Final Video-High Power Current Controlled Induction Heater

This is the final video in my series of videos on induction heater construction. In this particular build, we implement current control to prevent damage to components such as capacitors and transistors. This has been tested over several runs and seems to work really well for a much more robust build. For construction details , please refer to the following sites:
teslascience.w...
teslascience.w...
This particular build can use any type of induction heating coil and works on a feedback based mechanism similar to that employed in solid state, Tesla coils. This particular unit has been tested up to almost 10 kW. However, on regular runs, it is usually operated between 2and 5 kW. Levitation of metals in levitation type set ups requires much higher power levels. A lot of this is in the form of wasted energy. However, heating large workloads is much more efficient.
********** Just found out that I can do away with the current shunt and get way more effective feedback from the coupling transformer. In addition, the current in the coupling transformer ia minimum 16 to 20 times lower than that in the primary tank. Check this channel to see how this is done********
The video of the coupling transformer feedback is shown here:
• Final Video-High Power...
In summary, I hope this build comes in useful for your future induction project.  Care has to be taken during this construction, particularly if there are exposed live wire connections. Again, thank you for looking at this video. Your comments are most welcome. 

[@TeslaExplorer]

Your craftsmanship is over the top.

[@morganchandler5126]

Awesome work Mr Brian 💯.... Hooe you have been well

[@BrianDhvTinkerer]

Thank you. Wish I could do more of this type of stuff.

[@A.D1852hh]

Hi brain. I am following your induction heating project with excitement. You solve every problem in a smart way, this is great! Instead of current limiting, there are capacitors specially produced for this purpose, 5000A value and liquid cooled. I wonder if current limiting would be necessary if this capacitor was used?

[@BrianDhvTinkerer]

Thank you! That is very much appreciated. Yes I saw those capacitors and I’m tempted to just use those instead of messing around with current limitation.- dawn cap makes them!

[@jamessmith..919]

This is a great video, i love how passionate you are🙂

[@BrianDhvTinkerer]

Thank you. Really enjoy hobbyist electronics!

[@Magneticitist]

Awesome Brian. I was scrapping some stuff at work earlier thinking about this.. aside from probably running through like 100lbs of copper bus bars I realized I have access to some of the largest heatsinks I've ever seen. Makes me wonder if I could pull something off like this with a bunch of high power fans and a 30lb heatsink haha.

[@BrianDhvTinkerer]

Haha you could!

[@BrianDhvTinkerer]

I got this thing finalized. Took a bit of extra work, but found a great way of doing the feedback. Here it is: teslascience.wordpress.com/simple-and-effective-induction-heater-current-control-circuit/

[@Magneticitist]

@@BrianDhvTinkerer Cool! Thanks for the write up, I'm just waiting on some copper tube to give it a go. I got some 3/8 OD at first but found it too difficult to work with without kinking.

[@BrianDhvTinkerer]

If you get the soft copper from the hardware store, straighten it out, tape off on end and fill the tube with fine salt or with fine play sand, tapping the upright tube on the ground to tamp the sand down and get tight packing, then you can bend it around a piece of PVC or a bottle former to make your coil, then it works well without kinking. That’s basically how I did mine. It’s definitely challenging if you make a levitation coil 😅

[@Magneticitist]

@@BrianDhvTinkerer That's a great idea, I'd have never thought of that. Might have to give the 3/8 another go. I definitely had no idea how I was actually going to go about making fairly small diameter turns haha.

[@T3sl4]

Nice demo! Glad to see series resonant drive here, and the levitation coil is a fun pick. I'm a fan of the PLL style control, myself: frequency is varied to set power output, and all the process variables (output voltage, inverter current, phase shift -- to prevent crossing resonance, power) just act to adjust that upward. With fairly slow compensation (time constant of 10s ms?), this is well behaved and makes for a very reliable, robust and flexible (controllable) system. A few other protective features (input inrush / precharge? desat detection?) and you basically have a full industrial unit!

[@BrianDhvTinkerer]

Thanks - it has held up well. Got inrush via a bunch of thermistors. With over current all feedback shuts off. Takes around 10ms or less to get back.

[@contomo5710]

while i do like the working of PLL to control the power, something in my brain is keeping me from doing inefficent things, if the power factor isnt 1 im just not happy ^^ but it is very reliable, easy and well behaved!

[@T3sl4]

@@contomo5710 A nonlinear control with rapid-rise slow-fall behavior would run just fine at 120Hz ripple, if you mean to run it unrectified. The setpoint could also be modulated by Vin, so that Iout is controlled proportionally, and at normal control bandwidth (but mind the singularity, perhaps use a OTA to stop control movement at low Vin).

[@BrianDhvTinkerer]

@T3sl4 thanks!!

[@contomo5710]

@@T3sl4 not even starting to talk about mains power factor, just saying those switches need to start handling a lot as you get into higher power systems, not just the true power but also the reactive going into the off resonant tank

[@contomo5710]

i really really like this design! its very very nice to see different approaches to mine!! i also like monitoring the voltage, (be careful with the voltage on those chinese CCC caps, i blew one up before!) its quite opposite to my op amp peak detector and 74hc74 latch, all in all, very nice circuit and very nice build! i cant imagine how much levreage and packed sand/water it took to bend the tubing that tight without kinking it haha this is the level im happy with ^^ oh maybe one thing i wouldnt have done like it is, the current shunt on the small wire, you never really know the current through it, i think it would have been much simpler/easier/more accurate to just place the current transformer on the primary side, if you have 16 turns, the current through it is just 1/16th

[@BrianDhvTinkerer]

Thanks that may be worth trying. Thought about doing that at first, but I’ve always had better success with secondary feedback. I’ll mess around with that in my next experiment.

[@contomo5710]

@@BrianDhvTinkerer feedback for the overcurrent, not for the tracking, something i did for both tho is just have two fairly decent sized wires running back to the controllor, something that can do 1-2A, and place a second smaller resonance capacitor there (twisted wires going to it) ie if you have two a 10uF resonance capacitor in total in your tank circuit resonating at 1000A, a 10nF one would 1A resonating through it, that way you can just use an off the shelf current transformer, or just use a self wound one with much less requirements. 10uF/10nF => 1000A/1A

[@BrianDhvTinkerer]

Thank you. Bending that 3/8” copper was really tough but I was happy to get it without a major kink!

[@haronwepworking5806]

Hey Brian! Great work. I'm curious about water cooling, as I'm planning to water cool the IXDN614CI gate driver in my project, and I'm using a water-cooled Illinois capacitor. The coil and the capacitors are being water-cooled on the same water path. Essentially, I'm planning to water cool everything as follows: water enters the water block that houses the rectifier and the IGBT, then the water passes to another water block which contains the two gate drivers. From there, it flows through one of the tank's leads to the capacitor, then to the coil, and finally exits from the other lead back to the original source, which is the iced water bucket. My question is, how do you ensure there are no short circuits due to the water cooling shared among all these components?

[@BrianDhvTinkerer]

The gate driver chips don’t really generate that much heat and so don’t really need the water cooling. I would not put them on the same water cooling path as the high voltage electronics. However, if you have already set it up- shouldn’t matter. The iXDN share a common negative

[@BrianDhvTinkerer]

I’ve done the finalization on the project. It turned out that the feedback using the coupling transformer was not reliable. So I have it all set up and working great. Please check out the revised link. Link is here: teslascience.wordpress.com/simple-and-effective-induction-heater-current-control-circuit/

[@Magneticitist]

Hey can you point me to where you found those awesome resonant capacitors?

[@BrianDhvTinkerer]

I got them off AliExpress. Not cheap but well worth the money because they indestructible: Just found this amazing item on AliExpress. Check it out! $75.00 | 2UF 500V 650A Solid-state high-frequency resonant tank circuit machine low sense of film capacitors a.aliexpress.com/_mM9JKbe

[@Magneticitist]

@@BrianDhvTinkerer Oh wow, that's not bad at all. I was expecting something twice that cost or more. Awesome, thanks!

[@BrianDhvTinkerer]

I got a new build on the current handler l. Got some ideas off hvforum. This one will work at any frequency. Boards still in production. Once I get em, will test out and I can send you one if they work good.

[@Magneticitist]

@@BrianDhvTinkerer Awesome I would appreciate that. I haven't done much with the feedback but I can see how feedback + current limiting sounds the most appealing. I haven't gone over about 450W yet so I've been able to use a high ratio CT I had but I suppose at a certain point that's not going to cut it. I tried the PLL approach first but I don't like the idea of manually detuning to try reducing no load power. Right now I'm just using the VCO to manually tune cause I have to bring it way over resonance just to reach over 50VAC in without exceeding about 5A. I think my caps might pop doing over 600W.

[@BrianDhvTinkerer]

@Magneticitist sure no would be happy to send you a board

[@whatevernamegoeshere3644]

4:39 is it actually boiling or off-gassing?

[@BrianDhvTinkerer]

I think it’s boiling. Mp and bp for copper are close

[@tsclly2377]

you need to modify the induction coil so that it and the crucible can rotate to dump the content..

[@BrianDhvTinkerer]

Good point!

[@abeleski]

I want one

[@BrianDhvTinkerer]

It’s a fun project. Definitely worth messing with.