Also if I used the internal primary. What pins are the feedback? 9-7 or 9-4?

 

I missed your question regarding the caps earlier, my apologies. What does their capacitance measure? I would start with one though. The driver circuit you have listed is actually quite good, however if I'm not mistaken the fast diodes may be rather important for it to work. That circuit will give just as much output voltage if not more than the ballast. However for it to work at all you cannot use the internal primary, as it is designed for a modestly high voltage and has far too many turns. With transformers it's about turns ratio, with the internal primary having several hundred turns, and the output having several thousand. This gives a ratio of maybe 1:1,000 or so. By winding a primary of only 6 turns as in the schematic, it gives you a ratio closer to 1:10,000 or so. Allowing you to get an equal output voltage with lower input voltage and higher primary current.

 

Yes I did wind my own center tapped primary. I was asking that for a 2n3055 based circuit that uses a 3 turn feedback and a 6 turn primary only using two resistors and a transistor. I mostly see self wound coils for that as well but I recently saw a guy on youtube do that circuit using the internal primary and feedback so I was just wondering.

Success!!! I just got a super low power 9W CFL and tried that. it was very very week but it did arc, not like i expected either, It was more like a tesla output,very stringy and harry outputs unless i got super close, like 3-4mm. and only worked for like 2 mins. So Im done worrying, Im ganna try the big ballasts today. Ill just add to 20kv diodes in series on anode to help the internal diodes dissipate there voltage.

 

I'm glad you got it working. I've got a few broken 12v ignition modules (the driver always fails, leaving a perfectly good flyback in there), however they're potted. If I can get the flybacks out in one piece I'll send a couple your way, along with the fully functional ones I mentioned earlier. I currently only have 2 working.

 

Sweet man. Okay I'm test this TC tomorrow. I tried the big ballast. It works really good. Very strong arc. WOW. 6-9cm... This Ballast is for two bulbs I have two blues and one yellow wire connected. It's rated for 1amp not sure if it's half that since I'm not using the other yellow & 2 reds? Can I use 2reds and 2blues on same pin and do two yellows on other? Do you know why there is two blues if attached to same pin? couldn't be for current since only one yellow is on the other side.

If I use this for my SGTC, should I use a choke? I don't wanna use an inductor quite yet since it doubles your voltage. So if I need one, what value should I use? I have alot of high wattage ones. But I'm not sure if it's needed on this setup. If it won't hurt I'd like too..

Now on gap. I was ganna try static at first but I now see on a quick SG test with this how you need quinching. My 6 watt one had no problem breaking arc. But I only fliped the switch for a sec on this ballast one with 2 6.8nf caps in series. So that was 3.4nf and during that quick on/off I heard the gaps freq speed up rapidly to the point where my secondary arced to my primary so Im working on insulation now but why is it rapidly upping in freq? Is that normal? Maybe since I had no choke? And then I don't have a quiche setup yet so should I make a quick rotory gap? What would be a good BPS?

Ballast ---- GE240RS120-DIYB
Sec- 2.375 inch 648turns of .020 enalmeld primary- .24 copper pipe cone 8turns tunable clamps
5- 40kv 6.8nf caps

 

As for your ballast the two reds, two blues, and two yellows all have a low voltage with respect to each other (red-red, blue-blue, yellow-yellow). Each color set has a high voltage with respect to another color (red-blue, red-yellow, yellow-blue). The low voltage is for preheating filaments in a fluorescent lamp, the high voltage lights the lamp. The yellow set is meant to be parallel connected to the pins on one end of both bulbs. Since you don't have any filaments to preheat, you can cap off one of each color. You now have a red, a blue, and a yellow. I would recommend going to Walmart and getting one of their cheap analog multimeters just in case it breaks. The reason for analog being that the high frequency of electronic ballasts may give erroneous readings on digital. Test between each (red-blue, red-yellow, blue-yellow) and find the pair with the highest voltage and use that, capping whichever you have left.

You should not need an inductor for this, since the supplies you are working with were designed to be short circuited by an arc. My knowledge of inductors and their behavior is limited to dc and mains frequency, thus I can't say with confidence what would happen at high frequencies like these. It may behave the same, or it may break things. I'd suggest not using one since you don't need it.

As for arc quenching, one method is to direct strong airflow through the gap to literally blow out the arc. If constructing a rotary gap, try for a few hundred (300-600) bps. I'm not sure why the frequency would rise like that, but my best guess would be that the ionized air is not gone by the next pulse, and so there is an already conductive path for it to go through. Try using an air compressor or small inflation blower (such as for air beds) to quench your static gap.

 

Hey I haven't read ur last post so I'll respond to that in a sec. But I got a 400kv boost converter on Amazon. Can I make a mini sptc with this without diodes some how? I definitely don't have 400kv worth of diodes. Or if the they even make it. So a way without diodes, possible?

 

You could yes, just skip the diodes, it's a pulsed dc output anyway. Do note however that I have several of these and there duty cycle is very short, about 15 seconds max, then let cool down 2 minutes or more.

 

Okay Sweetness... Do you hook straight up to SG? Or do you use caps then SG just like I'm doing on the big one? And for mini saying 400kv what the hell caps should I use? Do I really need to roll new ones with 9sheets between each foil layer?? That seams excessive lol. Also can you walk me through how to calculate tank capacitance on this mini one. Rating is 3-6V 2-5A@400kv. Say I use a 7.4v lipo with a 5amp 5v linear regular.

Good thought for testing. But ya I have a really nice fluke rms meter with a 6kv x1000 probe I have never got wrong readings on this. Only thing that's about this accurate surprisingly is my DS202. But I only have a x10 for that. I also can't remember is its max 40v(80vpp) or if it's 40vpp so *10 is either 800 or 400 but big difference.

So besides inductors you think no resistance for a choke is needed either?

With bps I have a pwm motor driver on my Arduino. Will it be noticable enough to just adjust rpm so see what's better or should I make an optocoupler feedback loop for an exact rpm? Is doing all the math make things easier than just seeing what works better?

 

I would just hook it straight to spark gap and see what happens. The way it acts makes me think it has its own capacitance. With that module, 400kV is quite frankly bs. It's about 100-150kV before the insulation fails. It's a very slow pulse at those voltages, and it gets faster as the arc distance (and voltage) becomes less. I'm not really sure how to calculate tank capacitance with that boost converter as a power supply. Many of the parameters used to calculate (watts, bps, and voltage) are relatively unknown. The only one that is somewhat know is watts on the input side (5V x 4A = 20W). BPS will be all over the place with a static gap, bps is TBD, and voltage is all over the map with that thing. I don't even know if that can charge a cap, that might break it. A choke is primarily used to smooth out the charging of the tank cap so as not to stress the diodes in a normal supply with excessive peak current. Since all of your supplies so far are meant to deal with short circuiting (arcing), you really dont need to reduce peak currents. The choke also can double the input voltage. Not a good idea when the input is 150kV.

Yeah, you should be just fine with the fluke. Though when measuring ballast voltages (max 800V) I would just use the normal probe.

When adjusting rpm, it will be noticable. Though knowing the exact bps is very handy, so if you can build in a tach it should prove worthwhile. If you were working with well regulated and smoothed DC supplies, then I would say the math is worthwhile. Though if you want to plug in some voltages and practice with those equations, I'd be happy to walk you through it. As for functionality of your setup, I'd say see what works.

 

Okay well is the reason for not using an inductor in this situation is because of why it helps- how it resists high frequency when tank cap oscillates at resonant frequency, not allowing that high frequency to back track. But with a flybacks and the high frequency it puts out would be resistive to flow through the the inductor in the fist place right?. But theoretically wouldn't a resistive setup with filter cap have less losses the ours since it has a filter cap?

I'm just wondering if these a better way to set this up. Having 50% loss is alot since we have no filter cap and when our plused DC swings low we are theoretically standing there with our thumb up our a**.. so I just think there's gotta be a better way of doing this to help get our Ripple out. The ballast might make the flybacks output a high enough frequency to not matter altho my small ones like the booster coil I'm sure are much lower. Is there a way to do that type of inductive jump on a booster coil one? Is there any disadvantages other than making everything be able to handle those very high voltages doubling 200kv.

And for math let's just talk on my big one. How do I calculate tank cap. Is there no way? Since output is kinda unknown? At least my 6watt version I noticed as I changed capacitance it needed to be retuned on the primary. Which makes since. But let's say you have my setup with rotary, would you be doing anymore calculations prior to testing?.

Could I make a resistor choke to limit my power for testing? Maybe have like 3-4 different resistors to step up current slowly? If I could do that what values would be good. I have no clue what current will be flowing so I'm not sure where to start.

 

I would like to point out that resistance and impedance are very different things, and are not interchangeable. Resistors have resistance, while inductors have impedance. An inductor does not resist high frequencies, it impedes them. Getting the terminology correct will help you in the future.
That being said, the reason for not using an inductor with the boost converter setup is partially that it may impede the flow of the pulsed DC. The primary reason not to use one is that doing so would attempt to smooth out the output of the boost converter. There is no reasonable way to smooth that pulsed of a supply. Also the inductive kickback may cause unwanted voltage spikes that destroy the boost converter. I don't suggest putting too much effort into designing a could around one, as they break very easily, and you won't get that many runs out of one.

A resistive circuit with a smoothing cap would not be more efficient than the current setup, as it will have 50% loss. The current setup does not have that much loss. A tank capacitor will always swing low after the spark gap fires. The boost converter has an internal capacitor from what I can tell, and it will serve as the tank capacitor. You could likely add very small amounts of extra capacitance to it and be ok.

You really don't want that high of voltages on a TC primary. Even massive ones don't go much above 30kV, the only thing that increases is current. Therefore an inductive jump setup with the boost converter is unnecessary, and would almost certainly break the boost converter. The most I would do with one is try to add a little extra capacitance for more discharge current.

You could indeed use different resistor values to limit testing current. I would suggest a large rheostat for this. They make them into the 100+ watt range. I would suggest a 0-100 to 0-500 ohm. They're not badly priced on Amazon.

As for the math, I'm pretty sure some calculation could be done on the flyback model since its output is somewhat known. I'll ponder that at work today and reply separately on it later. 4am math before work is not my thing. I'll also link to some rheostats later if you'd like.

 

Word okay. Yes I knew the difference but yes I lacked the terminology my bad haha. But that all makes since on the booster.

I have a ridiculous amount of high wattage resistors some in the 400watt range. but on resisitive current choking, why is it on my boost converter running on a PSU with a 25amp 5v rail, I did 5v / 10ohm = .5amp I had to run closer to 3-4ohms for that to pull .5amp? and around 2ohms was 1.3amp.. what good does the math do me when I'm obviously missing some variables. is it just the circuits impedance screwing me on the math?

For most circuits so far I have been using a 600 watt dimmer. Worked great on the 12VAC step down 6 watt version, and with the CFL ballasts, which where non dimming. It's also used on my big ballast right now although I have only tried it all the way down and switched it on and got that good of an arc. Since it was that big already i figured it wasn't really restricting even though it might have been. I was trying to find a dimming ballast so the dimmer would work, probably really well but wasn't sure if anyone had tried that.

Why cant I find anyone doing a spark gap in a light bulb or something? like one with the filament broke. I also thought of an argon filled tube or so. but wouldn't that make it allllllllot quieter? Obviously that prevents the use of a rotary gap but yeah just a thought.

I did get my mini booster coil working with a rotary gap! YAY! I'll post some pics tomorrow. Me and the father in law are testing the big one around 1 tomorrow. Weather was shitty last few days. Still static single gap right now, If test goes well ill be making a big rotary gap for it. If this ballast says 40 watt is that what Im throwing at it? To compare to my 6watt? and if things dont break ill get a higher wattage ballast and with dimming if you think that'll work.

your totally fine on the math man take your time I cant tell you how much I appreciate all the help!!

 

Were you checking the voltage of the 5v rail while testing? Some PSUs do not actually regulate that well, especially on less utilized rails such as the 5v rail. If the voltage had dropped under load, that could have caused you to need a lower resistance to pull .5A. A resistor should have no impedance, unless of the wirewound variety. Even if wirewound, any impedance would be irrelevant when DC is used.

What brand/model dimmer are you using? A dimming ballast will be quite difficult to come by except in CFLs. Also, the most powerful electronic ballasts I know of are for 4 bulb T5HO (high output). 4x54W.

The problem with a spark gap in a light bulb is that the gasses in a light bulb are at a low pressure, and are more conducive to forming a glow discharge that will not stop as the voltage drops. This would effectively short circuit the supply. The other problem is that the lead wires to the filament cannot handle any appreciable current without melting right down to the glass seal. (from experience) The issue with an argon filled tube is that once he arc starts, it would likely not cut out until power was completely cut. Most blue colored "neon" signs are simply argon filled tubes.

Congrats and I'm definitely looking forward to the pics! All I need is some magnet wire and a 25W 0-1k rheostat to come in and I should have a functional tesla coil. The rheostat is on backorder though. A boost converter to give me a higher DC voltage to work with is also in order. Or just a higher voltage DC supply (around 60V). I have a 12V @ 50A supply, and from there it jumps to 3kV @ 2.5A.

If the ballast is rated at 40W, it will likely deliver up to 60W when connected to a flyback instead of a fluorescent tube.

You're very welcome, glad I could be of help.

 

I think my dimmer is a Lutron 2-3way. I tried the old school lutron one with a knob that said only for incandescent @ 600watt rheostat, that one popped when testing a self wound Transformer. So the new one is a slide type that says it's for incandescent, CFLs, or leds. Inside has quite a bit more circuitry. But it works with all of them so far. Even the big one!!!! Woohoo, I found that out when testing the big one today. When I cranked it up it most definitely gave it more current.

So I'm not that worried about a dimmable one. Although I did find quite a bit of dimmable big ballasts online.

The big coil did great today. Had about 6in streamers ehh okay. I figured quite a bit better especially if you say I should be at 60watt... But the spark gap was wicked as all hell! That was cool! So for tunning it I just messed with primary turns to get further Sparks and the added a cap and repeated till I had the best. My confusion is when I add more caps the spark gap obviously slows down but I'm I going backwards? Do I want a fast spark then use a rotory gap to slow discharging down? It does seem better with more caps and its slower but eventually it wants me to clap my primary at .5-1.5 turns with 4 caps. Instead of 8 turns with one cap. So i run myself out of primary shrinking wise. How would the rotary gap help when my gap is sparking slow? I'm obviously doing something wrong or backwards.

And for more power now lol. Can I get more current out of this ballast by using both yellows and both blues? Or am I already using all of its output current?

And for pics sry I got caught up working today. I'll post them as soon as I get a chance. I spaced it this morning when testing my bad lol

 

I spaced replying about the PSU. No I wasn't checking for any voltage deviation altho i figured it would be quite minimal pulling .5-1amp when it can supply 25amp. But thats something ill check and get back to you.

I keep wondering why, I seem to be getting a better output on my SSTC on a 20v .5 amp supply. I mean not 6inchs but i get streams off secondary without a ground probe right next to it. Every SGTC i got working only outputs on the secondary when I have a grounded wire to capture the sparks. why is this? I really like the constant streamers. Is this due to problem i have?

This is a lot to take in on these two posts so sorry ahead of time. Since my SSTC constantly strremed, I was thinking of making another with a 220-14 0 14 transformer out of my old tig welder. It pretty big, about both my fists. not sure on current output but if close two 10 that would be sick! My first TC was this schematic bellow. would this work with that power supply? maybe with a couple mosfets in parallel to divide out the current? Or is there a better more complex version that utilize's the center tapped low voltage secondary.

 

Those old school dimmers are pretty much just a triac, so any time they are used on a non-resistive load weird things start to happen that usually lead to popping it. That new Lutron is a pretty good dimmer for this application though.

I guess it's been a while since I looked for a large dimmable ballast, seeing how they're easy to find now lol. I suppose last time I checked magnetic ballasts were the norm.

As you add more capacitance, the only way to keep sparking at the same speed is to supply more power to charge it faster. Adding a rotary won't help that. Having a fast spark and using a rotary to slow it down slightly should result in more consistent pulses, but slow it down too much and you'll be leaving power unutilized. As for running out of primary, maybe having a faster spark gap is the key. Have you tried moving the points closer together?

As for power, you are already at the limit of your 40w ballast. If electronic ballasts have gotten better at power regulation when connected to a load other than a lamp, you may be right at 40w. As I said it's been a while since I've dealt with electronic ballasts for this purpose. Last time I tried running a flyback with one, they didn't regulate power well, and didn't last long as a result. Good powerful ballasts to check out would be 2 bulb F96T12 or 4 bulb F54T5 models (150W and 216W respectively). They can be had on Amazon for about $40 new, or $30 if you go for the "used" option of open box from Amazon warehouse deals.

No worries on the pics.

EDIT: You replied while I was typing lol, give me a couple minutes to reply to that next one. Also I left a message in your inbox.

 

On second thought that second question is going to require more time than I have this morning to answer, so I'll get back to you later today.

Your difficulty getting constant streamers on a SGTC is likely one of efficiency. Much more of your input power is wasted as heat, light, and sound by the spark gap. Therefore there is less to be had at the secondary. As I said I've got a 16kW SGTC in the works, and I expect about 12kW output.

 

Okay so today I was testing my mini with this big transformer. But it was connected to 120mains not 220. Just for testing. So this supply barely works with my medium that kicked as on my 20v .5a supply. Hmm.

I was just wondering if adding multiple mosfet helps more current flow. I added one more in exact parallel. It blew. Haha. Maybe this type needs a gate resistor to isolate form the other? I only ask that cuz as I turn up my pot I sent allot more current flowing altho as it goes up my Sparks get much worse... Well here it the mini in action as sstc @ 15v .3a. Sparks are decent for a 1.5in secondary is. CFL is bright as shit!!!

 

Probably does need a gate resistor on both to equalize. Then again I'm not particularly familiar with mosfets. I have more experience with spark gaps and vacuum tubes. Looks great though! By chance did you go off a YouTube video by keystone science?