induction heating levetation / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

induction heating levetation

previous 1 2 3 4 next

Move Thread

LAN_403

IamSmooth

Mon Jan 18 2010, 04:44AM

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

you have got to see this!

Here I am levitating, melting and boiling aluminum like a mini sun, suspended in a magnetic containment field.

I start at 1.5kw and bring the power up to 2.5kw. I am still afraid to just turn it on full. The aluminum spins because it is uneven in the field. Eventually, it liquifies into a sphere, which I poke like a soft marshmallow. The magnetic field is strong enough so the molten mass does not slip through the center. It is probably not evident on the video, but I can see it boiling. Finally, I release it and it plops onto the floor and immediately solidifies.

The levitation is very stable and the mosfets were all cool to the touch. The PLL-integrator held onto the lock just perfectly.

I have also liquified copper and will get this on video next.

Dinges brings up the point about the series diode. With regards to the schottky series diode 1n5822 I looked at the datasheet and it has an average of 3.0A forward current. How can this work when the mosfet sees 20A or more through it? Steve or Richie?

Experimentonomen

Mon Jan 18 2010, 08:39AM

Registered Member #941 Joined: Sun Aug 05 2007, 10:09AM
Location: in a swedish junk pile
Posts: 497

Nice!!

GeordieBoy

Mon Jan 18 2010, 11:40AM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

My mosfets blew because I was out of resonance. The piece of steel shorted some of the coils, lowering the inductance. This put me in a zone where capacitative reactance was greater than inductive, and this is bad for the mosfets

The series drain schottky diodes are to prevent this type of failure. Otherwise you get forced reverse recovery of the MOSFET body-drain diode whilst driving a net capacitive load.

Also note that one of the advantages of the parallel resonant LCLR network is that it is inherently fault tolerant of the work coil getting short circuited. If the work coil gets short circuited by the workpiece (or the tank cap should go s/c) then the inverter just sees the impedance matching inductor as its load. The inverter load current falls to a small inductive (lagging) current which is safe for the inverter until it is shut down. (Contrast this to short-circuiting the work-coil in the series resonant topology where the tank capacitor gets discharged into the output of the inverter. The on-going load is then capacitive for a complete or partial s/c of the work-coil)

-Richie,

Steve Conner

Mon Jan 18 2010, 11:51AM

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

Yes very nice!

IamSmooth wrote ...

Dinges brings up the point about the series diode. With regards to the schottky series diode 1n5822 I looked at the datasheet and it has an average of 3.0A forward current. How can this work when the mosfet sees 20A or more through it? Steve or Richie?

Well, a 1N5822 is obviously not a good choice for the schottky! Other bigger schottkies are available.

uzzors2k

Mon Jan 18 2010, 01:18PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

IamSmooth wrote ...

That's really cool, isn't it?

Hell yeah! Excellent demonstration Jonathan.

IamSmooth

Mon Jan 18 2010, 04:43PM

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

Steve McConner wrote ...

Well, a 1N5822 is obviously not a good choice for the schottky! Other bigger schottkies are available.

I have found a few schottky choices. They are all fast. The parameters I am toying with is the reverse voltage. How much is enough? If I have 340v, each mosfets sees half. So would this be overkill?

With regards to the free-wheel diode, does the reverse voltage have to equal the drain-source potential? If that is the case, I could have 170vdc.
How is this?

Also, what is so bad about capacitive loads? Why does this type result in failure compared with inductive?

Steve Conner

Mon Jan 18 2010, 04:53PM

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

The reverse voltage of the Schottky just needs to be a few volts. It only ever sees the Vf of the main freewheel diode, give or take some transient voltages. In practice 20 or 35V would probably do.

The freewheel diodes need the same voltage rating as your MOSFETs: 500 or 600V.

GeordieBoy

Mon Jan 18 2010, 04:56PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

> I have found a few schottky choices. They are all fast. The parameters I am toying with is the reverse voltage. How much is enough?

For Schottky diodes blocking the body diode of a MOSFET, 30 or 40 volts PIV should be plenty as long as you provide an alternative path for the free-wheel current.

> If I have 340v, each mosfets sees half. So would this be overkill?

Each MOSFET does not see half ! They both see the full 340v plus any switching spikes and ringing

However, the off-state voltage of the MOSFETs is irrelevant for rating the Schottky diodes. If you are thinking of using those diodes for free-wheel duty their 200V rating will not be sufficient. You'd be better with a 500V or 600V Ultra Fast recovery Silicon diode.

> With regards to the free-wheel diode, does the reverse voltage have to equal the drain-source potential?

Yes, ideally it would equal the MOSFETs Vds(max)

> If that is the case, I could have 170vdc.

Not if the DC bus voltage is 340V, you can't!

> Also, what is so bad about capacitive loads? Why does this type result in failure compared with inductive?

As I said in my last post. A net capacitive load on the inverter causes the free-wheel diodes to be subject to forced reverse-recovery. The load current changes direction before the applied voltage (capacitive load gives leading PF) so the route that the load current takes through the inverter changes before the MOSFETs are told to switch. When the MOSFETs eventually are told to switch, the current is flowing in the "wrong" free-wheel diode so these diode turns off rapidly and "under duress" If you use ultra fast recovery silicon diodes as free-wheel diodes and a tight layout to minimise stray inductance this is not too much of a problem, but if you try to use the MOSFETs body-drain diodes as free-wheel diodes with a net capacitive load or if the layout is sloppy then you are in for a miserable time.

-Richie,

IamSmooth

Mon Jan 18 2010, 05:36PM

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

I feel dumb. I understand now: capacitive loading just refers to the direction the current is going through the mosfet. It is reversed at this time.

I have been looking for some parts for the inverter, but have been having trouble finding some in the 30-40A range. I found a 40A schottky that seems good, but it is surface mount.

This isn't a big problem, as I have the heat station for it. I want to have some headroom above 30A. Do you know of any through-hole style for +40A schottkys?

I guess the MUR1560s are a good choice using a pair for the free-wheeling diodes.

Nicko

Mon Jan 18 2010, 05:59PM

Registered Member #1334 Joined: Tue Feb 19 2008, 04:37PM
Location: Nr. London, UK
Posts: 615

This is what annoys me - In the UK (e.g. from Farnell) the MBRB4030G is Â£4.14 or USD 6.77. RS don't even stock it.

From Digikey in the US, its USD 3.70.

Both in Qty 1 pricing.

Pah.

previous 1 2 3 4 next

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