Design-and-building a large valve Tesla coil (class C, Armstrong oscillator).

[Al (astral highway)]

Thank you, Ed, p.m sent!

[ionburn]

This is really interesting.
Hi Al

I used to spend a lot of time in my youth with ignition coils etc, generating lots of sparks (how naive we were then ) and it is still with me as I still put together inverters and multipliers to reach higher voltages, although recently only up to a lowly 900 volts. I am thinking about valves and linear amplifiers here.

I take your note on current inrush. It has been a bit of an issue for me as I have used ex computer power units which can be 10 or more amps output rated but shut down at the hint of overload. This does make me think of some of the problems you are getting as I have had more problems with current than voltage so with such high voltages I can only imagine the problems.

Please keep us updated

[Ian - G4JQT]

Interesting stuff and something I've got as a 'medium-term aspirational project'!

I have a couple of coils wound on paxolin formers. One is about three feet tall and 10 inches across, the other about 18 tall and four inches across both wound with 20-ish SWG enamelled wire that's a bit chipped in places. Maybe I can just re-coat it.

I've been to the Cambridge Teslathon a couple of times and met some of the guys in Tesla Coil Builders Of the UK group and am on their email list.

I also have the 'Ultimate Tesla Coil Design and Construction Guide' in pdf if you're interested, although it's probably something you already have.

Please keep us posted.


Ian

[G8HQP Dave]

Originally Posted by astral highway

1.3MHz

Sorry to ask an awkward question, but will this device be used in a screened room or will it flatten MW (plus harmonics) for some distance around? I realise that Ofcom are not very interested in AM these days but if you draw too much attention to yourself they might come knocking on the door.

[Al (astral highway)]

Originally Posted by Ian - G4JQT

I have a couple of coils wound on paxolin formers. One is about three feet tall and 10 inches across, the other about 18 tall and four inches across both wound with 20-ish SWG enamelled wire that's a bit chipped in places. Maybe I can just re-coat it.

Hey Ian, they both sound like good candidates, although that first one would be a monster needing a lot of power in! The Teslathon sounds like great fun, I should try to get to one!

[Al (astral highway)]

Originally Posted by G8HQP Dave

Sorry to ask an awkward question, but will this device be used in a screened room or will it flatten MW?...

Well, there is no MW reception whatever in my building - a new build of several floors high, made of cladding bolted to a steel superstructure And beyond that, more of the same - a nice natural Faraday cage for 300m waves? I think some reception might be possible with a long wire on the top floor but I'm sure nobody bothers to try. The main problem around here Is RFI on FM from walwarts and SMPS.

[Al (astral highway)]

Afternote to post #26...on a technicality, it's actually our good old national treasure, the BBC (I used to work there, but not in regulation) where the public are invited to go to report complaints about broadcast interference that severely affects their broadcast reception. As I say, I hope we don't end up in a big old OT diversion here, so I'm enriching my last response to cover all the bases.

I feel confident and appropriately mindful: all that I said about the structure of the building I live in, with its earthed superstructure, inside which are spaces only around 15m deep and wide; plus the multiple RFI measures I am designing in, make electromagnetic radiation extremely unlikely ever to be perceptible to anyone else on any broadcast equipment. If I do operate the thing elsewhere , I will consider the issue and take a decision on what screening can be put in place.

Also ,for others to consider, a well-tuned model not only runs more efficiently and produces longer plasma but also produces far less interference, by a long chalk. Running a de-tuned coil with a poor break-out is when things need looking at urgently.

Cheers!

[G8HQP Dave]

OK (why does this website insist on posts being at least 3 characters, even when two will suffice?)

[Brian R Pateman]

It's a function of the forum software.

If you use punctuation you'd be OK.

[Brian R Pateman]

OK!

a nice natural Faraday cage

You have a screened room, a Faraday cage is somewhat different.

[Al (astral highway)]

Progress report and block diagram

Hi folks, I'm posting a block diagram as well as some progress photos.

- EHT supply is now complete and more robust, including a bi-directional diode, 2X062H, which imitates the current way of protecting microwave oven power EHT supplies, instead of the traditional high-voltage fuse.

When the leakage current in the oil-filled caps goes high, this component goes S/C, thus shorting the capacitors completely, blowing the primary-side fuse and instantaneously latching the consumer unit RCD - and preventing them from exploding in an oily mess. Since I have 4 of these for a total of 4uF, this seems essential. I have no idea whether the HV diode would survive such a scenario but the manufacturers of microwave ovens now clearly regard this as the best way to do things, so who am I to question it?

I have also added inrush current limiting resistors in series with each EHT transformer primary. Switching on one of these beast without inrush protection is bad enough; two in parallel is a big old inductance indeed, enough to make the lights flicker. My solution completely resolves this problem and is robust enough to survive many many operating cycles.

After 2 seconds, the two big transformers are shorted by a relay, bringing the full current into play in each primary. I have also put 100nF capacitors, 750V rated, in series with the resistors, such that when the relay switches off, a snubber is instantaneously formed, comprising the resistor and the cap in series.

I have measured the voltage across each primary and my DMM shows that it rises to a peak over 2 seconds and decays to zero in around the same time. This is a vast improvement on where I started, when the lights did a transient flicker. Please note - do not try to measure the voltage across an EHT circuit like this with any probes - it's asking for trouble. Picture shows only the primary side voltage being measured. The EHT is all tucked away on the opposite side of the chassis interior, making it harder to produce accidental user contact.

Unfortunately, the brand new relay, which is Chinese, has a fault-- meaning that I can only use one of the two 'normally open' circuits. This is not ideal but the total current is stiill within limits.

I have also put a TG-14 surge arrestor, which is a gas-discharge tube with a breakdown voltage of 14Kv, in parallel with the EHT+, to discharge high-voltage pulses should they reach here from elswhere (the tank circuit).

- RFI added for EHT and 12V+ power supplies

-12V stabilised power supply installed. Currently under-used to operate a relay, but this is in place ready to also power a PLL once constructed, and a fan. The fan will blow straight onto the two EHT transformers, which run hot (by manufacturers' design).

-Safety earths complete. I noted from past experience that no power suppliy grounds and no metal surfaces should be left disconnected from safety earth since HF, high voltage currents can be induced in practically any piece of metal that is not at earth potential.

- New secondary coil under construction. This comprises 500grammes of 1mm wire, for 210 turns on a 90mm OD Lexan former.

The Q is around 4 times higher than that of the previous coil I designed, close to 600. I am currently coating this with multiple coats of glass varnish, each heat-treated with a hair dryer. I will aim for a final thickness of around 1mm. Inductance is 1.44mH.

You can see in the last photo the jagged shapes I cut for the two meters (not as shown - this is just for testing). This was the best I could do with my available tools, woodworking knowledge and equipment, but these holes will be completely concealed by the grid and anode current meters!!

Thanks both to Dr Wobble (Andy) for offering to help make some legs and perhaps with the cabinet finish, and to Ed Dinning for offering with advice on the usage of Litz wire and also a supply of Litz wire.

The very high Q resulting from usage of Litz wire makes for very sharp tuning of the primary and I think this has discouraged its usage in coil-building circles, since tuning is already difficult enough. However, I am keen to experiment with this.

Cheers, folks, watch this space!

[Al (astral highway)]

Mods - correction, please! Para 4... It should read...

'After 2 seconds, the [series resistors] are shorted,' not the transformers!

I was typing under pressure to get the big post done within 30 minutes!

Please can we not have any comments about this mistake as it's clearly unintentional and will be OT. Thank you.

Thank you.

[Sean Williams]

Hi Al,

As a builder of high power valve based RF amplifiers, your box for the power side of things has me a little worried - there isn't a whole lot of space in there should something go wrong - multi KV supplies can and do flashover, all sat inside a nice flammable wooden box....

Appreciate a good Tesla coil though

Cheers
Sean

[Al (astral highway)]

Originally Posted by Sean Williams

Hi Al,

...your box for the power side of things has me a little worried -multi KV supplies can and do flashover

Hi Sean, wood isn't the ideal choice, I know, but I couldn't do this with a metal chassis - it's way beyond my metalworking skills. And I'm now far too committed for any kind of rethink, but a note to bear in mind for anyone else thinking of heading down this road.

I absolutely respect what you say about flashovers. They are indeed likely and common, but I have put a TG14 gas discharge tube in the EHT power supply for precisely this reason, and there will also be one between the other main flashover point, which is between the grid feedback and primary coil windings. Indeed, such flashover destroyed the grid feedback coil of a previous incarnation of this.

Also, its' not visible here but there is a ton of metal on the upper side - all of the top of the wooden chassis is covered with copper foil and some by 3mm copper plate, which is at least fire retardant.

More to the point, I will operate this only with a carbon-dioxide fire-extinguisher close at hand, and never unattended even for a minute. Ideally, I will design some kind of current sense in the anode circuit to crowbar the thing down in an instant. Working on it!

I do appreciate this wise note of caution. Thanks, Sean. Safety first...safety second!!

[Bufo Bill]

Could I just mention what a fabulous project this is and fabulously well recorded as a journal here. If there was an award for the best thread on the site, my vote would lie here; well written, concise explanations and diagrams and photos too. We could all learn from this for our own topics. Well done Al!
Regards from Bill.

[Al (astral highway)]

Originally Posted by Bufo Bill

Could I just mention what a fabulous project this is and fabulously well recorded as a journal here. If there was an award for the best thread on the site, my vote would lie here; well written, concise explanations and diagrams and photos too.

Hello Bill! Wow, thanks for taking the time to say that. I'm happy that you're enjoying the project and my write-up so much. Very generous of you to be so detailed in your feedback, I'm moved.

I do have much more time, energy and capacity to think about the project at the moment than I might normally - long story and OT - and so it's become more important to me than I had first realised.

Welcome, we haven't spoken before.

Cheers!

[feathercurls]

Hi Al, awesome project, thanks for sharing the progress. As another high voltage project enthusiast, I would love to see the finished result. (I haven't done any high-power projects myself.) Good luck and stay safe.

[Red to black]

Hi Al,
I whole heartedly agree with the last two posters, I too am following this with great interest, I love stuff like this, I only never commented before now because you are presenting it so well and I had nothing really to add to your discussion.
Keep up the good work.

[Al (astral highway)]

Thank you Lucia and thank you Feathercurls!

I appreciate your feedback and support, it's wonderful that you took the time to post your comments, and I'm happy that you're enjoying this thread!

Cheers