BrOdin coil - High power Big sparks!

[davekni]

It could be just the rectifier.  It could instead be that one or more IGBTs fried, which then fried the rectifier.  Unfortunately, that would be my guess.  Fried IGBTs likely have a collector-to-emitter short that you can measure with a meter.  If no IGBTs show as shorted, then I'd run the bridge at low voltage (say 5-50V) and scope the outputs.

[fh89]

It could be just the rectifier.  It could instead be that one or more IGBTs fried, which then fried the rectifier.  Unfortunately, that would be my guess.  Fried IGBTs likely have a collector-to-emitter short that you can measure with a meter.  If no IGBTs show as shorted, then I'd run the bridge at low voltage (say 5-50V) and scope the outputs.

Just checked, no C->E shorts! Seems like I got lucky and it's just the rectifier!

The rectifier was only rated for 40A rms and I was pulling 180A+ even before upping the primary current, not only that, but after looking closely at it, the way I had it wired was not optimal, with all of the current going through a single diode pair instead of paralleling all 3 pairs like I should have! Kind of unreal that it didn't blow up before this! Just ordered the 100A version as a replacement.

[Uspring]

An interesting coil you got there. The combination of massive racing sparks together with only tiny topload breakout seems to imply, that the top load voltage can't alone be the source of the racing sparks. A much more likely cause is the voltage induced by the primary coil. The idea is, that there are 2 components to the magnetic field near the secondary. One component is the field caused by the current in the secondary. That field is pretty much constant along the secondary and causes a nearly linear voltage drop from the top down to the bottom.
The other field component is generated by the primary coil. That is localised around the bottom of the coil and causes an additional voltage drop there. The amount of voltage induced depends on the primary current and the geometry of the primary coil and its distance to the secondary.

The field of the primary coil drops off quite rapidly in the direction upward. The primary coil can be seen as a collection of rings, each one carrying the same current. The field of a single ring extends up to roughly its diameter, more precisely: The voltage induced in the region from the bottom to one diameter up the secondary will be about 90% of the total voltage induced along the complete secondary.

So there are two ways to avoid the induced voltage drop, which will cause the racing arcs.

1. Move the secondary away from the primary, i.e. up. The penalty for this is lower coupling.
2. Move the primary coil outward, i.e. make it larger and particularly avoid too small inner rings. That will cause the primary field to extend further up and will distribute the voltage stress along the secondary. There is also a coupling penalty, but it is less than solution 1.

You don't seem to have a large diameter primary, which possibly aggravates the arcing problem. But this just a guess and certainly a lot of work to change.

@davekni: The model implied by the above can be tackled almost analytically by calculating center fields of current rings. These are pretty simple. An even simpler model is a split of the secondary in 2 parts, one strongly coupled and the other not at all. Similar to a magnifier setup. I believe JavaTC obtains all these results but does not export them.

[Kizmo]

It could be just the rectifier.  It could instead be that one or more IGBTs fried, which then fried the rectifier.  Unfortunately, that would be my guess.  Fried IGBTs likely have a collector-to-emitter short that you can measure with a meter.  If no IGBTs show as shorted, then I'd run the bridge at low voltage (say 5-50V) and scope the outputs.

Just checked, no C->E shorts! Seems like I got lucky and it's just the rectifier!

The rectifier was only rated for 40A rms and I was pulling 180A+ even before upping the primary current, not only that, but after looking closely at it, the way I had it wired was not optimal, with all of the current going through a single diode pair instead of paralleling all 3 pairs like I should have! Kind of unreal that it didn't blow up before this! Just ordered the 100A version as a replacement.

That would be lucky indeed if only rectifier blew up.

Your currents and duty cycles are well within the region where i started killing CM600s

[fh89]

Finally got my new rectifier (11 day shipping from 50 miles away, WTF USPS?) and we are back making huge, hot streamers!  It was only the rectifier that went boom.

I am not able to lift the coil onto the stand by myself, so had to run at ground level, limiting max spark length. Anyway here are some short vids. This is the same run shot from two different angles, one a gopro hero8 2.7k@60 and a note 8 at whatever max is.

[AstRii]

Have you tried touching the sparks at low power? At these high ontimes and low BPS it should not hurt. I would try to decrease the power so that it makes only a few inch sparks and try it. Imagine how cool would it be to touch so big streamers. But I guess that this hot plasma would burn your skin quickly.

[fh89]

Have you tried touching the sparks at low power? At these high ontimes and low BPS it should not hurt. I would try to decrease the power so that it makes only a few inch sparks and try it. Imagine how cool would it be to touch so big streamers. But I guess that this hot plasma would burn your skin quickly.

I understand the skin effect and have touched the output from much smaller coils, even up to 4' sparks, but at much lower power levels, say 1-2kW. Those sparks leave pinhole burns in the skin and hurt a moderate amount. Holding something metal in your hand can eliminate the pinhole burns, but it still gives a strong shock feeling.

 The thought of touching the output of this monster is quite frankly terrifying, even at low on-time and BPS because of the energy in the bus caps/primary. One mistake there and it's pink mist and closed-casket.

At full power, the sparks are lighting the dry grass on fire and melting my frame bolts, I don't want to go anywhere near them, even with a 10' chicken stick. Getting hit by one of those suckers may not kill you, but I'm guessing it's a massive understatement to say it won't feel very good!



On another note, I just ordered materials to build a 14x52 toroid. Hopefully stacking that on the 10x36 will get the topload to where it needs to be for maximum spark length!

[davekni]

Great that it was just fried rectifier diodes!  Nice performance again.  Looks like there was a little left-to-right breeze.  If you can get that sort of breeze when the coil is higher (and eventually with a larger toroid), that would be great for stretching arc length.

[Kizmo]

Have you tried touching the sparks at low power? At these high ontimes and low BPS it should not hurt. I would try to decrease the power so that it makes only a few inch sparks and try it. Imagine how cool would it be to touch so big streamers. But I guess that this hot plasma would burn your skin quickly.

Total capacitance of large tesla coil is so large that sparks do hurt like hell no matter what the duty cycle is.

I have touched tiny sparks from my BiggerDR, it was not pleasant.

[fh89]

I had some friends over last night for a party, so brought out the big coil for a show. I was planning on a quick demo and playing some music, but ran into an issue and had to cut the show short. Finally got a decent camera to record on:

shot on Canon EOS rebel T7

Same run shot on GoPro Hero 8:

The smoke at the end turned out to be just a coolant leak. One of the hose clamps on the primary lead coolant inlet become superheated through inductive heating and melted right through the tubing, causing coolant to spray everywhere.



This is a pretty simple fix, just cut off the melted tubing and replace clamp with zip ties.

The sparking was from an arc forming between the isolated components of the frame. The two pieces the arcing was between were at opposite ends of the ground so to speak from the isolation I did earlier in this thread. Apparently a direct hit from the coil is enough to overpower this gap even though an alternate ground path was provided by drilling/screws through each bracket and rail to punch through the anodization.

 

[davekni]

The sparking was from an arc forming between the isolated components of the frame. The two pieces the arcing was between were at opposite ends of the ground so to speak from the isolation I did earlier in this thread. Apparently a direct hit from the coil is enough to overpower this gap even though an alternate ground path was provided by drilling/screws through each bracket and rail to punch through the anodization.

Interesting data.  An arc strike will have some high-frequency energy on first hit.  Apparently that high-frequency current generates sufficient voltage across the inductance of the path to ground to jump your insulation gap.  Either that or the secondary arc is hitting so close to the gap that the ionized air starts the arc.  Either way, once an arc starts, the inductive loop of structure members coupled to the primary will generate high current to sustain the arc and feed it high current until the end of a burst.

[Hydron]

As far as failures go, I think you could do a lot worse!

[fh89]

Got the coil all patched up and upgraded! After running the bus voltage drop calculations on Mads' site and ending up with a nearly 100v drop, I added 10,000uF more to the bus caps, now totaling 26,800uF and bringing voltage drop down to around 20v.

Another shot for scale (I'm 5'10" 195lb):


GoPro angle

DSLR angle

[TMaxElectronics]

Did you ever consider replacing the aluminium extrusions on top of the box with a plate of acrylic or polycarbonate? That would probably solve the issues with arcover that you are having without requiring a complete rebuild.

I have a thick sheet of Polycarbonate on top of my coil, strengthened with some of those extruded pieces (but without them touching). Like this:


That is very strong and didn't give me any issues with voltage so far

[fh89]

Wow, that is a fantastic looking coil, great job on the construction! The adjustable coupling is a great idea, would have been great to have on my coil. Maybe I will do something along those lines on my next build.

For now, I'm not too worried about the arcover on the frame. It does melt some of the aluminum, but it is not serious enough to compromise the structural integrity of the frame, and it does seem to fizzle out after arcing over for a few seconds. I figure that it will either become carbonized/oxidized enough or enough material will be vaporized that it stops arcing over in that corner.

[Mads Barnkob]

Wow, that is a fantastic looking coil, great job on the construction! The adjustable coupling is a great idea, would have been great to have on my coil. Maybe I will do something along those lines on my next build.

For now, I'm not too worried about the arcover on the frame. It does melt some of the aluminum, but it is not serious enough to compromise the structural integrity of the frame, and it does seem to fizzle out after arcing over for a few seconds. I figure that it will either become carbonized/oxidized enough or enough material will be vaporized that it stops arcing over in that corner.

Great run! and that amount of DC bus capacitance is going to give a show if you explode and IGBT

So keep the cameras rolling! You have the rock and roll attitude of Tesla coiling with "enough material will be vaporized that it stops arcing over in that corner." 

Your secondary is much much wider than I first thought, its easy to read a measure and imagine something, but your new pictures of standing next to it....

[fh89]

Great run! and that amount of DC bus capacitance is going to give a show if you explode and IGBT

So keep the cameras rolling! You have the rock and roll attitude of Tesla coiling with "enough material will be vaporized that it stops arcing over in that corner." 

Your secondary is much much wider than I first thought, its easy to read a measure and imagine something, but your new pictures of standing next to it....

Probably inevitable that I blow an IGBT eventually, hopefully not for a few years though!

Lol, she's a girthy beast for sure! Hard to gauge the actual size of the coil without any point of reference.

 I really need some easier way to get it higher off the ground, the stand I built is nearly impossible to get the coil on alone. I think it would really help horizontal spark length.

[Mads Barnkob]

Great run! and that amount of DC bus capacitance is going to give a show if you explode and IGBT

So keep the cameras rolling! You have the rock and roll attitude of Tesla coiling with "enough material will be vaporized that it stops arcing over in that corner." 

Your secondary is much much wider than I first thought, its easy to read a measure and imagine something, but your new pictures of standing next to it....

Probably inevitable that I blow an IGBT eventually, hopefully not for a few years though!

Lol, she's a girthy beast for sure! Hard to gauge the actual size of the coil without any point of reference.

 I really need some easier way to get it higher off the ground, the stand I built is nearly impossible to get the coil on alone. I think it would really help horizontal spark length.

You look to be working out quite a bit more than I do  I can barely lift my secondary coil in place alone, in a safe manner where I would not scratch it, so my large coil is also too big for a single person to manipulate beyond its original physical dimensions. I have thought about making a base, in a distant future, that would be ready-made for a cheap motorcycle scissor lift like this: https://www.expondo.dk/msw-sakselift-mc-150-kg-10060972 or something similar, it would properly need a bigger foot print for stability.

[fh89]

You look to be working out quite a bit more than I do  I can barely lift my secondary coil in place alone, in a safe manner where I would not scratch it, so my large coil is also too big for a single person to manipulate beyond its original physical dimensions. I have thought about making a base, in a distant future, that would be ready-made for a cheap motorcycle scissor lift like this: https://www.expondo.dk/msw-sakselift-mc-150-kg-10060972 or something similar, it would properly need a bigger foot print for stability.

Ha ha, well I am a bit of a "gym bro", hence the "BrO"din name of my coil! My secondary is also very difficult for me to lift and install or remove from the coil base. It is relatively heavy but manageable at 45kg/100lb, but the physical dimensions make it quite difficult to grasp and maneuver. Lifting the assembled coil requires at least two people. A jack like the one you mentioned is something to think about for future builds!

[alexhanyuan]

Have you measured the arc current? Must be tens of amps to be doing that to the aluminum frame. Certainly not 'high voltage, but low current' anymore haha.