Tesla Coil Build Thread

[Fiddy]

wound my secondary and polished my top load

tried to wind 40AWG but it was a C&%T! went for 32AWG instead, this is going to raise the resonant frequency a lot so im probz gunna add a extra bigger top load.

 

[Sigurthr]

Ooh, nice!

Even 32ga is damn fine wire. I use 26-30ga only for that reason.

Spent the day testing my old 666KHz secondary; it is toast. I can't remember what killed it, but it is dead. It passes DC continuity tests but presents a VERY high impedance to AC. I wish I had a function gen and LCR meter.

 

[Fiddy]

32awg is soo easy can do 200mm of it 20mins. 40awg is like laying hair its bullshit small.

Haha thats really weird what have you done to it?

 

[Sigurthr]

I can't remember, tbh, haha.

It still resonates, but at 178V input it only draws 20mA regardless of coupling. I remember it did this when I had it on an old solid state driver I was testing a year ago. Totally forgot about it, heh. I may have pushed 4kW through it and caused its death but I have no idea the failure mode. To go from low impedance to 9000ohms reactance... Yeah idk. It's only 55ohms DC resistance. It's like I turned 90% of the inductance into capacitance some how.

I'm scrapping it. I have three brand new ones on the way; 500KHz, 750KHz, 1MHz. The one megger is for my own experiments though. I'd like to try a class-DE half bridge.

 

[Fiddy]

Haha good stuff. Yay experiments :-D

 

[Sigurthr]

I also would like to try a center-fed push/pull topology. While it is easy to implement (due to no need for a GDT or isolated drive), no one has successfully done it and reported it (at least not at the usual places, and no google search results turn up). The issue is that one leg of the primary will have far less coupling than the other half, resulting in a load imbalance between the push and pull. Due to the coupling imbalance the coil will alternately resonate at two different frequencies. When a primary couples to a secondary the effective capacitance seen by the secondary increases proportional to the coupling factor. Getting a center-fed push/pull primary to oscillate at the same frequency and have a balanced load sharing will be a lot of fun to try. I'd like to use a monolithic oscillator IC directly feeding a UCC gate drive pair. Zero feedback. The entire electronics package would be two filter caps, four ICs, four decoupling caps, two FETs, and two resistors. I'd like to experiment with external ballasting and battery powered operation if possible (and if needed as for the ballasting).

 

[The Lightning Stalker]

What about an oudin coil? Wouldn't that balance the primaries, or is an oudin somehow not possible with
a SSTC?

 

[Sigurthr]

Oudin coils function by mechanically interrupting current through an inductor to generate HV via back EMF. No transistors can hold up to the amplitude of the back EMF generated by a traditional oudin coil, so a solid state version isn't exactly viable. You could change coil parameters to lower the back emf below 1200v (maximum transistor blocking voltage), but there wouldn't be much use to that. You could try driving a SGTC with the modified low voltage oudin, but losses adding would again reduce viability.

I remember reading about a design where isolated 900V DIAC modules were synchronously switched in series to effectively replace the SG in a traditional SGTC type design. It was an incredible work of engineering. Something like that could work for a solid state induction coil.

Anyway, it's all tangent to the original topic.

As far as push/pull goes, it works flawlessly and is straightforward to implement on a bipolar TC! If only I had a way to make a proper mounting support for a bipolar coil I'd whip one up in a heartbeat. Bipolar coils need no grounding and can utilize super high coupling factors. They make excellent solid state coils. I saw a bipolar DRSSTC a few years ago on a teslathon website.

 

[The Lightning Stalker]

Oh, well, when I said oudin, I meant bipolar. I guess you received my mental image even though there
was a difference in terminology.

 

[Speedy78]

Rebuilt the MMC I got from Sig today. I needed a smaller MMC to fit in the chassis I'm going to build.

Started out with three rectangles of plexi.


Cut up and drilled for the leads of each cap.


8 caps in series per rectangle.


All rectangles in parallel.


A bolt acts as my terminal.


Wanted a brief demonstration of the dB level of a SG. All 8 gauge car audio wire, can of green beans for a dummy load. 14 turns on the primary.




This night made possibly by Sigurthr and Jack Daniel's.

 

[Sigurthr]

This night made possibly by Sigurthr and Jack Daniel's.

Haha, an excellent combination, if I may say so myself. Though, these days I'm more of a Jim Beam kind of guy. <3 Devil's Cut.

Today I did the diagnostics on my test bridges for the new boards (once they come in). All tests Pass.

I also reworked the very first half-bridge I ever made, it is the "On PCB" style one. It's up to snuff now. Though, the core I used for its GDT is the old N87 stuff I used to get before I switched to the T38 stuff I use now. The old stuff doesn't like <100KHz or >1MHz signals, but works fine for normal SSTC/DRSSTC frequencies.

If I get a chance in the coming days I'll do photoshoots and write-ups for both bridge construction methods. Both of these Half-Bridges may be available for sale at some point, we'll have to see.

I think I may need to expand my dedicated grounding network. I have the materials to do so, but not the time or drive to do it, heh. Plus, I'm hesitant to cut up my $1/foot low loss HF coax. But, it really does need to be done, especially if I get that 1MHz coil running well. 1kW @ 1MHz is a LOT of RF.

Also, I'm going to start work on a microcontroller based interrupter. Much of the code will be based on the work by Gao Guangyan a.k.a. "Loneoceans". I'll be using the AtTiny85 based Digispark as a platform because I F^%$#NG HATE ARDUINO. Gao didn't work with the digispark though, just the ATTiny85 itself, and while I'd do just that, it means having to buy a $50 dev board. Fk. That. The Digispark is an ATTiny85 bundled onto a breakout board, Vreg, USB interface, and Bootloader for $9. It should all come together rather simply and easily. The hardware should be as simple as adding two 10k pots for PRF and PW, a 9V battery, and one of the Fiber Optic transmitters I've been using for years.

update: 12/29/13 I finished writing the code for the MCU based interrupter, and my Digispark has been shipped. Now just have to wait.

 

[Fiddy]

Tested my SGTC today, first without a secondary then with a really bad setup.

Wrong Secondary too little Inductance
Wrong sized Top Load(s)
Unknown Primary Inductance

heres a lil video:

Getting 50mm arc's, im surprised it did anything!

 

[Sigurthr]

Shorten those primary leads to under one topload diameter and you'll see major improvement. Also, SG coils really don't like helical primaries that aren't quarter turn tapped.

Nice first light though!

 

[Fiddy]

Yeah shes a bit long will be short once the roof goes on top where the resonator will live.

Ive got no problem tapping a helical primary at quarter turns, silicon wire loves doing that.

Its a horrible first light, when i get the proper tuns on the primary and proper toroids it should be awesome.

 

[Speedy78]

Winding jig complete. Test run was the ity bity coil sitting on it. Decided to make a toroid for it. Now I really want it to work! Resonant 1100Khz.

Coil for for the SGTC is 4.25". Received it in the mail from Sig as a packing container for the MMC and ran some numbers with it and it will work out for me. (plus the wife said I'm not aloud to spend money until the baby is here) Good thing my 28 AWG is already in the mail!



Edit: Built a small cap from aluminum foil and used the clear top from a binder as dielectric.

For sure the smallest SGTC I have ever seen though I'm sure one of you will link something smaller and more successful.



Video showing it working. Weak but still neat. Water is there because I was 100% sure I would be putting out a fire. End of the video shows the now dead cap I guess the dialectic couldn't hold up because there is an internal short. When I showed the wife I had a light bulb a few inches away that lit up and I could draw about 2 inch arcs from the breakout.

Anyone have any recommendations for making a new cap that can withstand out of household materials?

 

[Fiddy]

Winding jig complete. Test run was the ity bity coil sitting on it. Decided to make a toroid for it. Now I really want it to work! Resonant 1100Khz.

Coil for for the SGTC is 4.25". Received it in the mail from Sig as a packing container for the MMC and ran some numbers with it and it will work out for me. (plus the wife said I'm not aloud to spend money until the baby is here) Good thing my 28 AWG is already in the mail!



Edit: Built a small cap from aluminum foil and used the clear top from a binder as dielectric.

For sure the smallest SGTC I have ever seen though I'm sure one of you will link something smaller and more successful.



Video showing it working. Weak but still neat. Water is there because I was 100% sure I would be putting out a fire. End of the video shows the now dead cap I guess the dialectic couldn't hold up because there is an internal short. When I showed the wife I had a light bulb a few inches away that lit up and I could draw about 2 inch arcs from the breakout.

Anyone have any recommendations for making a new cap that can withstand out of household materials?

Great Job!

The mini coil is cute i see people making capacitors from beer bottles, aluminium foil and salt water, maybe should look at that, you need to get your capacitance up to lower the resonant frequency for bigger arcs.