I'm trying a different induction heater topology, but... / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

I'm trying a different induction heater topology, but...

Move Thread

LAN_403

IamSmooth

Mon Jan 04 2010, 07:04AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

I had really good success with my transformer coupled series resonant tank induction heater. See here in case you are not familiar. The schematics are at the end. I decided to try something different. The former topology had a coupling transformer connected to the inverter output. This energized a one-turn primary running through it which is connected to a capacitor and work coil. The coupling transformer barely got warm, but I was losing power in heating the work coil.

Instead, I wound a 40T primary of 14g wire around the powdered iron toroid and connected a capacitor in series with the primary. This LC pair is connected to the inverter output. I have 8T of 10g as the secondary which is connected to a 5T work coil. With very little input power I can get the workpiece warm. With under 100w and a little over 10v on the input, the primary coil has 150v across it. The work coil has about 30v. I verified that the inverter voltage and current are in phase with an oscilloscope. You would think this is good, except my toroid transformer is getting way too hot. With very little input power, the wires start to smoke and the work piece is only at the "too hot to touch" point. I feel I am not transferring the power into the workpiece. Am I doing something wrong? Am I saturation the core and need to stack some more toroids?

GeordieBoy

Mon Jan 04 2010, 02:16PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

You are right in your feeling that you are not transferring the power into the workpiece!

The problem with what you have done is that by placing the resonant capacitor on the primary side of the coupling transformer you now have to transfer the reactive VARs as well as the real power through this transformer. In essence the poor coupling transformer sees very high voltage and very high current, but these are not necessarily in phase so the real power throughput may be quite small. The high current causes lots of conduction loss in the windings of the coupling transformer, and the high applied AC voltage causes large core losses.

The choice of powdered iron core material only makes this situation worse since it is very lossy for large applied AC voltages and the coupling to the work-coil will not be as high. A power grade ferrite of the Mn/Zn formulation would give much lower core losses and tighter coupling. However this topology is still the wrong way to go about induction heating.

There are two popular schemes: The series resonant scheme you used before, and the LCLR parallel resonant scheme. I personally favour the LCLR scheme as this keeps the large circulating current of the work-coil out of the inverter and does not require a high-frequency high-power ferrite transformer to work, (although one may be used to achieve galvanic isolation from the mains supply line.) Some people have trouble determining the matching inductor and fabricating this part though.

The transformer-fed series-resonant scheme is equally useable. In my opinion it just has the downside of requiring a high-frequency high-power ferrite transformer with water cooled secondary.

The thing you need to remember about IH is that the voltages and currents seen by the work coil in this application are both huge for a given power rating. What I am saying is that the phase angle between the voltage and current is such that the voltages and currents are disproportionately large for the amount of real power that goes into the workpiece. The two commonly used topologies are chosen because they act to keep these large voltages or currents out of the inverter, and ideally out of any power transformers too. You want the reactive power to circulate as close as possible to the work-coil itself, and keep it out of the inverter or any transformers.

I hope this helps,

-Richie,

PS. I'd also like to say well done on your Induction Heating web page. There's lots of detail there for experimenters dabbling with IH.

Mates

Mon Jan 04 2010, 02:40PM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

Hi Iam,
it is a solid piece work you made on the induction heater guide! Good work!
I have a question - a bit out of topic. The idea of placing a Zener diode in series with the FET gate - where this came from? Could you comment it a bit?
Thanks
Mates

IamSmooth

Mon Jan 04 2010, 06:27PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Thanks, Richie. I'm glad you like the tutorial. As everyone must notice, a lot of power is wasted heating the workcoil with R*I^2. Is there a way to limit the power lost? Would increasing the work coil turns solve the problem, or does this introduce other issues? I know for one thing it will increase the coil inductance and lower Fres.

As far as the series diodes...

Mates, I have to give credit to John Dearmond. His website is here.

You can view the timing tracings here.

When I put together my first inverter using the TL494 there is a timing delay built into the chip. This can be increased with the dead-time controller input. This ensures that both mosfets are not conducting which would short the HV rails. I figured that there is a small Ton delay when the gate is turned on, but this delay may not be enough. The transformer gate drive will simultaneously go high on one side, and low on the other. If I am correct, it will only take 4v on the gate before the mosfet goes on, and this will occur before the other is off.

By adding the zener, the breakdown voltage has to be exceeded before the gate voltage starts to rise. Now, the slope is very steep, so this does not translate to a big temporal difference, but a difference none the less. If you look at the superimposed gate tracings with one of the diodes shorted you can see the delay to the gate caused by the diode. further down on the page you can see the delay in the time it takes for the gate to reach the same voltage when compared to the tracings without the diodes. I experimented with different values. There is nothing magical about 5.1v. However, too large a value and you won't have a smooth transition and you'll get overshoot; to soon and you will short the rails and draw excessive power. You may be able to get away with no timing delay diodes, but I thought at the power levels I was reaching I did not want to take a chance.

The only smoke I wanted to see was off the workpiece.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob