Thread: DIY Induction heater

Hi RC, Guys,

Look what I happened across whilst looking for a bearing !

21-12-2022-001.jpg

I knew that I had the spare part of the induction heater coil kicking around somewhere. If you look at the left hand side you can see where I had bent the tube over so that it hooked onto a screw in the wooden former that it was wound on. Also the lathe was running in reverse when winding so that the tube was on top and could be guided easily.

Well, I finally got some new mosfets for the replacement mosfets I blew up. Turns out I put the wrong zener diodes in when I replaced them.

Anyway since where the zener diodes sit I have completely lifted the traces from removing them, there has been some bodgy workarounds done. And upon carefully firing it up with it fused so not to blow anything, it works.

Now my one hung low 1500w 36V power supply, well turns out it does not like anything more then about 750W, then it cuts out.

I do wonder if a cheap 140amp inverter welder would power it. It puts out about 30V under load. Time to try it.

And the answer is?

20230112_181009.jpg

That is a bit of a pain but if you at first don't succeed........
Richard these vids might give you some clues.
https://www.youtube.com/watch?v=i9Ei1jVII5w
https://www.youtube.com/watch?v=SZEJJjYy7N4
https://www.youtube.com/watch?v=I_V9H_-5H-8
This one has subtitles, a bit of a quick flick through it looks like he gives a pretty good idea of what to do and he has more vids on the right https://www.youtube.com/watch?v=sicT...&start_radio=1
and this is thought provoking https://www.youtube.com/watch?v=AcbbhayfynE

Thanks for the links to the vids. Trouble is all these inverter welders are wired differently and I do not have the knowledge on what to do.

In any case I think my bought induction heater has some other problems. Connected to a 12V battery it works still.

But connected to my 36V PSU it does not work as the PSU will not power up (too much inrush current required I think) and it was arcing across the mosfet heatsinks when it sometimes did fire up.

So I am getting set up to build the Schematix induction heater and have all the components now. Got delayed as I ordered a TS101 soldering iron back a month and a half ago, and I do not think they even bothered to send it as the tracking never happened. I got a refund. So have ordered a cheaper 24V powered T-12 clone, which at least the tracking shows something has been sent.



When I had the Chinese built heater going, I put the cheap pocket oscilloscope on it. The voltage shown is with the probe on 10X, so would be 10 times higher.

20230113_190701.jpg

Originally Posted by .RC.

Thanks for the links to the vids. Trouble is all these inverter welders are wired differently and I do not have the knowledge on what to do.

In any case I think my bought induction heater has some other problems. Connected to a 12V battery it works still.

But connected to my 36V PSU it does not work as the PSU will not power up (too much inrush current required I think) and it was arcing across the mosfet heatsinks when it sometimes did fire up.

So I am getting set up to build the Schematix induction heater and have all the components now. Got delayed as I ordered a TS101 soldering iron back a month and a half ago, and I do not think they even bothered to send it as the tracking never happened. I got a refund. So have ordered a cheaper 24V powered T-12 clone, which at least the tracking shows something has been sent.



When I had the Chinese built heater going, I put the cheap pocket oscilloscope on it. The voltage shown is with the probe on 10X, so would be 10 times higher.

20230113_190701.jpg

Hi RC,

The scope picture is about what I would expect !

Well some time ago I got the PCB of the schematix 1200W induction heater built. But I only got to test it out today with a coil that was not optimised for this heater.

It certainly sucks some power and can turn steel bars red ina short time.

I was powering it with a 36V PSU and it outputs 41 amps, and if I stuck the bar too far into the coil it would draw more then that and cut out the PSU.

I have some spare boards here as I had to buy 5 from JLCPCB, say $5 each plus postage is anyone is interested. I have three spares as I will keep a spare for myself.

The components I bought from mouser and were about $100, although you could cut that back a fair bit if you do not use $20 mosfets as I got higher rated ones as the listed ones were all out of stock everywhere. .

Originally Posted by .RC.

It certainly sucks some power and can turn steel bars red ina short time.

C'mon Richard. You know the deal. Pictures or it didn't happen.

Originally Posted by GuzziJohn

C'mon Richard. You know the deal. Pictures or it didn't happen.

Well it took me awhile. But I finally got the proper coil made. It takes nearly 3 metres of 10mm copper pipe. I bought two metres thinking that would be enough. Had to join on some more, and 3/8" copper pipe is not easy to come by. Apparently plumping pipe starts at 1/2".

Anyway, got the coil sort of bent into a shape resembling a coil. Fired it up and it can certainly suck some power. Put something too big in there and it goes over 40 amps at 36V and the PSU cuts out. The inductor coils also get over 100C pretty easily, so forced air cooling is going to be a necessity.

In any case it does certainly work, and this one did not blow up components like the ready built chinese one did. yet.

20230507_171356.jpg

20230507_170924.jpg

20230507_171959.jpg

Hi Richard,

If you formed the coil ends out at 90 degrees, you could pump cold water through the coil and keep it cool ! You could also have gotten hot water to wash your hands as well.

Very nice!

I know that it’s common to run cooling water inside the coil. Is there a rough power threshold where that becomes necessary?

Edit: John beat me to the water cooling while I was watching YT vids of it

Steve

Hi Steve,

The coil gets very hot regardless ! The problem is any soldered joints on the coil ends will get hot just from the conducted heat. I've watched soldered on coils drop off at one end because of it. Then if you are lucky nothing else melts or drops off.

This coil does not get that hot for the brief uses I have used it for, which is not much at this stage as I am still experimenting. I cut my coil down a couple of turns as per the original design. I think you would get a coil out of 2 metres of copper pipe now.

I have had a problem though and after I had modified the coil, the trace to the coil blew right where the coil bolts on and where it was not reinforced with copper wire/solder,

Then it started making a high pitch squeal as I quickly cut the power. I repaired it (photo's later on) with soldering some flat copper I cut out of copper pipe.

I dreaded the worse as these things seem to blow mosfets as easy at looking at them. However it still seemed to work, which was thank goodness as they are bloody expensive mosfets in this one, but I still had a bad problem of heating issues where the trace blew and higher current draw at idle and the coil heating up with nothing in it.


I did wonder if it was a resistance issue where the coil bolts onto the PCB. So I did some fiddling with the connection and just like that it is back to normal.

Put it on the oscilloscope running at a low voltage and a nice sine wave.

SDS00001.jpg

Here is a scope reading of the mosfets going on and off. Yellow and Blue are the drains connected to the coil. Pink and green are the gate pins that turn the mosfet on and off.

SDS00006.jpeg

Hi Richard,

Pink and green are the gate charging voltages. Fet devices have quite large gate input capacitances.

The little hump is probably caused by the zener voltage turn on time.