Finding the appropriate connectors for an induction heater work coil

Thread Starter

Marcus2012

Joined Feb 22, 2015
425
Hey guys

Hope everyone has had a happy holidays :)

I was just hoping someone could help me with my induction heater assembly. I have one of those noobheaters and the unit works fine. However I attempted to have the output from the unit going to binding posts on the case with a cooled work coil and two metres of cable connected to the binding posts. Unfortunately I did not foresee that the binding posts (and cable) would heat up at a greater rate then the actual work piece. The result was that they melted and the cable got very hot :( So I was hoping someone may have some suggestions for appropriate materials for connectors? Or is this just a consequence of having the work coil that distance from the driver?

Thanks for any and all help as usual guys

xD

royer_ih.png
 

GopherT

Joined Nov 23, 2012
8,009
@Marcus2012

It is a question of the diameter of the wire you use to get the current from transformer to work piece. At these amperages, you are best off with large crimp connectors.

Also, google wire gauge, then calculate the power absorbed by the cable by multiplying amperage by resistance of wire to get the voltage drop. Then multiply by amperage again to get power drop. It can be quite amazing if you don't account for the cable resistance.
 

Thread Starter

Marcus2012

Joined Feb 22, 2015
425
@Marcus2012

It is a question of the diameter of the wire you use to get the current from transformer to work piece. At these amperages, you are best off with large crimp connectors.

Also, google wire gauge, then calculate the power absorbed by the cable by multiplying amperage by resistance of wire to get the voltage drop. Then multiply by amperage again to get power drop. It can be quite amazing if you don't account for the cable resistance.

Thanks the quick reply. I was using 17AWG with about 6.5A load. Strange thing is the cable from the driver to the bind post didn't heat up. The binding post melted and the two metre stretch of 17AWG got hot but not as hot as the posts. Anyway I did look it up and 17AWG has a resistance of 16.61 mOhm/m so cable dissipation is only 1.4W.

0.01661Ohm*2 = 0.03322

0.03322*(6.5^2) = 1.40W

I do not know the resistance of the bind post though or what it is made of for that matter so I'm thinking this might be the major source of resistance.
 

neonjohn

Joined Mar 3, 2013
7
Hi Marcus

That is a horrible circuit, one I posted about 10 years ago on my website when I was first learning about induction heaters. I'll be happy to explain the problems if you like but for now let me just direct you to a much better one and then answer your question.

http://www.neon-john.com/Induction/Roy/Roy.htm

Do you have one of the ready-made heaters that are widely available on sleazebay and elsewhere? Can I also determine from you post that you've made a transformer for the output? Assuming you have a transformer, the current is going to be quite high (relatively speaking, at least). You'll need to use heavy gauge wire, preferably about #4 welding cable.

What you really need is Litz wire. This is finely stranded wire with each strand individually insulated. Litz wire overcomes the main problem with conducting high frequency power: skin effect. The skin effect causes the electricity to flow only on the surface of the wire. With Litz wire the current flows on the skin of every one of the tiny strands. You can buy Litz wire in small quantities from my former company, Fluxeon.

Welding cable is about 60 to 70% as effective as Litz because the individual strands oxidize and partially insulate themselves from their neighbors. Fluxeon uses welding cable because we had yet to find a company that would extrude a neoprene jacket over Litz wire. I am actively pursuing that with my new company Tnduction.com (no website yet).
 

neonjohn

Joined Mar 3, 2013
7
The rest of my post. A storm interrupted my Hughsnet connection.

...
Hi Marcus

That is a horrible circuit, one I posted about 10 years ago on my website when I was first learning about induction heaters. I'll be happy to explain the problems if you like but for now let me just direct you to a much better one and then answer your question.

http://www.neon-john.com/Induction/Roy/Roy.htm

Do you have one of the ready-made heaters that are widely available on sleazebay and elsewhere? Can I also determine from you post that you've made a transformer for the output? Assuming you have a transformer, the current is going to be quite high (relatively speaking, at least). You'll need to use heavy gauge wire, preferably about #4 welding cable.

What you really need is Litz wire. This is finely stranded wire with each strand individually insulated. Litz wire overcomes the main problem with conducting high frequency power: skin effect. The skin effect causes the electricity to flow only on the surface of the wire. With Litz wire the current flows on the skin of every one of the tiny strands. You can buy Litz wire in small quantities from my former company, Fluxeon.

Welding cable is about 60 to 70% as effective as Litz because the individual strands oxidize and partially insulate themselves from their neighbors. Fluxeon uses welding cable because we had yet to find a company that would extrude a neoprene jacket over Litz wire. I am actively pursuing that with my new company Tnduction.com (no website yet). A rule of thumb for computing skin depth is dept(mm) = 400/sqrt(freq) in hz. Even with #4 gauge wire, our 1500 watt heater causes the cable to get very hot.

You can approximate Litz wire with as many strands of magnet wire as you can muster or even wire wrap wire. String out about twice the length you need and then twist it very tightly using an electric drill. It will draw up to about half its length. While not as good as the hundreds to thousands of strands in real Litz wire, this is a vast improvement over hookup or electrical wiring wire.

If you DON'T have a transformer then extending the coil is going to be problematic. The distributed reactance will affect the resonance of the tank and the added resistance will dramatically lower its Q. About the best you can do is tightly twist the wire/cables. This will cause the wire to act as a transmission line. The tighter the twist the lower the impedance.

I tried that technique in the beginning with just a foot of wire and had problems. I SPICED the circuit, representing the cable as a ladder network of L, C and R, and quickly saw the problem. That and life safety concerns (our heater runs directly off line current) led me to use a transformer.

Any other questions, feel free to write. I don't do forums very much but you can contact me at jgd@tnduction.com.

John
Chief Engineer
Tnduction Corp
 

Thread Starter

Marcus2012

Joined Feb 22, 2015
425
@ neonjohn

Thanks for the reply it was indeed on of the ebay version I bought. I was a little wary about adding 2 metres of cable as in the description the coil is just mounted on top of the cap bank. So thanks for clearing all that up for me and thanks for the original links to your work, much appreciated :) The binding posts just melted long before the work piece heated up, so I guess at that frequency they just didn't have the conductive cross-sectional area for the current. I've had a look into this Litz wire and I think I can get that no problem but I am a little confused as why I'd need a transformer between the cap bank and the work coil. Could you explain that please? Sorry if it's a major concept I've overlooked there :).
 
Top