Induction Annealer Component and another question

[itchyTF]

For the analog design types -

I built a Royer oscillator based induction annealer 6 or so years ago (kit from Fluxeon) but now I'm building a Gina/Eric version for a friend. People have commented that the caps get hot. The caps on my Chyna board are marked 50KHz. Is this the reason? I tried to find replacement caps that would be more suited for the 100KHz range but was unable to find anything. Any thoughts?

I plan to use the same output connection method I used on my original annealer. I added brass pads on top of the 3 standoffs to connect heavy gauge wire to the brass "connectors" on the front panel. The tubing slides through, as shown, and is clamped down on the front side. The work coil is soldered to these tubes. My concern now is skin effect. At 100KHz the depth of the conduction path is pretty thin so the center core of the stranded wire may not be carrying much (or any?). Should I use multiple lesser gauge wires or use copper sheet to make the connection? I don't really want to get into Litz wire if I can avoid it. Don't have a good way to strip the insulation off the individual strands (no solder pot).

 

[Gina1]

Only real advice I can give is build a basic GinaEric induction annealer as presented in the original thread.
It works ! The one I built is still going strong now, more than 5 years, with no problems or failures. The fan over the induction PCB keeps the capacitors cool. (simple low cost solution).
Just my two cents
Gina

 

[McFred]

Yes, several small gage wires in parallel would be better than a plain heavy gage wire or blocks of metal of similar cross section. There are some design features you can copy from the pictures in these threads:

forums.accuratereloading.com/eve/forums/a/tpc/f/9111080861/m/7681074262

forums.accuratereloading.com/eve/forums/a/tpc/f/2511043/m/5931089822

Basically the tank circuit has a LOT of current circulating in it so things will get hot no matter what you do, so it is generally easier to cool them than try to eliminate heat. In some of those pictures that use fewer pre-fab components you can see that the capacitors have very short leads that are soldered straight to the Cu tubing of the tank and are also lined up directly under a fan:

 

[McFred]

How is your project going? Was my information helpful? You're using water cooling, right?

 

[itchyTF]

I had already purchased the induction board, so I'll keep my fingers crossed. I am using liquid cooling as well as fans for the radiator (which I may not need) and the induction board.

I'll be using a 36v 1000w supply. How much is a LOT of current in the tank circuit?

I've seen charts that rate 20 gauge stranded wire between 7-11 Amps. I don't know how to debate due rto frequency. I guessed 3 amps was reasonable per wire so I'm trying 9 strands. Hope they don't go up in smoke. When I get it together I'll observe with no brass for a short time then gradually increase if they don't get too warm.

 

[McFred]

The tank circuit has roughly 150 amps of high frequency AC circulating. I would suggest cooling as much of the tank circuit with water as possible. Those machined brass blocks look nice, can you route cooling water through them? Big blocks of metal work well with DC current, like bus bars on battery and capacitor banks, but with high frequency you will have eddy currents circulating in them. Also, are those screws steel?

You could take the board apart if you want, and mount the components better. How far down the rabbit hole do you want to go?

Again, efficient electrical design is not very critical if you can keep it cool, and water cooling is hard to beat.