Help wiring a flyback transformer.

[Atomicrox]

Love boardless circuits, mine's like that as well

Sounds simple except for the "Iron-cored 21AWG home-made inductor", I have no idea how to make one.

Update: Rewound my primary after applying electrical tape to the core, haven't seen it arcing any more. One bottom pin is still arcing under the glue, unfortunately. There's also some corona (looks like it's under the glue near the HV- but I'm not sure) - is that harmful to the flyback?

Sigurthr, I found a calculator and according to it adding another 50k pot in series with mine should allow the frequency to go as low as 6kHz (which should be well below the point of resonance). It also messes a bit with the duty cycle.. what's the optimal duty cycle for this circuit?

 

[Sigurthr]

Optimal duty cycle is 50% AFAIK, but it isn't critical, just keep it between 33 and 66 %. Give it a shot if you have the resistor and report back any findings =). As I said self capacitance may be too low to achieve resonance, but you never know.

Glad the insulation worked. Arcing under glue can only be fixed by re-liquifying the glue and allowing it to reharden slowly. Most hot glue is an exceptional insulator but if the glue is not hot enough (low, or med temp variety/gun) it will not bond at a microscopic level, which leaves gaps and voids for the arcs to pass through. Submerging the transformer in oil will only help if the oil can get in to these spaces, so you may have to bake the transformer at a low temperature (whatever the melting point of the glue is) to get it to reflow. This does two things: 1) possibly fills in microscopic gaps solving the problem, and 2) is the only way to easily remove large amounts of glue. So, test it after baking and if it is still arcing pop it in an oil bath.

Corona won't affect the hot glue but it will deteriorate everything else... metal, plastic, you name it. Eventually corona weakens the dielectric (insulative) properties of the material and allows arcs to form. So wherever there is corona now, will become a problem spot later. Corona likes to form on sharp edges, so try to avoid any. Hot glue can block out corona but it must be HOT. As stated above, low/medium temp is not hot enough. I don't remember the temp of the glue I use, but I think it was around 245C - well hot enough to take your skin off. It seals exceptionally well against HV - I've tested it to more than 100kV/cm^2. If I use the same glue in a low temp gun it fails at less than 30kV.

Oh and an iron cored 21awg inductor is just 21 gauge wire wound on an iron core. Core type matters, laminated iron, or solid iron is not the same as powdered iron or ferrite. So if you do make the inductor make sure you get the right core material. I just salvaged a core from a stepdown power transformer today, took about an hour of sawing with a hacksaw to get the windings out. It's a laminated iron E/I core - GREAT for home made transformers, but not great for GDT or inductor use. Would be better if the E and I parts came apart but they don't, no biggie. I'd only be using it for step-down/up, isolation, or inductive coupling uses, perhaps a small inverter. Anyway, my point is cores can be found free or nearly free if you know where to look, and with some effort are reusable.

 

[Atomicrox]

Just to make sure before I do it wrong - by bake you mean inside a kitchen oven?

 

[Hiemal]

Atomicrox, would you like me to build you a ZVS driver? I could do it for $35, if you'd like.

 

[Atomicrox]

Thanks for the offer but I like to build the stuff, it's cooler when you get it working
I hope it'll also help improve my poor electronic skills.

 

[Sigurthr]

Ya, in the oven, and just the flyback only, not the driver circuit. Only set the oven to the glue's lowest melting temperature, which should be less than 250f.

Btw, Jared (Le Quack) makes a MEAN ZVS driver.

 

[Atomicrox]

Lowest temperature on my oven is 150C (356F). A couple minutes at most?

A mean one will most likely fry my poor flyback, lol

 

[Sigurthr]

Man, my oven goes down to 150F, haha. Experiment with short times and keeping watch, worst that can happen is you need a new flyback, and that'l happen soon with arcing anyway.

Don't put it on a metal pan, btw, as the metal will get too hot, put it on ceramic or glass. If you have a paint removal gun you could use that as well instead of the oven.

 

[Atomicrox]

Update!

I used my aunt's hair dryer to remelt the hot glue and it seems to have fixed the stray arcing

I also increased the power and added a fan to the MOSFET cooler. I think I can reduce the number of windings a bit more before it starts overheating.


Here's an improved Jacob's ladder:

It also has quite an interesting effect on water, note how it forms a "hole" when the wire isn't close enough to arc (don't worry, I was using a well insulated plier to hold it):

 

[Sigurthr]

That "hole" is the ionic wind applying pressure to the surface of the water, similar to if you were to take a straw and gently blow at the water's surface with the straw out of the water.

Grats on the increased performance!

 

[Atomicrox]

Thanks, I'll update the video with the explanation

Do you happen to know why the negative wire seems to take a lot more heat than the positive? I figured it should be the other way around, since electrons are coming out of it and hitting the positive.

 

[Sigurthr]

In welding a similar phenomena occurs:

in TIG welding where the electrode does not melt/get consumed by the weld and is shielded in a stream of fresh non-ionized gasl; a negative electrode polarity deposits most of the heat in the positively charged work piece.

in SMAW "Stick" welding where the electrode is coated with a reducing chemical mixture that oxidizes leaving an oxygen-free plume of fumes surrounding the electrode, and the electrode is consumed by the weld (by being transfered across the arc as a stream of droplets) having the electrode positively charged deposits most of the heat in the negatively charged work piece.

Now, back in TIG welding we encounter another interesting and possibly related phenomena; when having the electrode positively charged and the work piece negatively charged the work piece's surface exibits an etching force which blasts away surface material. This is shown to not be from electron emission but rather the striking of the surface by positively charged ions from the shielding gas cloud.

It is thus possible that the uneven heating effect we observe from DC arcs is from ionic activity rather than electron emission.

 

[Atomicrox]

Might be that, ions should be a lot heavier than electrons and might carry more kinetic energy.