Author Topic: Skm100 coil (Read 10907 times)

oneKone · « **on:** December 18, 2019, 04:48:35 AM »

Originally I wasn't going to post this coil, I thought it was just another coil, or not very interesting... Bit that's what this forum is about! It's convenient to post random pics and vids on Facebook but information does get lost there.

Scope of the coil:
Fullbridge of Skm100gb125
Epcos 4700uf 400v bus cap
MMC: 2 X kemet 220nf 3kv snubber capacitors in series
Driver: Profdc9 tht ud2.7c
Secondary is still up in the air.

*Incomplete post, just to save*

Mads Barnkob · « **Reply #1 on:** December 18, 2019, 09:43:46 AM »

Thank you for sharing, there is many small and interesting details in putting things together that is always nice to see!

At first I could not see it was two snubber capacitors bolted together, I thought it was some kind of primary tap clamp until I saw it was solid filled boxes

What kind of current limit are you planning for this? SKM50-200 have historically been running at some 350-700A ( http://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/igbts/ and https://4hv.org/e107_plugins/forum/forum_viewtopic.php?85229 )

Would you consider making another thread about your CNC router? Laboratories and Tools needs some content

https://highvoltageforum.net/index.php?board=29.0

Hydron · « **Reply #2 on:** December 18, 2019, 09:49:19 AM »

I have run SKM200GB125s up to maybe 700A (without any failure). They switch very fast and clean, if the 100s are the same or better you'll likely be able to run quite high frequency if needed (I have done over 300kHz, albeit for a short test only!)

oneKone · « **Reply #3 on:** December 18, 2019, 10:02:55 PM »

cheers @mads. tbh i think it would have been selfish towards the community not to share. there's so many documented TO-247/cmX00 builds, not so many skm builds. I'll most definitely post about the router, i'll also link to the parts.

also to @mads and @hydron,

I have two secondaries i'm going to try. The first is the one pictured, its 80mm diameter X 230mm winding length with 0.14mm wire (i think it was 170khz loaded). The second is 110mm diameter, 305mm winding length with 0.3mm wire (with a very rough calculation to be around the same frequency of the first coil). So far I've set the ocd to 500a, initially i think i had the ocd set around 600/700a as the primary was heavily detuned to get maximum streamer length, the only problem with this was i had a pretty hard time with racing sparks. I think with the 80x230 coil i got 660mm streamers, with the 110x305 it was 770mm (not really tuned at all, and running 150v ac). To me if i continued being haphazard i would have killed the bridge.

To the bridge! they're also aliexpress switches! they cost $27AUD (around $18USD). So far they've been rock solid, and awesome thermally.

Weston · « **Reply #4 on:** December 19, 2019, 06:44:05 AM »

The CNC'ed base looks really nice!

What is the part number for the MMC capacitor? It looks rather undersized and I am a bit dubious about it's ability to withstand a high peak current or modest repetition rate.

oneKone · « **Reply #5 on:** December 19, 2019, 08:02:20 AM »

Quote from: Weston on December 19, 2019, 06:44:05 AM

The CNC'ed base looks really nice!

What is the part number for the MMC capacitor? It looks rather undersized and I am a bit dubious about it's ability to withstand a high peak current or modest repetition rate.

Cheers,
This is the data sheet for the capacitors:
https://docs.rs-online.com/5fa1/0900766b81524bec.pdf

During testing I've noticed no heating from them, I like to think I'd notice heating from the caps before failure.

oneKone · « **Reply #6 on:** January 06, 2020, 04:06:28 AM »

Some time has past but there is still much to be done!

After my last quote of "I've noticed no heating from them" (the MMC) I started to notice heating, and the data sheet confirms why.... The most simplistic fix was to just order 2 more capacitors and have a 2p2s MMC. This did change the capacitance by double (going from 110nf to 220nf) but a higher amp rating.

The next part I need to change is the topload as it's too small now, i think I'll order a 300mm toroid in a couple of weeks. Aside from that I've upped the OCD from 500 to 600, the extra 100a really has helped streamer growth!

I'm now getting 670mm streamers to a pole. At 180khz i feel like I'm abusing these skm100s, but they're taking it! I think once I have the toroid I'll get some temperature readings with the bridge, atm the bridge is passively cooled but I will be switching to active cooling.

milkman2046 · « **Reply #7 on:** January 06, 2020, 08:02:05 AM »

This coil looks great! It looks exquisite!

What is the frequency of this coil?

Mads Barnkob · « **Reply #8 on:** January 06, 2020, 10:38:05 AM »

Quote from: oneKone on December 18, 2019, 10:02:55 PM

I have two secondaries i'm going to try. The first is the one pictured, its 80mm diameter X 230mm winding length with 0.14mm wire (i think it was 170khz loaded). The second is 110mm diameter, 305mm winding length with 0.3mm wire (with a very rough calculation to be around the same frequency of the first coil). So far I've set the ocd to 500a, initially i think i had the ocd set around 600/700a as the primary was heavily detuned to get maximum streamer length, the only problem with this was i had a pretty hard time with racing sparks. I think with the 80x230 coil i got 660mm streamers, with the 110x305 it was 770mm (not really tuned at all, and running 150v ac).

80mm diameter X 230mm winding length with 0.14mm wire would have a impedance above some 60-65 kOhm
110mm diameter, 305mm winding length with 0.3mm wire would have a impedance around the magic 50 kOhm

I would naturally prefer the latter, also for its height, so that you do not limit spark length because it all goes to ground.

Quote from: oneKone on January 06, 2020, 04:06:28 AM

After my last quote of "I've noticed no heating from them" (the MMC) I started to notice heating, and the data sheet confirms why.... The most simplistic fix was to just order 2 more capacitors and have a 2p2s MMC. This did change the capacitance by double (going from 110nf to 220nf) but a higher amp rating.

If you want to maintain the impedance of the system, you would need 8 capacitors in total, for 110nF, 2p4s. Now that you have lower impedance system, you should aim for shorter on-time and higher primary current to utilize the new configuration. Since you do worry a bit about the primary current, going back to 110nF and a higher impedance system, you could run longer on-time and lower primary current.

The picture of your sparks is lovely, love it

oneKone · « **Reply #9 on:** January 07, 2020, 06:25:30 AM »

@milkman2046
Cheers,
I think it's about 170-180khz loaded.

@Mads
You know it's bugged me since I wound it! I think when I order wire for the big coil I'll also order wire for this coil, I'll get a diameter to suit the 80mm diameter, but have the coil at a 4:1 ratio.

AstRii · « **Reply #10 on:** January 08, 2020, 04:54:44 AM »

Those are some nice streamers, yet i couldn't not notice you're running it with no shielding on the driver. Don't you have any problems with noise generated by the sparks?

oneKone · « **Reply #11 on:** January 08, 2020, 07:42:56 AM »

Quote from: AstRii on January 08, 2020, 04:54:44 AM

Those are some nice streamers, yet i couldn't not notice you're running it with no shielding on the driver. Don't you have any problems with noise generated by the sparks?

Cheers,
As for noise problems not yet. I've run all of my drsstc without the driver being shielded. I think bits more of a risk of a streamer hits the driver and risking the driver and/or igbts.

nevermore1983 · « **Reply #12 on:** March 07, 2020, 08:38:54 PM »

I have got a pair of Skm100's too, that i am basing my build around. Will be following this closely. I am in the very early stages of just research be for i buy any more parts. Hope my coil looks half as nice as yours, good work.

oneKone · « **Reply #13 on:** March 17, 2020, 04:11:03 AM »

Some time has passed on this coil, I work on it time to time... I think this weekend I'll fix a grounding issue and also increase the diameter of the toroid.

I had the bridge running upto 650a @ 100us on time, to me these skm100 modules are incredible! BUT it had to change. So I reverted to a 2s1p MMC (along with more primary turns) and have since lowered the OCD to 500a. Along with a bigger toroid the performance is also better.

I made a new toroid, following the same design as my larger drsstc. 2 plywood discs are cutout with slots routed in them to secure a piece of PVC, semi rigid ducting is then wrapped around the perimeter.

I can't remember the streamer length atm but it is over 3.5 X secondary length!

Mads Barnkob · « **Reply #14 on:** March 20, 2020, 01:57:26 PM »

Did you notice any difference in IGBT heat dissipation between the two configurations?

Bigger toroids are usually just better, since there is more capacitance to draw arc current from and it is less prone to get knocked completely out of tune

What did your resonant frequency end up at? I would like to add it to my IGBT list at the bottom here, as there is no SKM100 listed: http://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/igbts/

oneKone · « **Reply #15 on:** March 26, 2020, 01:50:03 PM »

to touch there was probably about 10 degrees difference, although the runs were at different interval/duration, this was still passively cooled. As well as i'm more comfortable running the switches at a lower apk.

It seems its actually tuned to 140khz! i think that's about 40khz below the unloaded secondary/toroid. I have also wound a new secondary, i wanted to keep it the same frequency to compare between how a secondary is wound. I'm just in the process of coating it with polyurethane, but am keen to post the results.

The new secondary is:
80mm diameter
400mm winding length
32awg wire double build (works out to be 0.25mm diameter)

It would be a massive honour to get an skm100 listed! I think once the new secondary is on i can get some actual "logged" thermals if you'd like?

oneKone · « **Reply #16 on:** March 30, 2020, 02:22:33 AM »

Update!

Last night I basically lashed the coil together to get a run in. I noticed despite the frequency being the same as the other secondary I was having trouble tuning the primary...... Until:

But let's be real, I knew this was coming, I was running them past their ratings and letting them overheat. Luckily I did also buy better rated capacitors (the same I'm using in my 160mm drsstc). I'm still going to have a 2s1p configuration, but the capacitors just have a higher rating in general. This will also bring the MMC capacitance from 110nf to 165nf.

This is the coil as it sat in the lashup stage last night:

Before I make my next run I still have to make a spacer so the secondary sits square, connect a better ground and maybe a slightly larger diameter toroid.

oneKone · « **Reply #17 on:** April 06, 2020, 02:16:56 AM »

*just adding some pictures of the bridge for reference*

oneKone · « **Reply #18 on:** April 11, 2020, 09:24:44 AM »

Tiny update.

I installed the new snubber capacitors for the MMC to make a 165nf 1p2s. I also (for the moment) turned the ocd to 550a, along with the new secondary I'm getting 900mm+ streamers with a very rough tune. I think this post will just be for reference as i dial the tune in, I'm also on the fence about making a slightly larger diameter topload.... Although I'm not getting any breakouts from it.

oneKone · « **Reply #19 on:** April 12, 2020, 02:14:28 PM »

Riding high in a win I decided to push forward with the build! I added soft start through a circuit I found online and active cooling, if I'm honest I was getting over using a variac to charge the bus cap.

/>
I think I still want to make a slightly bigger toroid, especially now with the breakouts. But tbh I'm happy with the results... The bridge now stays ambient temperature, the MMC doesn't overheat and it's just working at 550apk.

Edit:
Just found this pic in the mix... Over 1.2m streamer, I love when I get 3x secondary length streamers!!

Edit 2:
I forgot to mention the coil is currently running at 140khz. The topload is 370mm diameter. I think I'll make either a 450 or 500mm diameter topload. After that no more frequency changes will be made.

High Voltage Forum · « **Reply #19 on:** April 12, 2020, 02:14:28 PM »

Author Topic: Skm100 coil (Read 10907 times)

High Voltage Forum

Recent Topics and Posts