A Kapanadze generator replication - Febuary 2013

[TinselKoala]

Me too.

It is possible that you are underestimating the importance of a properly functioning spark gap. Also, for best power throughput you will want relatively loose coupling between your main primary and your secondary. Many many coats of insulating varnish (Krylon) on your secondary will help keep the voltage where it belongs. The secondary must be grounded to the Earth, either at the bottom for the standard Tesla coil configuration or at the center tap for the Odin-style double coil. The instantaneous currents in the primary can be very  large, kiloamperes even, which is why you need heavy conductors and good contacts for your primary. 1/4 inch diameter soft copper tubing is good, copper ribbon (reduces self-inductance) is even better. Thin wires like #12 solid copper can be used for a primary but won't work as well.
Do not try to rectify the output of the secondary unless you have a rectifier that can withstand at least a megavolt... which is unlikely. If you rectify the _input_ to your primary coil-capacitor tank, then you are no longer dealing with a system that requires the capacitor to charge fully during one AC input cycle and you can use a larger capacitor, hence requiring fewer turns for your primary.
For optimum performance you will want a spark gap that quenches (turns _off_) as rapidly as possible. If you are using a rotary spark gap with AC supply to your primary tank, you will encounter "picket-fencing" that will produce no output at all at certain vaues of BPS. Rectifying the tank input solves this problem and allows more freedom in gap rotation BPS. If you have a high-current supply then your sparks will "trail" in the gap and won't quench properly, vastly reducing the output of the secondary, so this means your electrode spacing and paths have to be constructed with prevention of arc-trailing in mind.
300-400 kHz is already getting into the "high end" of frequency for a spark-gap coil. As you go up in frequency the characteristics of the secondary emissions change and tuning becomes more critical.
You don't have to "spark that fast"; the secondary should make many cycles at its resonant frequency during each spark interval of the primary tank circuit. Ideally, with AC supply to the tank, you would like the capacitor to charge fully during one cycle of the mains supply frequency and have the spark gap fire just at the peak of the charging voltage from your primary transformer: once every full (or half) cycle of the mains frequency, 50 or 60 Hz.

http://www.youtube.com/watch?v=dTaIj5qLoH0
http://www.youtube.com/watch?v=tTLFlRhsa5U
http://www.youtube.com/watch?v=yjc9ilOAaQU

Of course, if you are trying for something other than extreme voltage and power throughput, you can ignore all of the above.

[TinselKoala]

The two spark shots above were taken at 1/15 second exposure. Since the DC coming out of the FWB is unfiltered, there is a strong 120 Hz ripple on this DC -- in fact, a 0 V to 3 kV ripple, about. In 1/15 of a second, there is time for 8 of the ripple peaks entering the tank to charge and spark, charge and spark.... and you can see very clearly the 8 sub-streamers in the psarks. Each of those substreamers is itself made up of substreamers generated at the coil's resonant frequency, far too fast for the camera to capture.

[DreamThinkBuild]

Hi TinselKoala,

Nice build, that is pretty cool. 

[Farmhand]

Good tips Tinsel, Can you give any tips on the camera settings, ISO, aperture, ambient light ect. I see similar arcs with multiple ripples but I find it difficult to capture them well by pictures, I need to use longer exposure times, I have only a (Fuji FinePix S5700) camera.

Anyway I was messing with a couple of other coils I had made for LV experiments and got two incandescent bulbs to light partially. The interesting thing I noticed was the different look to the light or the way the globes worked. They seemed to fill the entire inside of the bulb with an even light, the lower voltage one seems to pour the light out almost like onto the surface of the wood it was on, the filament kind of disappears in a ball of light, the 240 volt bulb only get's yellow but it is kind of the same.
Long way to go yet but I got some filaments to glow at least already. 

There's a blurry picture showing the odd way the lights light up attached.

This is the spark gap coil, it's pretty heavy duty, 2 x 6 mm primary turn 370 diameter and the rest is one mm wire I think from memory the secondary has 40 turns and the extra coil has about 170 maybe a few more, can't recall. Supply is two anti parallel MOT's with a DC resonant charging circuit, 12 nF primary cap.
http://i227.photobucket.com/albums/dd168/Toey1/TClights2009.jpg

Here's a video clip showing how it's setup ect.
http://www.youtube.com/watch?v=6Y1U1PSmAjQ

Please forgive the messy area and bad camera work.

[Lakes]

The two spark shots above were taken at 1/15 second exposure. Since the DC coming out of the FWB is unfiltered, there is a strong 120 Hz ripple on this DC -- in fact, a 0 V to 3 kV ripple, about. In 1/15 of a second, there is time for 8 of the ripple peaks entering the tank to charge and spark, charge and spark.... and you can see very clearly the 8 sub-streamers in the psarks. Each of those substreamers is itself made up of substreamers generated at the coil's resonant frequency, far too fast for the camera to capture.

Wow, that "ring of fire" pic is fantastic!