A feasible goal for a Miniature SSTC? / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

A feasible goal for a Miniature SSTC?

previous 1 2 3

Move Thread

LAN_403

Dr. Shark

Fri Feb 10 2006, 09:38AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Holy mother of Jesus, you actually managed to run IGBTs at above 1MHz? From what I hear so far, they max out at about 250KHz, so that is truely impressive. In fact this is exactly the coil I was planning to build, can you post some more information about it, e.g. what kind of IGBTs you were using, how you were driving them...?

Self Defenestrate

Fri Feb 10 2006, 02:27PM

Registered Member #87 Joined: Thu Feb 09 2006, 01:36PM
Location: San Jose
Posts: 191

That's incredible. Everything stays cool? Post a list of specs.

Marko

Fri Feb 10 2006, 03:06PM

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

When power is so low switching transitions make disipation but are not danger for igbt's life any more.
And I dont get proportions you say its just 1 inch ??? great..

JimG

Fri Feb 10 2006, 05:29PM

Registered Member #52 Joined: Thu Feb 09 2006, 04:22AM
Location: Austin TX
Posts: 57

The secondary is 1" x 3.5" using 36 gauge wire wound on an HDPE tube similar to the vitamin C tube. The top load is a 1" x 3" toroid.

The IBGT that I'm currently using is the IRG4BC40WS. I tried out a few different cores for the GDT to trigger them and found that a T50-2 gave me the best results at the frequencies that I was aiming for.

Unfortunately I haven't had a chance to run it for long periods of time since my mosfet drivers will stop working for after running for more than 20 seconds. I also wanted to run the whole thing off of a wall transformer so it only fires once a second after I've built up 175 volts.

Dr. Shark

Fri Feb 10 2006, 06:12PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

That IGBT has a turn-off time of 100ns which is really quite fast. Most lager die devices I have seen were above 200ns. Running at 1MHz this means that the switching transition lasts 10% of the cycle, which explains why it heats up quickly (and once hot the switching speed gets worse, for a runaway effect), but still sounds managable.
May I ask what kind of power supply you run it of, and what value your primary capacitance is?

JimG

Fri Feb 10 2006, 08:35PM

Registered Member #52 Joined: Thu Feb 09 2006, 04:22AM
Location: Austin TX
Posts: 57

At 1.1Mhz it's 454ns between zero crossings. At spec the total turn off and fall time of the IGBT is 174ns, my schmitt trigger has max propagation delay of 15ns and my mosfet driver has a max delay of 90ns. Looking at the numbers I should be spending 279ns transitioning and 175ns in the correct state. When I was putting it together I wasn't real sure if I would even get it to work at such a high frequency, but I've been able to get small visible streams at upto 1.8Mhz. I've only had it running for a few days now so I haven't had a chance to analyze if it is working because I'm using more than the specified gate voltage on the IGBTs or if it is because I'm just 180 degrees out of phase.

For the short bursts that I run it at the IGBTs don't seem to change temperature to the touch.

For now it is using a simple switch mode power supply using a 555 to charge a 250v 500uf capacitor. It's not ideal to wow people with, but I went this route because I wanted something I could experiment and take measurements with rather than generate continuous sparks.

Dr. Shark

Thu Feb 16 2006, 12:17PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I did some reading on IR's warp speed IGBTs now. To me this looks like the ticket to a new generation of miniature coils. Thanks Jim, for this great discovery!
With their 150kHz rated maximum frequency it seems we can easily drive them into the MHz, considering that most existing DRSSTCs drive IGBTs at about 10x the rated frequency. Turn off time is only 6ns, which is definitely MOSFET-territory, considering that even "ultra fast" IGBTs need some 30ns to switch off. Wow, this really does not fall short of the ideal switch very much. (according to the marketing guys here

that is)

But, where there is light, there is shadow: As these devices have not been around for long, its impossible to get them surplus, you have to buy them from IR, and they set you back $8 a piece (thats for the IRGP20B60PD which is the standard 600V 40A TO247 model. Datasheet:

Thats a lot considering that a few are certainly going to explode during the construction of the MHz driver circuit.

Well, the bottom line is, if anyone knows where to get these or equivalent devices for cheap, I'll be able to build my 1" secondary 5" streamer TC and be a really happy guy.

JimG

Thu Feb 16 2006, 04:02PM

Registered Member #52 Joined: Thu Feb 09 2006, 04:22AM
Location: Austin TX
Posts: 57

I got the IRG4BC40W IGBTs from Digi-Key for $5.83 each in the D2Pak package, but they only have the slightly more expensive lead free version, now. It looks like they still have a few of the lead versions in the TO-220 package left. I've been able to get them to peak at 200 amps without destroying them so if you're gentle you shouldn't have any problems.

previous 1 2 3

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