1

I have been tasked to gather data from an old running 335kHz 5kW induction heater system. It has no outputs available to give amperage, power etc. Basically there is an assumption that the unit is not doing a good job of controlling a constant power output. Given a constant power set point.

I was thinking about trying to measure the RF field generated by the unit to determine how consistent it is. Although most power meters for RF are for >1MHz... So I was considering using an AM radio front end design to receive and rectify the RF across a resistor, then log the analog voltage. The image below is a basic idea, obviously I would recalculate L,C to be resonant.

This question is also giving a solution similar... Measuring relative rf signal strength

I don't care about absolute accuracy, just relative. I can mount an antenna on a PTFE plate that will not move, so that controls device distance etc. Any feedback or alternative options?

enter image description here

W5VO
  • 18,303
  • 7
  • 63
  • 94
MadHatter
  • 3,506
  • 1
  • 19
  • 45
  • The fields nearby the heating coil might not accurately reflect the power delivered to the work, as the work will affect the fields. To get an accurate figure i think you'll need to measure voltage and current and phase angle. Try it with an oscilloscope before you build anything. – tomnexus Apr 16 '19 at 18:15

2 Answers2

1

I would think measuring the magnetic field of the induction heater with another coil might be the way to go. With an inductor the primary filed is magnetic. Since magnetic fields fall off with the distance cubed (1/R^3) it would be difficult to pick up other sources of magnetic noise at 300kHz. You could find a coil with a response at 300kHz or a magnetic sensor with a response at 300kHz.

Voltage Spike
  • 75,799
  • 36
  • 80
  • 208
  • Interesting information. Do you think a coil of wire with a band pass filter would have a similar effect to the purchasable ones? At least for testing so I don't need to wait for shipping. – MadHatter Oct 18 '18 at 18:34
  • Yeah, I think a coil\inductor with a response at 300kHz would be the best option, if the signal was too low you could amplify it. I would think you'd be able to see something with just a coil and a scope. If not, you could throw in an opamp after the coil. – Voltage Spike Oct 18 '18 at 18:38
  • My thought exactly. Thanks I'll give this a try while waiting on one of the antennas you posted. – MadHatter Oct 18 '18 at 18:41
0

Well per @laptop2d advice, I made a rectangular loop antenna and a resonant filter. Then used an op-amp to buffer the signal, rectify and read out on an analog recording system.

It is working well for tracking relative changes. The parts here are not all ideal, but it is what I had laying around. The schematic does not show it, but the op-amp input is coupled via a 0.1uF capacitor and biased with 1M ohm resistors to set it at half the rail voltage.

enter image description here

MadHatter
  • 3,506
  • 1
  • 19
  • 45