WooHoo!!! Built and now just waiting for the toroid.

[stegu]

Air humidity and to some extent temperature have an influence on the electrical properties of air, so the weather does come into play. Thanks for the videos. I will probably cut and paste and use some stills from them in the tuning guide. I have yet to produce such long sparks with my own coil, but now I'm more eager to get there!

[DJKOR]

It has arrived!!! It is so beautiful



Popped her on and measured resonance. I have 323kHz and if I put on my modified secondary toroid (hollow in the middle), I get my previous 270kHz. I bought some more cable today however and will try to re-wind my primary to see how it goes.

[DJKOR]

Well, I think I have reached the point where it is going to stay for the time-being. I really wanted to keep the clean look of the stock toroid so I caved in and decided to adjust the primary winding. Also, when I used my additional toroid on the original 6-turn primary, I was getting corona loss from all areas of the second toroid due to the stamped toroid being so smooth.

As discussed in the other thread, I was aiming to get a primary resonance of around 298kHz. Unfortunately, the closest I could get with 5 windings spread out was 288kHz. After seeing some undesirable results from running with 4-turns from Tetsula (resulting in a blown tank cap), I didn't want to risk dropping another turn on the primary so decided to leave this as is. This now gives me a primary to secondary resonance of 288kHz/321kHz (took the breakout point out a little to drop 2kHz), or a difference of 11.46%.



I'll get some current readings later to make sure I'm not overloading anything however I managed to blow the $15 fuse in my multimeter the other day so need to change that first. All I remember is plugging in the coil, disconnecting the power then reconnecting it after about a second. Must have put a surge through it or something.

Edit: Gave a quick test run on fixed frequency mode. Saw a bit of arcing from the primary to secondary. Looks like the top turn lacks insultation due to being close to the top of the coil form. Guess my work isn't entirely complete. Bugger.

[DJKOR]

Well.... that wasn't fun. Fixed the arcing issue by smoothing off all sharp edges and increasing the height of the inner PVC sleeve above the windings.

New $15 fuse in the multimeter, I then proceeded to measure current draw on fixed frequency (mid-frequency). Current draw figures at mid to full power matched those I had measured before with my 246/270 setup. Checked the scope trace at about 45% power and it looked okay. Comparable to what I had seen in other tests where I had been able to run full power.



I then went into midi mode and ran the first 20 seconds of "Beetlejuice" as I knew the current figures from those from my previous testing. First ran at half power to ensure all was good, checked the primary temperature for any excess heat, then ran again at full power. Everything appeared to be measuring the same. Visually, everything looked and sounded like previous runs with noticeable change to the spark intensity (not length as noted before) as I went above 50% power. So then I clicked stop, loaded "Paranoid", turned the power to full, then clicked play. All went good until 7 seconds in. All of a sudden... silence, except for the multimeter beeping like mad like last time the fuse in it blew. Figuring the damage has already been done, I remove the multimeter from the circuit and plug the mains back in. This time, all the onboard LED's fade off. Mains circuit breaker was tripped.

IGBT's and/or diodes are shot.

My very first oneTesla failure. Might celebrate with a beer

Edit: So both IGBT's are dead and two of the four diodes. Inconveniently, I don't have any spare MUR460's so looks like I will have to wait until Tuesday to get them as they have to be shipped from interstate.

I'm tempted to reset everything again and continue with my known good "double toroid" configuration... provided I get no further catastrophes upon rebuild. I'm going to incorporate some TVS diodes across the IGBT's for added measure, but I think I will accept my oneTesla for how it was before with it's 6 turn primary and dual toroid. Yes the primary does get a little warm on full power runs of about 2-3 minutes, but everything felt "safer" despite that. I will proably just set up a cooling fan to keep the primary cooler.

Also, for curiosity, I cut apart the ceramic fuse I used on my main board. The insides showed signs that there was a good arc inside there which is what led to the circuit breaker tripping. I reckon if it weren't for KevinNewZealand's post about his fuse exploding which then led me to change out my fuse, I probably could have had the same explosive catastrophe on my main board.

[DJKOR]

Well, got my replacement diodes today and went to work putting everything back together.

I wanted to make the IGBT's easier to replace however so attempted something. At the moment, the heatsink is supported with nylon posts and hot glue. One is set up to be "failsafe" in that should the hot glue melt, the heatsink isn't going to fall under it's own weight. I will fabricate some method of mounting it solidly, but this will do to test how it holds up. It isn't the most ideal way to mount it, but I got to the point where I didn't want to try again with another method, so this is how it will stay. Hopefully having everything mounted in this method causes no issues with performance.

[KevinNewZealand]

that looks like a really cool way to make IGBT replacements easy should they fail ... What are you using in between the board and the IGBts? I am starting over on my main board and might like to try this out , in case of issues

[DJKOR]

that looks like a really cool way to make IGBT replacements easy should they fail ... What are you using in between the board and the IGBts? I am starting over on my main board and might like to try this out , in case of issues

I used 20A mains rated cable, stripped out of the flat 3-core cable that you usually wire your home in. It consists of 7 copper cores per cable and total area of 2.5mm2.

HOWEVER..... I have fired her up this morning and am not getting any good output. The spark coming off the breakout point is considerably smaller and quieter than my previous experience (even when I was out of tune by 25% or there abouts) and turning the power up a little is not doing anything.

As all I can think of that is different about my circuit since I blew it up is the change in the mounting of the IGBT's, I'm considering remounting them the way they were in the first place and seeing if that helps. If that doesn't work.... then I'm going to be a little less than impressed after putting in the effort to mount them the way they are now.

So far, I have 2 new IGBT's, 4 new MUR460 diodes, and I replaced the Gate Drive IC's. All I can think of is either I am getting some sort of loss in the cables up to the IGBT's, or the cables are creating a bit of interference of their own preventing the GDT from switching the IGBT's properly.

This will have to wait till I get home from work this evening.

[DJKOR]

SUCCESS!!!!!!! She lives once again. Had to mount everything as per the manual, but at least it works.

Here is the interesting part. If any of you have been following me, you may have noticed that I observed my oneTesla had a somewhat reduced performance compared to earlier runs even when its configuration didn't change. I just wrote that off as environmentals as we have just moved into Summer here. After replacing the IGBT's, diodes, and gate drive IC's (just to be sure), it seems to have its "bite" back again.

Maybe all those full power tests were stressing the IGBT's more than I knew and the test run at a higher primary frequency to match the secondary was all it needed to push it over the edge. Needless to say, half power has provided me with equally satisfying results as full power (loud is loud at the end of the day, just lacking a few more sparks) and I will leave it in it's current configuration... with the exception of fixing the secondary toroid so it doesn't have any corona losses... and live with running it half power at max. I have an extra cooling fan on top of the bus caps to give more directed airflow to the heatsink and have fitted heatsinks to the voltage regulators and gate drive IC's. I will all mount some TVS diodes to the IGBT's via the underside of the board when they arrive in the mail.

[Bayley]

that looks like a really cool way to make IGBT replacements easy should they fail ... What are you using in between the board and the IGBts? I am starting over on my main board and might like to try this out , in case of issues

I used 20A mains rated cable, stripped out of the flat 3-core cable that you usually wire your home in. It consists of 7 copper cores per cable and total area of 2.5mm2.

HOWEVER..... I have fired her up this morning and am not getting any good output. The spark coming off the breakout point is considerably smaller and quieter than my previous experience (even when I was out of tune by 25% or there abouts) and turning the power up a little is not doing anything.

As all I can think of that is different about my circuit since I blew it up is the change in the mounting of the IGBT's, I'm considering remounting them the way they were in the first place and seeing if that helps. If that doesn't work.... then I'm going to be a little less than impressed after putting in the effort to mount them the way they are now.

So far, I have 2 new IGBT's, 4 new MUR460 diodes, and I replaced the Gate Drive IC's. All I can think of is either I am getting some sort of loss in the cables up to the IGBT's, or the cables are creating a bit of interference of their own preventing the GDT from switching the IGBT's properly.

This will have to wait till I get home from work this evening.

So the reason mounting IGBT's that way causes them to die is that it adds a *significant* amount of parasitic inductance in the bridge, and that causes ringing which will kill the transistors.

[DJKOR]

Was in the mood to give it a run tonight to make sure things are back in order. I really need to fix up the secondary toroid however I don't mind it too much at the moment. More spark I suppose. Still only giving it half power until I get a better idea on what level of tuning is safest for this thing at full power.

Smoke on the Water: http://www.youtube.com/watch?v=N9I91oTX3GA&hd=1

Imperial March: http://www.youtube.com/watch?v=s0cdOv59a64&hd=1

Tetris: http://www.youtube.com/watch?v=qzuiuku-cps&hd=1

Zelda: http://www.youtube.com/watch?v=dxogio7Ndk0&hd=1

Dr. Who: http://www.youtube.com/watch?v=ZoU8lptPmnc&hd=1

And for some photo spam for the fun of it using my only decent camera... a point-and-shoot Sony DSC-RX100. All photos done with interrupter set to fixed frequency, highest frequency, mid power. You can right-click on any image and open in a new window/tab for a larger size.

2 second exposure, f1.8, ISO 125:


2.5 second exposure, f1.8, ISO 125:


5 second exposure, f1.8, ISO 125:


10 second exposure, f1.8, ISO 125:


Continuous burst mode. 1/8 second exposure, f1.8, ISO 3200: