to verify that they and “me” were still sane. Had to remind my self of the voltage ratios… When working with 10kV you don’t want to mess up by a factor of 10X :D

So after the voltage probes were figure out I put a big one on the anode voltage.



But the voltage and stuff seemed super high at low variac setting.



I have turned it up far higher without actually looking at the voltages… Maybe, I was overloading something, again :D I am thinking once the streamers get high the loading takes the voltage way down...

So that is where I am at.
I have a lot of knowledge about Tesla coils, high voltage, test equipment, theory…

But rest assured, when dealing with tube coils I am definitely playing with things I do not understand at all! :D

Terry

Re: Rebuilding VTTC
RoV IW3IPD, Sun Jan 10 2021, 08:28PM

well, peak anode voltage without loading should be close to double the power supply. The power supply architecture is a doubler, so 1900 Vpk RF mean 950 Vpk from the power supply and 425 Vpk at 60 Hz. Don't know how your variac was set, but seems reasonable. May be too much for those tubes, but this is another problem...
Didn't understand the first part, the weird problem...