What do you usually do when you're going to plug something in the wall? / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

What do you usually do when you're going to plug something in the wall?

previous 1 2

Move Thread

LAN_403

Marko

Sat Mar 19 2011, 08:58PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Some words of advice.

If you are using a ferrite transformer, I assume it's a large flyback core or something of similar size. If it's so, you are going to need more like 200+200 turns rather than 16+16. Why? You're now trying to put around 700V on your transformer which can practically handle only like 2-3V/turn without cooking up, and you certainly saturated it with only 16 turns. The other thing is, if you put the transformer in between of your tank cap and work coil like shown on schematic, it would also need to carry likely tens of amps through those 200 primary turns, because you are pushing all your tank circuit reactive power through it. Constructing such transformer on a flyback core just isn't feasible, and in my opinion, doesn't make sense at all. The transformer should go onto the output of your mosfet oscillator, and tank L and C should be robustly connected together, with the transformer feeding only real power to them.

That doesn't exclude the possibility that the circuit will still blow up after that, though. IGBT's have too large collector-emitter voltage drop and this + diode drop makes the diode feedback drive scheme very unstable. Diode recovery losses are also likely to be gigantic, UF4007's might overheat and blow in seconds if you pushed the frequency too high. At 1Mhz this tends to happen to me with 15V in with such circuits.

Hence if you still think you have time and patience for this, put an overcurrent protection on your circuit to turn it off if it latches up. It needs to act very fast though since IGBT's will go boom in microseconds with 320V supply.

If you really want a quick and dirty (and yet powerful) induction heater, just get a rewound mot and use a standard circuit with IRFP260's. Quite a few people got this to work already and you also get isolation from mains. I've newer seen anyone make an royer induction heater running straight from mains though.

Marko

Steve Conner

Sat Mar 19 2011, 10:23PM

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

Thank you Iamsmooth and Marko for the hints.

Personally, I put on safety goggles and maybe hide under the table with a fire extinguisher if it's something really big.

Mains power is a different ball game to transformers and batteries. Most wall outlets, even in 120v countries, can deliver over 100 amps short circuit current, and if your circuit isn't solid it may simply disappear in a blue flash, leaving nothing but a kind of shadow of vaporised copper on the table.

Inducktion

Sat Mar 19 2011, 10:31PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Marko wrote ...

Some words of advice.

If you are using a ferrite transformer, I assume it's a large flyback core or something of similar size. If it's so, you are going to need more like 200+200 turns rather than 16+16. Why? You're now trying to put around 700V on your transformer which can practically handle only like 2-3V/turn without cooking up, and you certainly saturated it with only 16 turns. The other thing is, if you put the transformer in between of your tank cap and work coil like shown on schematic, it would also need to carry likely tens of amps through those 200 primary turns, because you are pushing all your tank circuit reactive power through it. Constructing such transformer on a flyback core just isn't feasible, and in my opinion, doesn't make sense at all. The transformer should go onto the output of your mosfet oscillator, and tank L and C should be robustly connected together, with the transformer feeding only real power to them.

That doesn't exclude the possibility that the circuit will still blow up after that, though. IGBT's have too large collector-emitter voltage drop and this + diode drop makes the diode feedback drive scheme very unstable. Diode recovery losses are also likely to be gigantic, UF4007's might overheat and blow in seconds if you pushed the frequency too high. At 1Mhz this tends to happen to me with 15V in with such circuits.

Hence if you still think you have time and patience for this, put an overcurrent protection on your circuit to turn it off if it latches up. It needs to act very fast though since IGBT's will go boom in microseconds with 320V supply.

If you really want a quick and dirty (and yet powerful) induction heater, just get a rewound mot and use a standard circuit with IRFP260's. Quite a few people got this to work already and you also get isolation from mains. I've newer seen anyone make an royer induction heater running straight from mains though.

Marko

The only thing I've really been trying to achieve through all of this, is just the ability to melt metal. Iron, Aluminum, or steel were ones I had in mind. Copper would be cool too.

Marko

Sun Mar 20 2011, 01:20AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi inducktion,

There was a thread around called "OMG induction heater" which is probably the first time someone melted steel with homemade induction heater, known to this forum. Of course, the secret were celem conduction cooled caps. No matter what kind of driver you choose, it's very hard to get the required power level without those caps. You could parallel a very large amount of small FKP caps(like 100 10nf caps) with lots of forced air, but that might still end up costing as much as a conduction cooled cap + lots of work to solder everything together.

Think of it this way: You're trying to melt copper, which is like a single turn secondary coil inside your work coil. But your work coil is also made of copper, and despite the current in the coil is lower due to more turns, it's resistance is higher so I'd say that the heatage the coil receives is in the same ballpark as the heat received by copper workpiece. What normally stops the coil from melting at this point is water flowing through it.

And even the best caps around are still far more lossy than the solid copper conductor, probably by like a factor of 10. So if you put whatever power required to melt the copper workpiece, you might easily end with your cap having to dissipate 10x as much. In other words, most caps would melt far before the workpiece does.

With metals like iron situation is easier, but losses in the cap and work coil are still going to be very large, and construction of the tank circuit itself a rather big challenge alone. After it's done you could actually use various topologies to drive it, perhaps even with royer circuit too.

I recommended a MOT-powered royer circuit because it's proven to work and it seems to be within your budget. It might even melt small amounts of steel if the workpiece is well insulated. Here is a nice example of such one:

I would add more tank caps and water cooling to it though!

By the way, the kim ladha's "OMG induction heater" thread is here:

Marko

previous 1 2

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