| ▲ Electronics |
I wanted to see if it was possible to modify an induction cooker to use an external work coil, for heating actual metal parts. The cooker in question is this one I got ages ago. It uses a quasi-resonant drive circuit (single IGBT) with a parallel-resonant tank.
The original tank capacitor is 270nF. The original pancake work coil is 118µH, 65mΩ DC resistance and is made from Litz wire. I removed the tank capacitor and work coil, wired the up externally and proceeded to measure the effective parallel loss resistance across the tank for different loads:
The maximum loss resistance which still causes the cooker to switch on seems to be around 200Ω - any more, and it displays an error message (corresponding to "pot not present") and switches off.
I made up a bobbin from bits of PVC plate and tube and wound a new coil. Bobbin dimensions are roughly 30mm ID, 100mm OD, 40mm thick (see here for a PDF). Wire used was about 4mm diameter 7-strand copper mains cable. Approximately 60 turns gave an inductance of 118µH and a DC resistance of 48mΩ, a good match to the original coil. The new coil has an unloaded parallel loss resistance of 350Ω, probably because of skin effect losses. A nice coil could be wound using copper foil (for example, from http://stores.ebay.co.uk/Der-Kupferladen).
Once hooked up in place of the original coil, it did actually heat things! It managed to heat a bit of ¾" steel tube to red heat. However, objects need to be a resonable size in order to present a low enough loss resistance that the cooker stays running.
 Original coil |
 New coil |
 View showing the coil connected to the main PCB |
 |
 I had previously removed the tank capacitor to measure it accurately, so it's wired on with flying leads here. |
| ▲ Electronics |