Small quick Class-E sstc. / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Small quick Class-E sstc.

Move Thread

LAN_403

Arcstarter

Fri Mar 12 2010, 08:51PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Having a few parts sitting around, and always having wanted to try class E switching, i decided upon a small sstc. I wound a 2.25 by 9.5 inch 600KHz secondary that i wanted to try out anyway.

So i just grabbed a 1.5 inch prototyping board and through some stuff together. I made an interrupter with a 555. It uses a pot to vary the current from discharge to the timing cap to change on time, and for off time control i used the output of the 555 to recharge the timing cap through a pot. Simple circuit.

I used a random mosfet driving IC i had sitting around. It worked, but the IC died so i switched to a UCC37321. It gets the driving signal from the antenna, blah blah blah. Like a simplified version of the mini sstc cut in half. Probably somewhat like Steve Ward's Class E sstc. Using a 16 amp 400v mosfet as the switching device.

It runs well with very little heat. I can run 40v input with no heatsink and 40-50% duty cycle. I am pretty surprised how little heat there is since it is a 2.6nf gate device, and it is running more than 600KHz.

Here is a video, and some pictures of the driver and such, and the spark pic is with a faster interrupter than on the video. All with 40v input.

This is under-damped, causing alot of losses in the anti-parallel diode. This is what it looked like in the video. Do not mind the ringing on the gate, that was due to the ghetto coax cable i was using for scoping the gate. This is with streamer loading. Excuse the bad scope shot, but i was running the interrupter, so it is not a sharp image due to the touchy triggering.

This is the waveform now. As you can see, the voltage drops to zero right as the mosfet turns on. So hopefully, i have near zero volts stored in the miller capacitance during turn-on.

ragnar

Fri Mar 12 2010, 09:11PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Cool, Arcstarter. Put it in a box!

Corey

Fri Mar 12 2010, 09:30PM

Registered Member #1902 Joined: Fri Jan 02 2009, 07:59PM
Location: Lancaster, NY
Posts: 75

Very good Matt!! I too am working on a class E coil, looks nice so far. Wish me luck as this will be my first sstc. It will will under the tesla coil thread and not the projects thread. Good Luck!!

-Corey

Arcstarter

Fri Mar 12 2010, 10:17PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Thanks for the kind works

. Some sort of enclosure is definitely needed.

Here is an 80 volt run. Wait till the end, i run with no breakout point.

Oh, and good luck on your sstc, too, Corey!

Update: well, the max voltage across the D-S on my mosfet is 400 volts, at this moment. That is the very limit of this fet! I wanted to go to 120v input. I need to find a higher voltage mosfet! 500 volts is cuttin' it close. I do have a 900 volt mosfet from an ATX PSU, but it's on resistance will be high. It is rated for 7.1 amps, and 1.1 ohms on!. Shoot.

StevenCaton

Sat Mar 13 2010, 04:31AM

Registered Member #1845 Joined: Fri Dec 05 2008, 05:38AM
Location: California
Posts: 211

Just watched your video. Looks great Matt.

Anyways, I actually just purchased the class E kit from EastVoltResearch. I can't wait to get to it.

Its the one that plays music.

Steve Conner

Sat Mar 13 2010, 10:28AM

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

Oh, well done Arcs! Your waveforms look Class-E-tastic.

Arcstarter

Sun Mar 14 2010, 10:43PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Thanks for the kind words guys!

A weird problem i am having... While messing with the coil, i set the interrupter to almost 100% on-time. The mosfet was not getting very hot, so i wanted to go for complete CW. I took the 555 out, and tied the enable pin on the UCC to Vcc. Three different times, this caused the UCC to explode, and usually the mosfet gets stuck on and silently, or explosively shorts. I am not sure why this happens? The voltage on the enable pin is 12v from the 555, and even when only interrupted twice per second and a very low off time the UCC runs cold. I am running short on ucc's and i have one 16a 400v mosfet left... So this time i do not want to screw it up

.

Again, with about 80v input (that is the AC rating, by the way, not DC, which would be around 113 volts) i am seeing 400v peaks from drain to source. That is about right, according to Richie's awesome website 3.3 times the input voltage is correct, i am seeing 3.5 times. But would it be safe to go for a 500v irfp460 and 120vac input? I am not sure how much the mosfet would like that, if the voltage is just a tiny bit higher it would avalanche. I have never seen a mosfet avalanche, so i am not sure what to expect. And if the voltage was 20v more than the non-negotiable 500v, it might avalanche every cycle and burn that 20v up. That is no good!

Heh, i will just try it. Not needing a power transformer would be great, then i could box it up with a 15va transformer for the driver and all the other parts, and mount the secondary and primary on top. That would be too neat...

EDIT: If the rise from the supply voltage was the same, i would be seeing more than 595 volts! Gah!

GeordieBoy

Mon Mar 15 2010, 02:55PM

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

Maybe the enable pin on the UCC driver chip you are using is only spec'ed for a high logic level of something like 5 volts? or maybe you are taking it to the Vcc line that is driving the Class E PA at something like 100V by mistake ?

Peak Vdrain for a MOSFET operating in Class-E is around three times the DC supply voltage. The exact peak drain voltage varies slightly depending on the Q of the load and particularly the duty ratio of the drive to the gate of the MOSFET. The longer the MOSFET is switched on the bigger the inductive kick is from the resonant tank when the switch turns off, and the higher the drain voltage swings during turn-off.

If you can easily control the duty-ratio of the drive to the gate of the MOSFET you might be able to eek a little extra performance out of a Class-E PA by messing about with the gate-drive duty ratio. Just make sure you don't make the peak drain voltage too high and avalanche the device continuously.

Most HV MOSFETs will avalanche about 20 - 30 volts higher than the datasheet absolute maximum Vds rating at 20'C. This value may be higher for a hot device. You will easily see if the device is avalanching because the trademark smooth Class-E waveform will get the top clipped off abruptly. Dissipation is high when this happens because the device sees simultaneous high applied voltage and current!

The avalanche voltage from drain-source breakdown has a positive temperature coefficient so you might actually be able to increase the supply voltage slightly when the device has warmed up! ...but do so at your own risk!

Another thing to bare in mind is that although the MOSFET may not repeatedly avalanche in a smoothly running HFSSTC it may avalanche momentarily if Vcc is suddenly applied, or if you use an interrupter to chop the drive or supply on and off.

-Richie,

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