Induction heater, what frequency to use for heating copper to brazing temperatures? / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Induction heater, what frequency to use for heating copper to brazing temperatures?

1 2 next

Move Thread

LAN_403

Henry

Tue Dec 01 2009, 10:15PM

Registered Member #2440 Joined: Sat Oct 17 2009, 05:49PM
Location:
Posts: 9

What would be a good frequency to use to be able to heat up a 10 mm copper tube to brazing temperatures, what i guess would be maybe around 800 centigrade or so?

Have wanted to do an induction heater quite some time now and finally got my thumb out and made one. Its a smaller induction heater based on the ZVS schematic that the first time i connected it worked just as it should and as a test i heated up at hacksaw blade to its curie temperature in about 2 - 3 seconds. Now, which initially was the reason for this heater, i want to heat up a copper tube so i can braze it since i dont have access to a oxy-acetylene torch and i have started to increase the frequency so it can induce eddycurrents in the tube. And sure the coppertube started to get slowly hotter the more capacitors i removed but also with some destroyed mosfets along the way but it has been my own fault.

Before i did this so did it draw 2 Amps at 45V at idle which increased to 11A just in the beginning of heating of iron, it did not change at all with copper. After i had increased the frequency twice as much so did the copper get quite hot after about 15 seconds but you could still touch it. I then wanted to try with iron to se if it would melt and 1 mm into the coil so was the current 13A so i thought it was ok. I then put it in 3 mm more but this time the current hit the roof with 118A which instant fried one of the mosfets and draw the zener and diod down the drain with it. Think i have to make a current limit for it all because it a little work involved to change a mosfet every time.

I have experiment some with capacitors and coils but anybody know what frequency i should aim at? So would it go faster to hit the right point. The transformer are good for about 20A+ continuously at about 40 - 45V so i have around 1KW+ to play with but dont know if i might need to have a higher voltage or so or this might be enough? I rather dont wan to connect me to the main. The mosfets are irfp260 and the coil right now is 7 turns bifilar winded that get hot as hell so i have to have a fan on it. Have seriously thoughts on water-cooling.

Tonskulus

Wed Dec 02 2009, 04:05PM

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

Copper requires much more workcoil current/reactive power because copper has no hysteresis losses while iron does have. So all we have here is eddy current losses to be play with. And copper is good conductor so we must induce very high eddy currents for copper workpiece to heat it up to a required brazing temperature.

We need specific amount of inverter power there, something like 600W or even more. Workcoil current should be hundreds of amperes (depending on turns and frequency, they all are depending on each other).

Henry

Fri Jan 22 2010, 02:25PM

Registered Member #2440 Joined: Sat Oct 17 2009, 05:49PM
Location:
Posts: 9

Yes eddy currents is the only way to go no doubt but as i see it so is the problem that copper cant conduct a magnetic field unless you go, as i guess, up quite high in frequency so it start to be magnetic conductive, just as iron but at much higher frequency. Only then can the power in the coil be transferred to the copper tube and heat it up, not otherwise. If i need to get more power into the coil so would that be equal to more capacitance but by doing that would lower the frequency and make the copper less magnetic conductive so that shouldnt work? So im a little lost right now..

The frequency is at about 130kHz and a copper tube would just get slightly warm after 20 seconds but not more then that. A hacksaw blade however will glow bright red after under 3 seconds so something is missing but i dont understand what.

Tonskulus

Fri Jan 22 2010, 02:38PM

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

Well I have melted coppersheets @ 5kHz. I used multi turn work coil and flux concentrator (one half of big E core ferrite).

So it is possible using low frequency for brazing purpose. There just has to be strong alternating magnetic flux density for inducing enough circulating current for copper work piece.

Also very high frequency is useful for copper brazing, like 1..3MHz. This high frequency will make considerable skin effect losses on surface of the workpiece. However, it will be difficult task making >3MHz high power mosfet inverter. Vacuum tube oscillator would be good idea there if anything.

Afterall conculsion is that almost any "high" frequency CAN BE used for copper brazing. The only thing we need is enough reactive work coil power and flux density.

IamSmooth

Fri Jan 22 2010, 02:57PM

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

I gotten thick and thin pieces of copper to melt and boil without a problem. My problem is containing the melt. I am going to try and make some crucibles using furnace cement. I am using about 80khz. I have fount that a tightly packed coil (little spacing between turns) and 1.5 to 2" ID worked well for me.

Henry

Fri Jan 22 2010, 05:23PM

Registered Member #2440 Joined: Sat Oct 17 2009, 05:49PM
Location:
Posts: 9

Had no idea that even a low frequency could melt copper no matter what, was absolutely sure that high frequency was the only way to go for that to make the copper magnetic so it would take the current from the coil. Cant really understand how a high current in the coil at a low frequency anyway could get into the copper without any flux concentrator or similar, dont get it the copper had to be magnetic?

As mention i use the ZVS schematic and the coil used is 1.8" ID and about 2" OD so the wire are quit thin its about 0.1" maybe i should use bigger tube? The coil i use are 7 turns bifilar winded but maybe 5 turns would be better and smaller ID? The wire, or actually tube to water cooling soon, are as it could and the coil is about 8ÂµH. The capacitor bank are right now 6 x 0.22Âµ capacitors with high frequency, high current and pulse capability and they wont get warm at all.

But maybe then put more power into the coil by increase the capacitance and make the frequency higher by making the coil little smaller and take of a rev or two to compensate the lower frequency due to the higher capacitance or this might not matter to heat the copper as far as the capacitance is high?

IamSmooth

Fri Jan 22 2010, 06:01PM

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

Give me a day or two, and I will put a video on youtube showing me taking some thick chunks and converting it into a bubbling red liquid. This will be my stationary coil, and not the levitation coil. I have been able to melt copper with 1/4" tubing, but the 3/8" works better. You need tubing so you can run a coolant like ice water through it. I tried turning up the power without cooling the coil and it started to smoke before I was able to accomplish much on the workpiece.

My coil is 3/8" tubing, about 1.75" ID, 4-6 turns. I will double-check the frequency. Input Power about 1.5-2.5kw. Resonance lock is with a PLL/integrator feedback.

I will put some pictures of the coil on there tonight or tomorrow.

Henry

Fri Jan 22 2010, 07:24PM

Registered Member #2440 Joined: Sat Oct 17 2009, 05:49PM
Location:
Posts: 9

That would be very interesting! Aha so you using that thick tubing good to know.

Im using the Mazzilli ZVP schematic due to its simlicity just so i dont have to build and test out a PLL control. And if its possible to get such high temperature anyway with the ZPV instead without having to crank up the powerlever insanly much so would that be just perfect.

Yes i had it going before without any cooling of the coil only a fan att full speed and that worked, but after i removed some capacitors so did the idle current go up to about double and then the fan wasnt enough and the coil started to smoke, so bigger tubes and watercooling is what im aiming for. I think i will go and buy coppertubes already tomorrow.

Tonskulus

Fri Jan 22 2010, 08:47PM

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

Basic mazzilli ZVS is way too low power for melting copper, there is just not enough tank voltage and thus, not enough work coil current. We want to see some hundreds of volts there (not just some tens of volts as mazzilli ZVS) to get reasonable heating results.

btw, it should be noted that heating iron above curie point (up to yellow-bright white glow), is much easier and requires less power than melting copper.

Flux concentrator is good idea for low freq. systems if smallish workpieces are being heated.

IamSmooth

Fri Jan 22 2010, 10:00PM

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

Tonskulus wrote ...

.btw, it should be noted that heating iron above curie point (up to yellow-bright white glow), is much easier and requires less power than melting copper.

You are right about that. The resistance of steel is higher that the RI^2 losses generate much more heat.

I am not familiar with the mazzilli, but the series tank I have has a Vcap voltage between 150-200v when I'm melting the copper. I've calculated the current to be in the hundreds of amps just like you mentioned.

Henry, you should check out the tutorial to see the circuit and tank I used, as you can do this off of MAINS voltage.

1 2 next

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