Yeeeeha! Microprocessor controlled induction heater / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Yeeeeha! Microprocessor controlled induction heater

first previous 3 4 5 6

Move Thread

LAN_403

The Ideanator

Sat Sept 19 2015, 07:42AM

Registered Member #54599 Joined: Sat Mar 07 2015, 06:09AM
Location:
Posts: 8

I've not been able to find your code anywhere and am trying to write my own in the meantime, so I have some questions.

Does the uP simply scan frequencies for one that looks like resonance and then let the PLL take over?
Will the tank tend to stay at f_res even when it shifts and thus would encourage the PLL to track? How well does it track on its own? Would it be necessary to have the uP track f_res and kick the PLL towards it?
Does the uP look for the peak current draw to determine f_res or is there another way to find it?

I'm sure a few days of trial and error would answer these questions and perhaps answer more that I haven't put my finger on. I relish the challenge, but the thought of blowing something expensive up gives me some pause and since you've already succeeded, I may as well ask how you did it.

IamSmooth

Mon Sept 21 2015, 11:58PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

This was a while ago and I sold the unit to a blacksmith. I don't have the code. The unit would like right on the frequency no matter what the object was inside the coil. It would stay locked on even as the Fres changed with heating (and it does change).

I had a large range the uP would scan during bootup. I would start high and work down. During this process I would track the output from the PLL and the primary current. I found, using a scope and multimeter, how the PLL output changed with frequency. I developed a table that let me determine when I was right on top of Fres. Once I had this value I entered a loop that kept checking the value and stayed a little above Fres. I had a subroutine that monitored the current. If the current got too high I detuned the tank by raising the frequency.

The website should go over the input to the PLL and the appropriate output.

This heater stayed dead on the frequency and could handle a sudden lose of load without blowing up. If I was to do it again I would add a temperature input to monitor the mosfet surface temperature. I would shut the unit down if they overheated.

If you have not built one yet I would make a simple PLL-based circuit that you tune by hand with a potentiometer. The PLL will get you close, but not quite on it and it will not track the change in Fres. This is the reason I went with the uP. However, you should invest some time into the simple circuit so you can see what is happening and understand the data the various components offer you for constructing a more sophisticated circuit. If you jump into a big circuit you will not appreciate everything that is happening, you will have trouble debugging and you will most definitely blow some expensive switches.

Wolfram

Tue Sept 22 2015, 10:20PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

IamSmooth wrote ...

If you have not built one yet I would make a simple PLL-based circuit that you tune by hand with a potentiometer. The PLL will get you close, but not quite on it and it will not track the change in Fres. This is the reason I went with the uP. However, you should invest some time into the simple circuit so you can see what is happening and understand the data the various components offer you for constructing a more sophisticated circuit. If you jump into a big circuit you will not appreciate everything that is happening, you will have trouble debugging and you will most definitely blow some expensive switches.

What you're describing is a free-running oscillator, not a PLL. A correctly implemented PLL will track the resonant frequency of the tank. Fast hardware overcurrent protection is essential to avoid blowing expensive parts in high power induction heater design.

IamSmooth

Thu Sept 24 2015, 03:10AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Wolfram wrote ...

A correctly implemented PLL will track the resonant frequency of the tank.

Unless there is a really robust PLL design, my PLL inverter did not accurately track the changing Fres. Yes, it did get close and it did change as the frequency changed. However, it was still became off over time as the metal went through the curie point.

The Ideanator

Fri Sept 25 2015, 08:05AM

Registered Member #54599 Joined: Sat Mar 07 2015, 06:09AM
Location:
Posts: 8

I had a large range the uP would scan during bootup. I would start high and work down. During this process I would track the output from the PLL and the primary current. I found, using a scope and multimeter, how the PLL output changed with frequency. I developed a table that let me determine when I was right on top of Fres. Once I had this value I entered a loop that kept checking the value and stayed a little above Fres. I had a subroutine that monitored the current. If the current got too high I detuned the tank by raising the frequency.

Yep, that's what I figured it would be like. Thanks!

Patrick

Sun Nov 29 2015, 07:43PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

When you guys use high power induction heaters and SSTC's, do you power them up using a variac until your sure you've got the bugs worked out? or just use a fuse and plug it in? I ask because I'm moving more into the line operated circuits.

IamSmooth

Mon Dec 07 2015, 01:24AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

I always powered it up with a variac or other power-varying device (boost converter). I got it working for a while with the latter, but blew the mosfets. Getting it right wasn't what I wanted to invest time in, so I went back to the variac. The variac was 240vac rated at 50A continuous. It was a beast.

The reason for the variac is my unit had to find the right frequency. If the frequency was off at a high power level the mosfets would blow. Once I had a lock, I could ramp it up all the way in seconds.

first previous 3 4 5 6

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