Asides

Induction Heater

IMG-20160823-WA0004

Common information

I created this induction heater based on this schematic from http://www.rmcybernetics.com:

induction-heater-schematic

BOM

One of the most important things would be the BOM I used. On the internet you’ll find a lot of suggestions. Due to the fact, that I want to use this heater as an induction furnace, I modified some parts so that the whole system can handle more power.

  • R1, R2: 220R, 5W (yes, a little bit overpowered)
  • D1, D2: UF4007
  • T1, T2: IRFP260N
  • C1: 1,98µF
    • 6 x WIMA  MKP10 630V 0,33uF in parallel
  • L1: as shown on http://www.rmcybernetics.com
    • 6mm copper tubing with 60mm inner diameter – yes, we will need some active water cooling 🙂
  • L2: 2mH
    • 1,5mm magnet wire on a ferrite toroid (ø37x21x13,3mm).

Power supply

I have two 24VDC/12,5A SMPSs in parallel. These bring sufficient power for heating medium pieces of iron until they’re glowing. For an induction furnace I will need more power to heat my graphite crucible to at least 700°C for melting aluminum (melts at 660,4°C).

Evolution of construction

Here I offer you the chance to have a look at my progress of creation. I hope this helps you to avoid some unnecessary mistakes.

Back to the roots – wood and metal!

IMG-20160811-WA0001

Because my breadboard went on fire I had to change my prototyping platform. Yes, beyond the driving circuit there is a lot of power!

First (working) assembly

IMG-20160811-WA0004

  • water cooling
  • circuit is working 
  • insufficient cooling of T1 and T2
  • CAPs attached externally: a lot of heat on the wires and capacitors

IMG-20160822-WA0004

But hey… its working! 😀

Next iteration

IMG-20160815-WA0002

  • CAPs now mounted on L1 directly:
    • better power/heat transfer
    • water cooling affects capacitor temperature as well
  • T1, T2 cooling improved
    • bigger heatsinks
    • active fan powered by a LM317 circuit at about 12VDC (when I recoil the choke this can be removed)
  • T1, T2 cooling still not sufficient 🙁

Fun fact: This construction has a clever safety feature! If the FETs get too hot (yes, really hot), the attached drain wires snap off. 

More cooling!

IMG-20160823-WA0002

  • T1, T2 heatsinks connected to existing water cooling. Fortunately these chunks of aluminum had a horizontal hole all the way through!

Some results

received_879516348849358

  • I need at least 700°C!
    • cooling is working perfectly
    • need more power! 😀

IMG-20160813-WA0005

 

Facebooktwittergoogle_plusredditpinterestlinkedinmail