How to design the correct choke value for 30 amps?

[gary350]

Since my electronic class had no Lab I am trying to learn this on my own 50 years later. My digital meter shows my home made choke is 6.6mh. These mosfets explode at 35 amps I would like to have current no higher than 30 amps. Ohms law shows a .4 ohm resistor will limit current to 30 amps. XL=2 pi FL online calculator shows my choke is 2,654 ohms. My digital metal shows this circuit runs at 64KHz with no load, a 3/8" diameter steel rod drops the frequency to 60 KHz. With the 6.6uh choke in the circuit meter reads 1.5 amps no load and 10 amps with the 3/8" steal rod as a load. It appears to me 2,654 ohms of inductance is different than .4 ohms of resistance??? It seems logical to me if I make another choke that is 885 ohms circuit amps will be 30. It is not logical to me that 885 ohms of inductance is the same as .4 ohms of resistance? I should probably use actual battery voltage to do math instead of 12V because 13.2v = 33 amps in this circuit. Choke is 15 turns of wire, 15/3= 5 turns is it logical to assume 5 turns will drop 6.6uh to 2.2uh?

 

[alec_t]

Choke is 15 turns of wire, 15/3= 5 turns is it logical to assume 5 turns will drop 6.6uh to 2.2uh?

No. The inductance varies as the square of the number of turns.

 

[gary350]

I removed 8 turns from the choke. My LCR meter says 7 turns is 2.9mh. Choke connected to the circuit meter reads 1.5a no load, 15a loaded. It would be nice to know how to calculate the correct mh value to get 30a loaded. I should do calculations for 60KHz when circuit is loaded. Online calculator says 2.9mh at 60KHz = 1093 ohms. Is there a way to calculate target XL ohm value to get 30a? I can continue to do trial and error remove 1 turn each time sooner of later get it right.

 

[gary350]

Trial and error method does not work. 5 turn choke = 15a on meter. 4T = 15a on meter. 3T = 15a. 2T pegged meter past 50a blew both mosfets. I know people that design circuits don't do trial and error until they get it right. When I worked at Potter & Brumfield 1979 engineering dept gave me circuits to build then we tested them for months. Sometimes parts burned out they modified the circuit then I built it and tested it again. Circuits are not always 100% perfect on the first try. My choke design is way wrong 1 turn makes too much of a change in mh value. I have 6 or 7 toroids taped together & 15 turns of wire is all that will fit. I'm going to make a choke with 3 toroids & 15T it should be less sensitive per turn. Maybe I can find an online choke coil calculator.

 

[alec_t]

Theory and practice are likely to vary here, because online calculators (and typical text books) assume the choke is passing nice sinusoidal current. I suspect in your circuit the current is far from sinusoidal.

 

[gary350]

Theory and practice are likely to vary here, because online calculators (and typical text books) assume the choke is passing nice sinusoidal current. I suspect in your circuit the current is far from sinusoidal.

It is a whimpy mickey mouse circuit but it is also a good learning project. I finally got choke at 2mh with 3 toroids taped together and 10T of #14 copper insulated wire. I swapped out the 10T work coil for the original 8T work coil again. No load is 1.5a and 18a loaded. It works good now heats up a 1/4" steel rod red hot in 30 seconds on 13.2VDC.

 

[gophert]

Just based on the size, 15A will saturate that core unless it is iron powder construction. But ferrite, MPP, Permalloy, nickel or anything else will likely be magnetically saturated at 15 amps. Check the core manufacturer's datasheet.

 

[rjenkinsgb]

Those crude self-oscillating circuits are likely to be very inefficient as both transistors are turned on for a short time each half cycle.

I suspect the inductor is needed to allow commutation during the changeover time, when both transistors conduct; the short (=high frequency) spike sees a high impedance whilst the average supply current is maintained.

I'd think from looking at the schematic that using a proper drive circuit with a dead time during changeover would significantly improve output and decrease heat, plus no longer needing the choke?

 

[shortbus=]

Those crude self-oscillating circuits are likely to be very inefficient as both transistors are turned on for a short time each half cycle.

Been told so many times to use a real induction heater circuit and falls on deaf ears.

 

[gary350]

Been told so many times to use a real induction heater circuit and falls on deaf ears.

Where is a real induction heater circuit?

 

[shortbus=]

Go back to any of your induction heater posts and see what Maxheadroom linked for you. Which you completely ignore for your own reasons, reasons that aren't working for you so far.

 

[gary350]

Go back to any of your induction heater posts and see what Maxheadroom linked for you. Which you completely ignore for your own reasons, reasons that aren't working for you so far.

Is this what YOU want me to buy? I hope this makes YOU happy. I bought these long ago just not ready to build a "REAL" industion heater not until I learn more about parks that I have to build.

 

[gary350]

I finally have the math figured out. Using 12v a .4 ohm resistor limites current to 30a in this circuit. A 2uh choke with 3 toroid taped together is .75 XL ohms = 16a. If I can make a choke 1.07uh = XL of .403 ohms that should limit current to about 30a. I finally found a toroid choke calculator online it has 38 choices for toroid material I have no clue what my choke material is. My meter is having trouble ready a choke below 2uh so that is a problem too. Problem could be 1uh choke only has about 6 turns with 3 toroids.

I have a question about C1 will this stabilize the DC voltage on the mosfet gates?

maybe 1 less toroid and several more turns of wire meter can read it.