DIY High voltage Jacob's Ladder Build

[Ant]

It seems to be the left hand rod moving in and out slightly in relation to the other one. Wasn't sure if the increased distance (albeit slightly) between the two rods would require a slightly larger voltage to go the same height. 

But I suspect your "heat" suggestion is nearer the mark. 

Good stuff!

Ant  

[angryowl]

Couple more updates, the holder is finished and has had its first layer of wire wound and I think it turned out pretty nicely and compact. The ends were glued with regular two part epoxy. 101 turns on the first layer. Three layers of regular plumbing PTFE tape between each layer and I'm onto my second layer now, though I've ran out of wire so more wire from the strands now needs to be unpaired and straightened. So will be a long day inside today for me...

[angryowl]

What do the more experienced DIYers out there with knowledge in such things think about my approach? Will this ever work or should I just stop as to finish it I've a LOT of wire to process...

[barkis]

The Arcs look extremely bright, almost like an Arc Welder, so any eye protection 
required? What Current  is flowing in your device? 
I know little or nothing about them
Just thought the Eye danger was worth a mention in case it is relevant here .

[angryowl]

50 minutes ago, barkis said:

The Arcs look extremely bright, almost like an Arc Welder, so any eye protection 
required? What Current  is flowing in your device? 
I know little or nothing about them
Just thought the Eye danger was worth a mention in case it is relevant here .

Yep, definitely worth mentioning this. My one in the video did indeed have a few miliamps of current (no means of measuring) making the arc very hot and bright, so you are right in suggesting adequate eye protection should be worn, especially when looking at ones that put out a lot more current thus brighter and hotter arcs.

Not only that, but simply the UV emission from hot arcs I would imagine giving long exposure could cause some serious skin damage.

Thanks for pointing this out!

[angryowl]

Just a quick update on this. The entire bunch of Cat5 wires in my previous image has only produced 5 and about three thirds layers on the core, this is including PTFE insulation in between each layer.

By my (poor) calculations, I should be able to wind approximately 21 layers and still have room for the ZVS wire. On average there's 100 turns per layer so that would give me 2100 turns in total. #

I know there's a calculator executable somewhere on the inter webs which allows you to calculate input and output power, number of turns, voltage amperage the whole deal and I tried it. The only problem with this is that the ZVS output frequency is between 30KHz - 60KHz whereas the calculator only deals with I think 50Hz.

I'm mostly concerned about the secondary getting hot as I stupidly used cheap two part epoxy which I know it gets malleable when warmed up. also one of my major concerns is corona between the secondary windings, I'm hoping that the main thing stopping that from happening will be the nice coating on the copper wires. According to this there are various types of insulation used on Cat5 UTP cables and I'll try and see if there's a way of finding out from the cables what type of insulation the individual copper wires use. That way I have a chance of finding the dielectric strength of the material and thus either put put mind ant ease or I'll be biting my nails.

Anyway I'll certainly try and squeeze in as many layers as I can on that core and see how things go with the ZVS. The way I see it, two outcomes; nice high voltage and current arc or smoking and popping everywhere! Anyway that remains to be seen. Will keep updating as I go along.

[angryowl]

Finished winding the first ten layers giving me about 952 turns. It's gaining some weight now with all that copper in there! I'm contemplating giving it a quick test with the ZVS now at low power, something like 12V rather than the 30V transformer as I have one... Hmmm

[angryowl]

Almost 12 layers in giving a total of 1135 turns. Ran out of Teflon tape so will have to get some more but in the meantime I'll setup a test for tomorrow with just 12V going into the ZVZ rather than 30V and with a low number of primary coils to see what sparks I get. Mostly I'm concerned about the secondary heating up as it's drawing a lot of miliamps. See how it goes and can even give the 30V a try see what what output I get.

I'll update tomorrow with how things go. 

As a note: SAFETY FIRST and I'LL BE TAKING EVERY PRECAUTIONS WHEN WORKING WITH THESE HIGH VOLTAGES. I HAVE EXPERIENCE WITH HIGH VOLTAGE EXPERIMENTS SO I KNOW WHAT I'M DOING!

[angryowl]

Did a quick test this morning with 12V and 30V with 5x5 turns on the ZVS side, also tried 3x3 turns. Good news is it works! Bad news is that at 30V it generates plenty of corona inside the layers, enough to be seen and heard. This is mainly at the bottom and top of the windings and of course when the secondaries are unloaded.

30V 5x5 turns ZVZ side generates an impressive amount of current, but not that much voltage I think. I tried my usual setup with the two rods seen in the first video posted and could not get it to spark across, it needed help. 

This is 12V arcing 

20180430_122804.mp4

 

This is 30V arcing

20180430_123032.mp4

 

And this is the corona

20180430_123138.mp4

[angryowl]

A quick update on this, I'll finish winding the remainder of the Cat5 wire I have which hopefully should get me close to the 2000 turns mark.

I'm not sure if the corona at the bottom and top of the coil is caused due to poor insulation between the layers themselves (2-3 turns Teflton tape, plus thick insulation of the Cat5 wire) or whta's actually happening is the bottom of the layers are forming corona to the ferrite core. I tend towards the latter as I've no idea of the dielectric strength of the layers of plastic I've found and taped together to form the core.

I know that roughly doubling the voltage by increasing the number of turns will definitely increase the amount of corona generated and there's a high chance that it'll arc to the core if ran on 30V, which I intend to do.

But I think I found a solution that'll prevent the corona altogether even at 30V, transformer oil. Or mineral oil will do just fine, I was studying the dielectric properties of regular transformer oil Shell Diala B and some other alternatives and they seem well suited for higher voltage applications than mine so I think submersing this whole core together with the coil in mineral oil should do the trick, plus it's a great coolant!

Of course if I just pour oil into a container containing the transformer, even with manual shaking and vibrating there will still be plenty of air bubbles/pockets. For that reason I shall find a saleable container large enough to fit the whole contraption in and luckily enough I found through my rubble of scrap a perfectly working refrigerator compressor that's been properly discarded of it's Freon. With that being said, what I plan on doing is running the transformer mineral container through a vacuum for a couple of days to ensure all or most of the air is out. 

I fear this project went from a simple quick DIY Jacob's Ladder to a whole new level, but I'll roll with it and see where it leads me 

That's the plan at least anyway, will update as I go along

[angryowl]

Hi all, back after a rather long break and after some setbacks I'm determined to finish this project.

Right after my last post I pushed the coil to its breaking point by adding about 10000 uF caps in parallel with the 30V input. Although the arcs were unlike anything I've seen from it, it shortly arced across two turns and upon disassembly it turns out one of the last turns found its way down close to the first few layers. Due to the very high voltage differential between the two turns, the PVC insulation on the wires broke down.

After researching what materials and techniques the professionals use for such high voltage applications I decided to go for a split bobbin as this greatly lowers the differential voltage between adjacent layers. I've gone with 5mm thick acrylic sheet for the top and bottom of the coil holder and also 5mm between the ferrite core and first layer of turns. For the separators I used 4mm thick acrylic which seems to be doing its job very well. The acrylic was welded using acetone based paint remover which resulted in incredibly strong bonds. Reason I went with acrylic was of it very high dielectric strength of 17 - 30 KV/mm and that it's readily available to me.

After each layer was wound I applied two generous coats of lacquer to both help keep the turns in place and fill in the air gaps to further avoid arcing. After applying the lacquer I used a turn of 0.06mm Kapton tape to hold everything together nicely and as additional dielectric protection between adjacent layers. 

The total number of turns came to 1645 turns with an average of 12 turns per layer. 

First test was a complete success at 30V without the capacitors, no sings of corona present with no load across the high voltage output, and a very nice, long and thick arc that travels all the way to the top of the ladder and seems to start by itself at the bottom every time.

I'll post a video shortly showing the arcs but I'm very pleased with the result and I can say all the time I put in the new coil has been worth it.

Now I just need to figure out what I'll be using for a box and how I'll mount everything.

[Alien 13]

Nice job, I have to ask about where you got the lacquer from? I have looked around and found it difficult to source,  I need some for my Tesla coils.

Alan

[angryowl]

38 minutes ago, Alien 13 said:

Nice job, I have to ask about where you got the lacquer from? I have looked around and found it difficult to source,  I need some for my Tesla coils.

Alan

Since I didn't need very much of it, I just used a can of automotive acrylic based clear lacquer I had laying around. I used a fine brush to apply to each layer after spraying some lacquer in a small can.

I looked online for general acrylic based lacquers and they all seemed to have very high dielectric strength and although on my generic can this wasn't specified I think it should be close enough to the ones I saw online.

Doing a quick search got me this, there's several affordable ones on there with rather high dielectric strength. Of course as per the usual shipping would probably come in more than the product, but I reckon you should be able to find other sources.

[angryowl]

Video without any smoothing capacitors across the input of the ZVS driver.

I might start out with testing a 2200uF capacitor and if it helps smooth out the ripple I could keep adding until I reach a smoother/less noisy arc as currently it's a bit louder than I'd like.

[Alien 13]

Thats impressive, I actually like the sound its a bit like those Hammer Frankenstein movies.

Alan

[angryowl]

Hehe, yeah I suppose it does.

However my phone's microphone doesn't to it justice as it's much louder in person.

Would be nice if I had a way of measuring the voltage and current. Tried estimating voltage with software but haven't had much luck, but I reckon it's putting out a decent amount of current judging by the thickness of the arcs.

[angryowl]

The enclosure is now completed and currently testing different length and spacing combinations of the electrodes to find one that produces the effects I'm after.
I think shorter electrodes and a larger distance between them at the top produces a more pleasing arc than my previous longer electrodes.

I also added a single 2200uF smoothing capacitor on the input of the ZVS driver which eliminates most of the crackling noise present before and increases the arc's thickness a bit.

For cooling I had planned to use a 70mm medium speed fan, but that turned out to be inefficient at properly cooling the ZVS MOSFETs and transformer which are the only components getting hot. This was replaced with a ..mm high power server fan with an added 8 ohm 10 watt resistor to lower the speed a bit. This generates a fair bit of pressure inside the enclosure pretty much being able to cool all parts.
I had to use a separate mains connection for the fan's 12V circuit as when connected to the same transformer mains input the fan would simply not start.

Just need to decide on a glass dome size and figure out how to mount and seal it on top of the enclosure.

A most interesting reaction takes place when an arc such as this is operated within a sealed environment.

Due to the high temperature of the arc some of the oxygen and nitrogen in the air combine to form nitric oxide, a colourless gas, which upon cooling combines with the leftover air to form nitrogen dioxide which has a brownish colour. This explains why after roughly two minutes of operation the colour inside the beaker changes to brown.
Another interesting fact is that this nitrogen dioxide gas is short lived as when exposed to wavelengths shorter than 400nm (deep ultroviolet), photolysis occurs to form nitric oxide and atomic oxygen. Therefore a few minutes after the arc is stopped and the brownish gas is exposed to natural lighting, it looses all of its colour, but the characteristinc odour of nitrous oxide is still present.
I also believe a small amount of ozone forms as its smell can be easily detected.

The above reactions were used in the Birkeland–Eyde process to produce nitric acid by dissolving the nitrogen dioxide in water. However the process was highly inefficient requiring large amounts of energy for relatively small yields and was replaced by other more efficient processes.

Although nitric oxide is generally considered non toxic, nitrogen dioxide is quite toxic and considered harmfull if inhaled in large quantities as it can form nitric acid with the water in the lungs which can lead to pulmonary edema and eventually death.

That being said, even though the glass dome I plan on using will be completely sealed, I'd still like to avoid generating highly toxic gases if I can. For this reason I have been looking into operating the ladder in an inert gas such as argon which is readily available. However this still needs some further research. 

The other aspect I need to look at is the material of the electrodes as when operated in air copper corrodes incredibly fast. The small diameter of the copper wire I used for testing might accelerate this further but if I go with copper the final electrodes will be much thicker. I am looking at other materials, but I'm mainly interested if there will be any corrosion with any metal if operated in an argon environment. Again more research needed...

Link to video of new test

 

[angryowl]

An interesting side effect of the ZVS operation is the induction heating of the steel plate and two long screws which hold the entire ferrite core pressed against the back of the enclosure. 

I should have used aluminium for the plate, but I found that steel plate with the right size and right holes so didn't bother. I don't plan on running the ladder for longer than 15 mins at a time so and it doesn't heat up to such an extent that I would consider replacing it.