JimG's Smaller DRSSTC

JimG, Sun Feb 12 2006, 08:55AM

9/20/2006

I just got these in today.





------------------------------

9/10/2006

This weekend I built a dedicated optical interrupter for the second tiny DRSSTC. It uses a sliding variable resistor for an on time of 50-500us with a constant off time of 11ms.





------------------------------

7/11/2006

I built a ground target for the tiny DRSTTCs. I've been using it as a counterpoise so that I don't have to have a good ground to show off the coil.

The small size coil made it easy to take with me and show off during my July 4th vacation. It turned out to be very popular with children.





------------------------------

6/19/2006

I tried playing around with different break out points over the weekend and found that location and length drastically reduced the streamer length. With the reduced output the IGBTs would heat up very quickly, almost too hot to touch after only 3 or 4 seconds of run time.

Experiments with the interrupter timing have resulted in much thicker and brighter streamers. With the current timing I think I'm getting close to finishing up all of the tweaking I can do with this coil and will start cleaning up the wiring and build a dedicated external interrupter. If I find time I may rebuild the top load to something a little more presentable.

------------------------------

6/ 10/2006 Update:

I've replaced the metal supports with acrylic rods and have added a small toroid. The total height is currently at 7.25 inches. This finishes up almost all of the hardware portion of this coil.




Here are photos of the front and back side of the controller board. The lower half of the board has the controller and the top half is the h-bridge. The board size is 2 1/2 x 1 3/4 inches.




------------------------------

6/1/2006 Update:
I'm getting close to finishing up my mains powered tiny DRSSTC coil. I stole the primary design from Steve since I couldn't think of anything better. It currently puts out 2 inch streamers and at least one 4.5 inch arc to ground with my variac set to 80 volts. Anything higher and it arcs from my ground connecter at the bottom of the secondary to my metal supports that hold up the coil platform. I also have my optical detector connected to the 12v power supply so it takes a lot longer to turn off than I would like. Hopefully I will get all of these bugs worked out soon.

The calculated Fres was 950 khz, but the measured Fres is 750 khz. I'm happy with that since I was aiming for a frequency lower than 1Mhz. Hopefully when I get a more permanent topload it won't change the resonance too much.





------------------------------

4/24/2006 Update:

I didn't want to start a new thread since this just feels like I'm making small changes to my previous design.

I wound a new secondary to be smaller and allow for a lower resonant frequency. My previous secondary had a 3.5" winding height and used #36(0.127mm) wire. My new secondary has a winding height of 3.0" and uses #40(0.08mm) wire.

My previous circuit board was 3 1/8" x 1 3/4" with an on-board interrupter. My new one is a 2 1/8" x 1 3/4" on a double sided board, with an external interrupter. I also switched away from using a dual 4A UCC27425 mosfet driver to a 9A UCC27321/UCC27322 pair.

I've also switched my capacitor over from using a 500uf cap to a 340uf one. Instead of charging it through a switch mode supply I plan to have this powered directly from mains using some sort of currently limiting.

I've finished building the circuit and done continuity testing, I just want to hook it up to a signal generator to make sure that the singal is propagating correctly before I do any smoke testing.

Picture of the old board and secondary sitting next to the new board and secondary:




Schematic for the simplified board:





------------------------------

2/12/2006

Last October I was in a local surplus electronic parts store when I ran across a 1.375" x 3.5" 250V computer grade capacitor. Having just built two DRSSTCs I figured it might be cool to build a DRSSTC centered around this small capacitor. I then decided I would challenge myself with a few constraints.

1. The secondary shouldn't be any bigger than the capacitor.
2. The whole circuit should be kept small using SMT parts.
3. It should run off of a single power supply using a wall wart.
4. The interrupter should be on-board and easily updated.
5. It had to produce streamers, not just corona.
6. I wanted it to be a full bridge if possible.

On top of this I had never worked with SMT, I had never designed or etched a pcb, and I had never gotten either of my other two coils to work without flashing over with greater than 50 volts.

At home I found a small 1"x4" HDPE container similar to the vitamin C tube that I had seen others use and a small spool of 36 gauge wire. I figured that the unloaded coil would be about 2Mhz and loaded with run between 1.6Mhz and 1.8Mhz. At first I didn't think I would get anything to work based on the advice that I had read from a lot of people.

I chose to use a pic uc for my interrupter, direct feedback through a single schmitt trigger, a dual inverting/non-inverting mosfet driver and abstained from using any overcurrent, soft switching or dead-time circuity to keep everything very simple. I had my first circuit up and working toward the beginning of January using surface mount mosfets and a flat pancake primary with 6nf of primary capacitance. I experimented with a few GDT cores and found that a T50-2 gave me the best results.

This first test gave me very modest results with small 1/2" streamers. I figured I would need more current if I was going to go any bigger, but I had scoped the primary current at 40 amps and my mosfets were only rated for 14A. I almost gave up thinking that there was no way that I could get IGBTs to work at such a high frequency but decided to give it a shot anyway. To my amazement I could get a set of IGBTs to generate streamers just as long as the mosfets by tuning the primary circuit to a lower frequency.

At this point I figured it might be feasible to actually get this to produce some slightly larger streamers so I built a conical primary, added a toroid that I purchased from John Freau, increased the primary capacitance to 20nf, and set my on time to 75us. Once I had it together I set the interrupter to wait until the cap was up to about 175v before firing and applied power to the circuit. After about a second a bright white spark jumped one and a half inches from the break out point to a grounded target. I was very happy with the results and playing around with the tuning in that configuration I was able to get 4.5" ground strikes fairly easily and on occasion a 6" strike.

However, I still feel that I have a few things left to do.

* Right now the simple switch mode power supply that I've built can take a couple seconds to build up a sufficient charge for good sparks. I'm sure I can cut this down through some proper tuning.

* The primary current peaks out at about 200 amps. I think I can get some bigger sparks with shorter on times at these currents. Hopefully I can get shorter times between bursts if I can keep the main capacitor from fully discharging with each burst.

* I would like to experiment with different interrupter settings. I had pretty good success using a random number generator for random on and wait times on my larger DRSSTCs.

* The driver circuit will stop working after running for short periods and will start working a short time later. Everything is cool to the touch so I will need to investigate further into what is going wrong.

* Maybe I'll get a chance to revisit the circuit to add soft-switching.

My secondary:



On the left is the h-bridge and controller, on the right is my 12v and 200v power supply:



The final coil:



A 5" strike to a target 4 3/4" inches away: