Tesla Coil Build Thread

[Polonium210]

I would recommend starting with a half bridge until you are sure your driver works otherwise you will be blowing 4 times the parts instead of 2.

Just a thought

 

[Sigurthr]

Aye, only problem is I have zero experience with half bridges and am unclear on their power supply topology, where as I have experience with full bridges, just not as an inverter.

 

[Hiemal]

I'm here Sig, what's up?

Also.. I've had that weird pattern appear on some of my mosfets that got too toasty as well. It's strange and I'm not sure why it happens, but it does.

If you take a look at the plasma speaker I made you it uses a half bridge and GDT... you may be able to infer some design ideas off of it.

They're both really easy to do.

http://t3.gstatic.com/images?q=tbn:...p1Tru67pB_IeAuC_KuTj6zpHxNP67mjnspG0ZRi-0irxg

More or less, that's basically what you need to do. You can also look through the 4hv.wiki site. It contains a good amount of information that you might find useful.

 

[Seoul_lasers]

Well, FETs and other semicoductors are packaged in epoxy. IF the internals get hot enough you'll get warping of the outer case. Had this happen with transistors before as well.

 

[Sigurthr]

Well, that supports the idea it was thermal death then. I figure once the deformations were created the thermal contact was compromised which lead to thermal run-away. Shame, I was enjoying it immensely!

I LOVE my big coil but it is so loud that I can't leave it on for more than 30sec without annoying someone rooms away. I wonder how hard it would be to later convert it to an OLTC.....

I decided I'm going to make Steve Ward's SSTC-5 next (after I get my first paycheck of the new job I hope to land during next week's interview). It is the "Mini SSTC", the one Jared linked earlier, which is a IRFP260 half-bridge fed from mains using UCC37322/UCC37321 gate driver pair and standard GDT. I ordered enough fets and heatsinks and freewheeling diodes for two of them last week and they should arrive some time next week. It may be a few weeks until I get that said paycheck, so don't expect progress quickly. I'll have all the parts for two ordered, so I can either make a second or replace any blown bits. I also drew up the PCB for this driver in CAD software, but it would cost $125 to have two boards made (and that's the best cost), so I may do it eventually, but not for a long while. If the SSTC comes out nice on my prototype I may invest in boards and offer it as a kit.

As far as needing another bunch of eyes to look over things....
Here's the quick schematic I drew up for a full bridge SSTC. Let me know what you think please! I don't plan on making this for a long while, but I would like to some day. I've never used a bridge for an inverter, it has always been motor controll, so I never had to use a GDT as it always ran continuous DC for motor reversal. Please take a look at the phasing array of the GDT and see if it looks good. The actual drive section is a copy of what is used in the "Mini" SSTC. To be honest I'm not qure sure what Steve Ward is doing with the 555 based interupter... I've never seen a 555 used with two pots between pin 6/2 and pin 3... but I trust it works somehow... I assume one is PRF and the other is PW, but I've never built a 555 PWM controller.

Btw, ignore all the differences in the GDT's winding sizes, they're all ment to be identical number of turns and in parallel on the secondary side. I haven't figured out yet if it would be best to do pentafilliar (is that a word!?) or have the secondary more/less turns. Without a scope and actually building it I wouldn't be able to tell if the Vpk on the secondary side is the needed 10-12V or not, or if there is enough current in each winding to feed the gate capacitance. When you're making a GDT how do you estimate these things?

 

[Hiemal]

I'd stick to doing a half bridge for now; full bridges are a bit harder to work with. That and winding the GDT is a lot easier.

But, yes, you more or less have it down correctly.

 

[Sigurthr]

Ah, good, thanks!

Yeah I'll do the half bridge first for sure. I am SO tired of half-assing things with cheap parts and sloppy builds. I can't wait to do something the right way once I'm working again.

I got the topload for my SGTC today and got 16" streamers out of it for a few hours today! Unfortunately somehow the tuning seems to have changed without me changing anything (changed since I retuned for the new topload) and I can only pull 13" streamers out of it again. Primary went from 13.25 turns to 11.75 turns to correct the detuning. I think it must have been freak luck earlier which allowed the streamers to grow ontop of themselves several times and capacitively load the coil allowing it to run with too much primary inductance and now I can't repeat it so I have to tune the coil classically. The only thing I did between the post topload addition tuning and the new corrective tuning was to clean and adjust the spark gap because ablation wore them down to the point that it was not firing routinely.

Here's video from before the mysterious detuning:
Spark Gap Tesla Coil V2.1 - YouTube

 

[Seoul_lasers]

I think I see your tuning issue two fold now.
You're using brass as your gap instead of using copper pipe. Brass will melt/vaporize quickly in spark gap operations. I can hear your gap now loosing quenching. Doesn't sound like a continuous roar, but rather poppy.... that's a bad sign.

Remember the copper pipe gap I posted earlier.. this is what I seriously recommend. You'll notice a much bigger improvement in performance too.
Also, please raise that primary as well, it will also transfer power from the gap to more of the windings on your secondary coil. Right now probably ~40% of your actual power is wasted in the gap as it isn't being used by the secondary. Bump the primary up, and you'll notice the gap sounds less like a rotary machine-gun and more like a roar.

The secondary acts as a load on the primary circuit.

 

[Sigurthr]

YES I think you are absolutely right.

I just noticed I still had white jumper wires going to the spark gap about an hour ago. I am such a dunce some times. The gap leads need to cary the HF HV HI current in the tank circuit. I changed the jumpers for 12ga w/ soldered on terminals and the tuning point shifted right back where it should be on the 13th turn. I have it at 13.0 turns now instead of 13.25 because I think it was just a bit far from resonance before and relied too heavily on streamer loading. Right now I'm actually seeing 16" regular and 18" peaks. I'm so glad it is working tip-top again!

Yes, I do understand and will implement the all-copper spark gap and raise the primary once I have the funds and materials to do so. The brass is vaporizing and that is definitely an issue (it requires me to check them periodically and correct for it, which is a pain) but it does not seem to actually affect the resonant frequency.

I can say this is definitely my last SGTC. I love the performance but damn, they are loud and real finicky about mechanical details. It's all SSTCs (and DR's) from here on out!

 

[Hiemal]

What about VTTC's?

Oh, and regarding the GDT;

Typically it is a 1:1:1 transformer. Each winding is around 12-15 turns. You need to find a ferrite toroid. It can't be one of those clamp on filter ones; they're lossy and don't make for good transformers.

 

[Sigurthr]

Yeah I forgot to mention VTTCs and OLTCs, I'm still undecided on them. I'd like a small lower power desk VTTC, but I'm not in to the sword like streamers and insanely hot flame like arcs. It would be very cool to have (and esp. to have made myself) since I'm learning EE/RFE. I just have no practical knowledge or experience with VTs, and every VTTC build I've ever seen requires pretty intense familiarity with VTs as well as excellent mechanical lay-out/logistics. I would want to make one which could run for hours with no problem, and have both the ability to run CW and interrupted with varying PRF and PW.

Re; GDT: Aye, 1:1 on all windings is all I've ever seen so far. All the tech. papers I find on it say better to have too many windings than not enough, to prevent saturation even if the leakage inductance rises. You'll be able to see on a scope if there is ringing from unbalanced leakage, but you can't see if the core is saturating quite as easily especially when "walking saturation" comes in to play (which I didn't even know existed before!).

Here's the core I'll be using: B64290L0616X035 EPCOS EMI/RFI Suppressors & Ferrites
I picked it out using a design guide by one of the DRSSTC founders, I can't remember which though. It looks like it should work fine, 1" dia, good and thick, high grade material, decent price, etc.

I do wonder though, if a 1:0.7:0.7 (say pri 15t, sec1 11t, sec2 11t) GDT would be useful in a situation where you can run your gate drive chips off of say 16V (most are rated around 18V iirc), and then get 11V @ 1.3x the current out on the secondaries. This would compensate for gate capacitance, especially at higher switching speeds and reduce the heating/loading of the gate chips.

I've got just about all the components saved for the project on Mouser now, if you want I can share the project file on mouser to ya via email. I have enough parts down for two complete builds plus extra parts where I think they'll be needed. The only things I might have omitted are some resistors I know I have plenty on hand of.

 

[Seoul_lasers]

YES I think you are absolutely right.

I just noticed I still had white jumper wires going to the spark gap about an hour ago. I am such a dunce some times. The gap leads need to cary the HF HV HI current in the tank circuit. I changed the jumpers for 12ga w/ soldered on terminals and the tuning point shifted right back where it should be on the 13th turn. I have it at 13.0 turns now instead of 13.25 because I think it was just a bit far from resonance before and relied too heavily on streamer loading. Right now I'm actually seeing 16" regular and 18" peaks. I'm so glad it is working tip-top again!

Yes, I do understand and will implement the all-copper spark gap and raise the primary once I have the funds and materials to do so. The brass is vaporizing and that is definitely an issue (it requires me to check them periodically and correct for it, which is a pain) but it does not seem to actually affect the resonant frequency.

I can say this is definitely my last SGTC. I love the performance but damn, they are loud and real finicky about mechanical details. It's all SSTCs (and DR's) from here on out!

SGTCs are much more forgiving than
SSTGs. I'd seriously encourage you to fix what you got already
before moving on. jumping from SGTCs to SSTCs is like jumping from
a porche to an alcohol fueled dragster. Lots can go wrong on a dragster!
You're relatively new to TCs. I would hope you take a bit more time
and make the corrections I've noted before moving on.
I'll say it again you have 2 problems. Both easily correctable.

 

[Sigurthr]

SGTCs are much more forgiving than
SSTGs. I'd seriously encourage you to fix what you got already
before moving on. jumping from SGTCs to SSTCs is like jumping from
a porche to an alcohol fueled dragster. Lots can go wrong on a dragster!

It's kind of apples to oranges though. In a SGTC a lot of what is forgiving is the electrical side of things, which I have a lot of experience with and have down pretty well. The sensitive side is the materials side; their construction and placement. It is a lot easier to tune a SSTC where you can wrap the primary directly on the secondary coilform with just some plastic dielectric between them. You can raise or lower it as needed and easily add or remove turns because it is just wire.

You're relatively new to TCs. I would hope you take a bit more time
and make the corrections I've noted before moving on.
I'll say it again you have 2 problems. Both easily correctable.

I do not mean to sound rude or unappreciative at all, but really, I have been trying to say this for a long time now and have frequently written it in my posts. You have the attitude of someone who has only ever lived in cities and you really don't seem to understand what it is like living hundreds of miles out from any real city.

Where I live it is VERY hard to get the necessary raw materials to make physical modifications to the structure. There is no Home Depot or Lowes or public-access lumber yard. To get the plywood I purchased I had to buy a full 8'x4' sheet for $50 from a commerical contractor and pay him to cut it down for me, he only did it because of my connections with the local amateur radio club and as a favor. For PVC I only have access to 1", 3", and 4" stuff and only the 1" and 3" can be cut by the supplier in to custom lengths. They will not take small orders frequently, and not without a tip/bribe. To cut the 4" stuff I got I had to use a torch and a piece of steel to melt a hole in it enough to get my keyhole drywall saw in to it and slowly work it by hand, it took hours and did not cut straight at all. Our hardware store is in the back of a small town grovery store. They only cary what there is call for and most of the farms and businesses moved away from the area years ago. It is a VERY limited selection. 90% of the population here is over the age of 65 and do not do home improvement or hobbies which require building materials. When someone wants to have repairs done they typically hire someone in from downstate 200+ miles away. The only stores I have access to are Walmart, Walgreens, Big Lots, and RadioShack. That is IT, there is nothing else for hundreds of miles.

Now, as I said two posts ago; I WILL fix the primary coupling issue and build the recommended multi-static gap from copper pipe when I can afford to get the materials for it. Getting those materials involves driving >100mi to a real hardware store downstate and bringing them back up. It also involves buying the tools downstate to manipulate the materials into the shapes and sizes necessary. If it can't be done with screwdrivers, a hand drill, soldering equipment, a torch, a keyhole saw, or a 5" coping saw I can't do it (yet).

For the following reasons I would have a lot more success with a SSTC than a SGTC:
1) no physical primary support needed - No need for power tools or machining equipment.
2) small form factor means it could be built in to materials found at Walmart.
3) the major components are semiconductors and related parts, all of which I have a source for (Mouser).

I know an oscilloscope is damn near necessary, if not completely necessary depending on the build. I am planning on ordering one in when I can afford it (after I land a job).

 

[Seoul_lasers]

Sorry... I am not trying to push you around here.
I understand you live 100's of miles away from civilization. My point is just to take your time with your first build. That is all. Get familiar with it before you skip to SSTCs.
That's the point.

BTW, back in Canada I live far out in the country as well. In British Columbia
I happen to be in Seoul now.

 

[Sigurthr]

It's all right, no hard feelings. =)

And yes, you are right too. It is very important to go slow and do things right. However I will add an addenum; when you've found that what you were doing/planning is not all you thought it would be and then find something similar that turns out to be all you had hoped for in the beginning, it's okay to shift gears and then refocus and do things right. That's kind of how I feel right now with TCs. The noise level, physical build sensitivity, and sheer size of SGTCs are quite off-putting to me. After even what I consider to be a pathetic attempt at an SSTC with old, used, and insufficient parts; I realized that SSTC's are what I've been looking for.

So, I will work out the bugs in my SGTC just because I feel obligated to because I put so much time and effort in to it, but my heart isn't 100% in to it if you know what I mean. Where as with the upcoming SSTC project I'm planning, which will be done right (correctly, and slowly) from the start I feel quite passionate about it.

 

[Seoul_lasers]

It's all right, no hard feelings. =)

And yes, you are right too. It is very important to go slow and do things right. However I will add an addenum; when you've found that what you were doing/planning is not all you thought it would be and then find something similar that turns out to be all you had hoped for in the beginning, it's okay to shift gears and then refocus and do things right. That's kind of how I feel right now with TCs. The noise level, physical build sensitivity, and sheer size of SGTCs are quite off-putting to me. After even what I consider to be a pathetic attempt at an SSTC with old, used, and insufficient parts; I realized that SSTC's are what I've been looking for.

So, I will work out the bugs in my SGTC just because I feel obligated to because I put so much time and effort in to it, but my heart isn't 100% in to it if you know what I mean. Where as with the upcoming SSTC project I'm planning, which will be done right (correctly, and slowly) from the start I feel quite passionate about it.

Agree with your thinking I do..
Tesla coil like making works of art is a bit of meditation.
A local buddhist monk once told me, "be mindful of what you are doing in the moment. Do not allow passion, anger nor ego to cloud the moment. let those feelings go as they serve no purpose. Concentrate on what is in the moment." ~ quote Yonggeun Sunim of Jogye order.

Also, this same monk I see occasionally hanging out around Insadong, near my apt.
Another saying, Don't be like the rock, but be like the water flowing over it.
~ take your time and you'll get to your goal.

---- > Tesla coil sutra ....