Induction Heater IGBT help

Status
Not open for further replies.
The neon john circuit is a so called LCLR parallel resonant tank with a series impedance matching inductor.


When you increase the frequency, you get closer to resonance, which increases the current in the inverter, if your mosfets are undersized, they will pop as the current exceeds their ratings.

If you increase frequency even further, current should drop off again

Heres a video from my project using four IRFP450, two parallel pairs: http://www.youtube.com/watch?v=DpOcGPsennU at 1:36.

Here i have a peak detector with a current transformer showing the peak inverter current: http://www.youtube.com/watch?v=jiRwKpmBGqw

The analog meter shows the current beeing drawn from the mains.
 

CDRIVE

Joined Jul 1, 2008
2,219
The neon john circuit is a so called LCLR parallel resonant tank with a series impedance matching inductor.


Heres a video from my project using four IRFP450, two parallel pairs: http://www.youtube.com/watch?v=DpOcGPsennU at 1:36.

Here i have a peak detector with a current transformer showing the peak inverter current: http://www.youtube.com/watch?v=jiRwKpmBGqw

The analog meter shows the current beeing drawn from the mains.
Yo! Dino De Laurentiis, is a security clearance required to post the link to Mr. Neon's schematic? His name keeps getting tossed around here but for some reason it still isn't posted ... and I have no intention of searching for it. While entertaining.. sort of... your vids don't help in resolving this thread. :rolleyes:
 

CDRIVE

Joined Jul 1, 2008
2,219
The neon john circuit is a so called LCLR parallel resonant tank with a series impedance matching inductor.


When you increase the frequency, you get closer to resonance, which increases the current in the inverter, if your mosfets are undersized, they will pop as the current exceeds their ratings.

If you increase frequency even further, current should drop off again

Circulating currents within a Parallel resonant (tank) circuit are very high at resonant frequency but the external input current decreases sharply at resonance. This translates to increased Z looking into the tank. Also, the signal voltage across the tank at resonance rises sharply.


The above statements are bases on an unloaded (Hi Q) tank, while your metal stock is your load, which will lower the Q dramatically via coupling the energy into the stock.

Since you have built a working model I think you can answer the following question...

If you remove the metal stock from the crucible does the input current drop?
 

CDRIVE

Joined Jul 1, 2008
2,219
Actually, I think you pretty much answered the question. My guess is that initially input current would drop. Voltage across the tank would increase dramatically, and finally.. input current would spike when the components broke down from over voltage.
 

Thread Starter

yappers

Joined Apr 13, 2010
16
Actually, I think you pretty much answered the question. My guess is that initially input current would drop. Voltage across the tank would increase dramatically, and finally.. input current would spike when the components broke down from over voltage.
So your suspicion is that over voltage that is building up in the tank circuit is a problem? Hmm....I wonder if there's a way to calculate that voltage buildup.

Also, I tried taking a look at other induction heater design including the video that Experimentonomen just posted and there is something I that is bothering me. From what I know, the number of capacitor in the resonant tank is used to spread out the current that is going into the resonant tank, but what is bothering me is that I don't know how to determine the number of capacitors. In Neon John's schematic, it uses a set amount of capacitance, but physically, there are 24 0.022uF, 400 volt caps and 2 0.1uF 400 volt caps for a total capacity of 0.7305uF. For my purposes, I could only get my hands on some film capacitors that when put in parallel get me close to 0.66 μF. Also, here is a picture of my circuit and my coil, for those who curious

Except my circuit was series resonant. I don't use the parallel resonant LCLR circuit anymore.

What was the reason? Because it sounds like you tried it
 

Attachments

CDRIVE

Joined Jul 1, 2008
2,219
Except my circuit was series resonant. I don't use the parallel resonant LCLR circuit anymore.
Yes, a series resonant circuit is a whole different animal. At resonance Input Z drops, thus input current increases and source voltage drops do to source Z.

Can we see your circuit?
 

CDRIVE

Joined Jul 1, 2008
2,219
So your suspicion is that over voltage that is building up in the tank circuit is a problem? Hmm....I wonder if there's a way to calculate that voltage buildup.
Not really. It's more akin to thinking aloud.


You need an oscilloscope to see whats going on at your tank circuit and check the inverter for ringing as that is what killd mosfets/igbts.
I agree. This is a must for this kind of work.

Edit: Yappers, where did you find that coil design? I know you said that you're trying to heat sheet stock but that looks like 14AWG wire and I would think that if it did work the insulation would melt from the hot sheet stock. Besides, that gauge wire is far too small for the currents we've been discussing here.

EDIT, EDIT: Actually, that looks like 16AWG!!
 
Last edited:

Thread Starter

yappers

Joined Apr 13, 2010
16
You need an oscilloscope to see whats going on at your tank circuit and check the inverter for ringing as that is what killd mosfets/igbts.
I tried scoping once before. However, the many times I have done this, the oscilloscope shorts to earth ground killing both my FETs as well as damaging my IC. Now I'm really afraid of using my o-scope to touch anything that is higher then 30 VDC.


Not really. It's more akin to thinking aloud.

Edit: Yappers, where did you find that coil design? I know you said that you're trying to heat sheet stock but that looks like 14AWG wire and I would think that if it did work the insulation would melt from the hot sheet stock. Besides, that gauge wire is far too small for the currents we've been discussing here.

EDIT, EDIT: Actually, that looks like 16AWG!!
Actually, it is 12 gauge. But I doubt you're satisfy with that. The mess called the coil is something I came up with and am using for the time being just because I want to start out small, then scale it up. Heck, it not even suppose to rival the induction heaters I've seen on youtube, but if I can feel heat from something this small, then I'm sure I can scale it up with more confidence.
 

CDRIVE

Joined Jul 1, 2008
2,219
I tried scoping once before. However, the many times I have done this, the oscilloscope shorts to earth ground killing both my FETs as well as damaging my IC. Now I'm really afraid of using my o-scope to touch anything that is higher then 30 VDC.
<snip>
All scope measurements must be referenced from ground. With these high voltages you should be using a 10x probe. Are you doing otherwise?
 
I never drew a circuit of my build.

It was basically around 100 150nF 1000V film caps in parallel and the core from a 3.5kW smps as the coupling transformer, 2 turns secondary and around 20-30 turns primary.

The control circuit was a basic PLL setup with a CD4046 with the phase comparator input reading the tank voltage. Fres was around 30kHz.

Another video: http://www.youtube.com/watch?v=3gHDi9HHdIE
 

wayneh

Joined Sep 9, 2010
17,496
I never drew a circuit of my build.
In the real world, there's a useful expression: "If you didn't write it down, you didn't do it."

There's only about one good thing about getting older, hopefully you become wiser. Take a word of hard-won wisdom, remember the expression, and document your work. You won't regret it.
 

CDRIVE

Joined Jul 1, 2008
2,219
In the real world, there's a useful expression: "If you didn't write it down, you didn't do it."
"Schematic!" "I donn née no stinkin schematic!" :D You have to be old enough to get this. ;)

By the way, there's a rumor that these videos have caught the eye of several safety agencies. Last I heard, one of them are holding a strong 2nd place for OSHA's Safety Video Of The Year. :rolleyes:

Just kidding cause I'm no saint. ;)
 
Status
Not open for further replies.
Top