Induction Heater Help / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Induction Heater Help

Move Thread

LAN_403

Wilson

Tue Feb 14 2006, 10:43AM

Registered Member #78 Joined: Thu Feb 09 2006, 11:27AM
Location: Sydney, Australia
Posts: 133

Hi
I"ve recently built an induction heater, utilising the LCR arrangement as outlined on richie burnett's site.

However, for some reason, i cannot get a ferrous object to the curie temperature. At first i thought it was due to my low coupling, so i wound a tighter coil, but with no luck.

Below the waveform from a current transformer in series with the inverter's output, at 1A/div and 45V in. Its close enough to sine wave, so i'm assuming that the system is in parallel resonance (at 131khz). The voltage across the tank also rings up to 160V, almost the limit for my IRFP250s.

Any idea why my workpiece isn't heating up? I've seen people with glowing nuts at 300W after a minute or so. I left mine on for 10mins and was still <400 degrees.

Thanks,

Wilson

Hazmatt_(The Underdog)

Tue Feb 14 2006, 10:17PM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Keep trying you'll get it!

I had a hard time too when I wanted to copy an idea from a friend. He mods PSU's for brazing. I tried and tried, and finally got the coil to about where it needed to be to see heating results. It was just trial and error.

Wilson

Wed Feb 15 2006, 01:42AM

Registered Member #78 Joined: Thu Feb 09 2006, 11:27AM
Location: Sydney, Australia
Posts: 133

what i don't quite understand is why when i have no load in the work coil, the current draw is still ridiculously high.
To my understanding, an unloaded LCR based induction heater is purely reactive, the only resistance, which could draw/dissipate power besides the workpiece would be the ohmic resistance of the connecting wires, and the ESR of the caps. It shouldn't draw an odd 300W with no workpiece

I'm waiting for a second CRO probe so i can look at both the inverter's output and CT at the same time.

Steve Conner

Wed Feb 15 2006, 10:52AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

2A peak at 45V is only about 90W of real power, so that is why your nuts won't glow. It also looks like you are out of tune, since the switching noise is visible near the peaks of the sine wave, whereas it should be near the zero crossings. Adjust the drive frequency and the size of the matching inductor to get more power. An ammeter that shows the DC current drawn from your power supply is useful here- you can tune the drive frequency for maximum current.

The voltage across the tank circuit has nothing to do with any voltage seen by your MOSFETs: they only see the DC supply voltage, which in your case is 45V.

About unloaded current draw: I found that with my induction heater the current draw increased with no workpiece. Because it's series resonant, an increase in load resistance across the work coil gets transformed to a decrease in the load resistance seen by the inverter. I think the LCLR behaves in a similar way. To stop a catastrophic meltdown if you remove the workpiece with the power on, you need a current limited power supply or some other form of protection.

Wilson

Wed Feb 15 2006, 12:34PM

Registered Member #78 Joined: Thu Feb 09 2006, 11:27AM
Location: Sydney, Australia
Posts: 133

Sorry, my bad. I meant to say 1V/amp. The peak is at 11V, so peak of 11A.

I'll try adjust the matching inductor. What would be a good value? enough reactance to not exceed the current limit of the supply?

I'm glad to hear that the voltage ringup is only confined to the tank. This means i can take my bridge all the way up to 160VDC.

I'll buy an extra probe in a few days, so i can see the mosfet's gate waveform vs that of the currrent transformer.

By out of tune, do you been by +/- 1khz, or by a whole harmonic (not too sure if this even makes sense). What is quite intriguing is that at some frequencies, i can get almost 1/2 div of that noise at the peaks.

I have also found that the LCLR arrangement behaves in the same way to your series resonant one- it draws a bit more current when there is no workpiece. However, i don't see how this is possible as when the system is in parallel resonance, the impedance should be at its highest. Only when there is a workpiece that there is a resistive component which can draw and dissipate power :S
However my weird results might be due to detuning.

Thanks alot for your help Steve.

I'll try to upload a pic of the dual scope shot in a few days if school isn't too harsh

EDIT:
I"m considering upgrading my tank capacitor, as it looks like the old one isn't up to the job. How does metalised polyester caps (like the ones below) fare against the recommended PP caps in this application?
(<link>http://cgi.ebay.com.au/ws/ eBayISAPI.dll? ViewItem&item=7590457256&fromMakeTrack=true</ link>)

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