Current limiter for induction heater / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Current limiter for induction heater

1 2 3 next

Move Thread

LAN_403

Tonskulus

Mon Mar 02 2009, 09:49PM

Registered Member #1223 Joined: Thu Jan 10 2008, 04:32PM
Location:
Posts: 133

In case of LCLR topology:
Would it be good idea to use constant current source instead of Zmatch inductor? Because I have noticed some serious problems using Zmatch coil to limit input current.. It should be adjustable or something to keep input power always constant no matter how much there is load in workcoil.

If there is fixed Zmatch coil, input power varies with different workpieces. As example, heavy steel tube inserted to workcoil and input power drops from 1000W down to 300Watts before curie point is reached..after this point, it climbs up to 1kW again.
So before curiepoint, heating effect is too low.
Up to this point, I have removed ferrites from zmatch to reach curiepoint faster and then ferrites has to be plugged in again or input current rises too much.

I would like to have a system that keeps input power always the same. How about simple forward converter style constant current source?

MRacerxdl

Mon Mar 02 2009, 09:55PM

Registered Member #989 Joined: Sat Sept 08 2007, 02:15AM
Location: SÃ£o Paulo, Brazil
Posts: 476

Maybe using a OCD like DRSSTC's? I think it might work, and stabilize the current, but I dont know how the workpiece will go in that way.
But in theory, that is a good thing, you can adjust the current limit limit a potentiometer.

uzzors2k

Mon Mar 02 2009, 10:52PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Weren't you the guy with a mechanical Lmatch inductor controlled by a servo? At any rate I plan to make an air gap adjuster for my PLL induction heater which will be as easy to use as a potentiometer. You can use the same topology with just a little bit of mechanical handywork.

GeordieBoy

Mon Mar 02 2009, 11:32PM

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

Power throughput in LCLR is usually controlled via variable frequency control on the high-side of resonance, or sometimes by varying the DC link voltage.

If you think about it, you don't want constant input power regardless of load. If you pump constant power into the low-loss tank circuit when the workpiece is small of absent this will pump up the tank voltage and circulating current to very high levels. It's better to control the work-coil current, so that it doesn't skyrocket when the workpiece is absent or a magnetic workpiece goes through Curie.

-Richie,

Steve Conner

Tue Mar 03 2009, 02:03PM

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

My induction heater used one of my DRSSTC driver boards. They have an OCD that pulses the gate drive on and off. This seemed to limit the work coil current fine, although it made some audible noise. I prefer a series feed arrangement to LCLR, so for me limiting the inverter current was the same thing as limiting the work coil current.

Note that this method puts a lot of stress onto the antiparallel diodes of your power devices. The current commutates to them when the gate drive is off, so they may suffer nasty forced recovery every time it comes back on. If it dithers on and off at high frequency then the diodes may overheat.

I was using IGBTs with ultrafast co-packaged diodes, and my PLL driver only allowed whole cycles, so it didn't seem to be a problem. But MOSFETs might need their body-drain diodes isolated.

You can also run your DC bus off a current limited power supply.

GeordieBoy

Tue Mar 03 2009, 02:39PM

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

If the PLL controls the inverter current to be exactly in-phase with the inverter's output voltage then the free-wheel diodes shouldn't see any forced reverse recovery when the inverter starts up again. In fact they will only carry current during the deadtime within the burst, and of course during the time when the inverter is actually disabled by the OCD and the current is ringing down.

This is probably easier to do with a series resonant circuit because you can get the inverter load current exactly in-phase all of the time. For the LCLR arrangement it is usually slightly lagging, but this is better for MOSFETs as it allows ZVS.

You can surely only get away with this burst-fire power control approach because you are using a PLL and the inverter always comes back up with the voltage in-phase with the ringing tank current when the current falls to the "power up again" threshold. It would not work at all well with a fixed frequency oscillator as the phase would wander when the inverter's output was gated off.

-Richie,

Tonskulus

Tue Mar 03 2009, 03:00PM

Registered Member #1223 Joined: Thu Jan 10 2008, 04:32PM
Location:
Posts: 133

Varying DC link voltage is almost what im looking for. It works same way as constant current source anyway.
Manually adjusted Zmatch coil is not an option, as it has to be automatically controlled however.
Its not good to always adjust zmatch impedance manually when iron reaches curiepoint. It has to be autoamtic.
I dont care about tank voltage rising "too high" if there is too low load. Its not a problem if tankcapacitor can handle that current. And circulating tank current is not infinite, it is limited by LC impedance and by losses in tank circuit itself.
And if im heating low loss workpieces like copper or aluminium or "over curiepoint steel", huge amount of workcoil current is required to get enough heating power.
In case of iron, it is so easy to heat up to curie. We dont even need any parallel resonant circuits for that.

I cannot adjust power by turning frequency over resonant as my induction heater uses direct feedback from tankcircuit. There is no separate oscillators (except starting oscillator).

Ok, i was thinking about using my 3kVA servocontrolled variac to get variable dc link voltage :)

GeordieBoy

Tue Mar 03 2009, 03:14PM

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

> I dont care about tank voltage rising "too high" if there is too low load. Its not a problem if tankcapacitor can handle that current. And circulating tank current is not infinite, it is limited by LC impedance and by losses in tank circuit itself.

In a well designed induction heater the losses in the work coil and induction heating tank capacitor will be low compared to the losses when the workpiece is inserted. This achieves some reasonably efficiency, but it means that the unloaded Q of the tank circuit will be very high. Therefore circulating current and tank voltage can really skyrocket if you keep pumping in constant power from the inverter with little or no load to soak up the power. It's not a case of "the tank capacitor can handle it". In my opinion if the tank capacitor can comfortably handle the high VARs under this out-of-control resonant condition, then it is not being very well utilised when actually heating lossy workpieces!

-Richie,

Hernan

Sun Aug 09 2009, 05:06AM

Registered Member #1614 Joined: Wed Jul 30 2008, 03:08PM
Location: Argentina
Posts: 52

HI guys , I decided to build a mechanical Lmatching coil with some parts from an old scanner
I've finished the water cooled with mechanical adjust matching inductor for my LCLR IH to control the current to the tank circuit. I use a stepper motor to open or close the core . but now I would like to use a control circuit to keep the current constant... for example like tonskulus said when steel load reaches curie point the inverter current starts to increase so I have to close the Lmatch core to keep the current at save levels. I plan to do this with a pic microcontroller with some PI routine that compares a setpoint current from potenciometer with the inveter current and then drive the motor to open or close the core automatically so I`ll preciate the suggestions with the routine because I have no idea how to start with the program.

L match picture

Steve Conner

Sun Aug 09 2009, 06:39PM

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

Hope that clamp isn't a shorted turn around your core! :P

By the way, what's wrong with just waiting until the steel reaches the Curie point? My induction heater did the same with steel workpieces, I could see the line current shoot up as it passed through Curie.

1 2 3 next

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