Author Topic: Drsstc 3 (Read 34546 times)

oneKone · « **on:** March 26, 2018, 04:13:08 AM »

While my second drsstc is on hold i thought I'd start trying to design another drsstc. The two coils I'm thinking of are:

110mm diameter, 440mm winding length, 0.3mm wire, 1460 turns. As for the bridge I was thinking about a full bridge of 82N120. Eric Goodchild wrote that 82N120 can run at the same frequency as the ixgn60n60. I'm thinking of using the 82n120 so I can go up to 600v bus.

Or

160 diameter, 640mm winding length, 0.3mm wire, 2130 turns.
With a fullbridge of cm300du-24.

As for the driver I'm on the fence to go with a 1.3b or an 2.7c driver, especially because I have an unpopulated 2.7c board, or if there's a point when phase shift should be added to a coil, at what point it's necessary. I'm hoping someone could maybe provide some insight with these more larger coils.

Mads Barnkob · « **Reply #1 on:** March 26, 2018, 09:45:09 AM »

Be careful with multiprojectitis, it is a horrible syndrome and we are many that suffer from it

I would advise you on building the larger CM300 coil, simply because it is a different beast of a inverter than the TO- package parts you have been using so far. It can take so much more abuse just from the shear size of the silicon die.

Both 1.3 and 2.x with phase lead will work great on your coil. Phase lead is used to minimize hard switching current when we are not hitting ZCS spot on, so as long as you can manage the extra temperature rise, there is nothing wrong with a little hard switching, but if you start pushing things to the limit and beyond it suddenly gets critical! If you run your CM300 at some 600-800A there should be no problems without phase lead, but if you want to push them above 1kA, you better well use it

You could always do the math according to the temperature calculations in my IGBT selection guide and see how much phase shift you can allow and still stay below 30 degree Celsius temperature rise per pulse (and be able to cool that off before next pulse).

oneKone · « **Reply #2 on:** March 26, 2018, 12:45:27 PM »

Mads...... You simply never fail to give good advice! I'll go with the cm300 config, as well as reading your guide. I figured since I have the spare time I might as well wind a coil. But I think to add my own flair I'll water cool the system, especially with aio kits being relatively cheap.

oneKone · « **Reply #3 on:** April 16, 2018, 05:23:22 AM »

Ok.... After a bit of thinking I'm going to go with the 110mm diameter by 440mm long secondary. I think I'll be too space limited to do full power runs with a bigger coil. But the plan of cm300du fullbridge is still going to happen!

Mads Barnkob · « **Reply #4 on:** April 17, 2018, 03:06:07 PM »

If you get too many ground strikes with a short secondary coil, you can always place the coil up on a table

It is only a problem if you end up with a coil that only makes topload to strike ring sparks, but with a good long break out point that is properly not an issue.

oneKone · « **Reply #5 on:** April 17, 2018, 03:31:42 PM »

literally the only thing i focused on " short secondary coil " (man my mind operates differently after gaming).

A large portion of my decision to change the secondary length was because i don't have enough wire, but i'm going to have to order more anyways to build a new pll sstc. So the big build is back on! realistically it's only about $130 price difference. Another thing that i'm thinking of changing is rather than going for cm300du-24 bricks i might go for the cm300du-12. part of me wants to try the Chinese bricks to see how far i can get, even though the sting of those ixgn60n60 still feels fresh!

I think this build will take a while. with bigger ticket items it's going to put put more of a strain on my project/hobby budget.

oneKone · « **Reply #6 on:** May 08, 2018, 02:37:46 PM »

First proper update on the build!

I've ordered 3 used cm300dy-24 bricks (thank you Mads for the link) and i wound the secondary. In the past i've been winding secondaries by hand and this has proven to be very time consuming. This time it was different! i held the form between centres and used a rotary buffer connected to a variac for speed adjustment. the coil took about an hour and a half to wind.

I've applied 5 coats of polyurethane and will apply about 5-8 coats (with sanding) of oil based polyurethane to get a nice finish and shine.

the coil sitting next to the mini drsstc:

oneKone · « **Reply #7 on:** May 15, 2018, 10:13:50 AM »

another small update!

I've received the cm300 bricks.... man they're big! I'm also just going to get a spun aluminium toroid, its kinda expensive but they look incredible. hopefully this weekend i'll get the driver assembled and tested.

Mads Barnkob · « **Reply #8 on:** May 16, 2018, 10:16:17 AM »

Beautiful secondary coil, how tall will the complete system be? Around some 1.5 meter?

Big is good! Almost impossible to destroy with that amount of silicon on a huge copper heat sink.

oneKone · « **Reply #9 on:** May 16, 2018, 03:48:06 PM »

Cheers,

My guess a the moment is about 1.2m. I'm still unsure if I'm going to go with air or liquid cooling. With air I could find some heatsinks on eBay or get some vapour chamber type CPU coolers, with water I was just going to go with 2 x140mm aio.

I'm unsure if I should buy the toroid first or buy the bus and tank capacitors and other parts to build the base.

Mads Barnkob · « **Reply #10 on:** May 17, 2018, 10:49:04 AM »

Water cooling is only necessary if you are going to push it very hard, long on-times, high bps and for long run times. Forced air cooling on a large heat sink will be good for most cases. If anything needs water cooling its usually the primary coil before the IGBTs, due to self induction heating of the inner turns.

Get the bus and tank capacitors first, there you have much more work to do and surprises to get, parts missing etc. A toroid is pretty much plug and play on top

oneKone · « **Reply #11 on:** May 18, 2018, 02:21:17 PM »

Cheers for the guidance. It's insane how much more expensive a larger coil is to build vs a small one. I'm guessing as long as the bus caps have a rms current rating of at least 15A they'll be ok?

As for the heatsink I think I'm just going to turn to eBay. There's nothing local that I know of, I'm thinking of this one:

Mads Barnkob · « **Reply #12 on:** May 18, 2018, 02:56:13 PM »

You can calculate the needed RMS rating from the math supplied here: http://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/dc-bus-capacitor/

That is pretty expensive for a new and relatively small heat sink, is there no industry / scrap yard near you where you can find a scrapped variable frequency drive and scavenge the heat sink from that?

How about local HAM / radio amateurs, they usually have a good stock of heat sinks too

oneKone · « **Reply #13 on:** May 19, 2018, 01:23:32 AM »

Cheers for the link, I'll have a look and hopefully order the bus caps Tonight.

Unfortunately in Perth there's not a lot around. Back in my earlier days of electronics the scrap yard would have random parts like that, these days there's almost nothing. I'll have a look at AliExpress.

This is the only suitable heatsink I have for them. They have a bit of overhang but main portion of the die is making contact.

Mads Barnkob · « **Reply #14 on:** May 22, 2018, 10:29:22 PM »

If the IGBT module can not fit on the surface it is not hard to deem that it is too small a heat sink

Large enough heat sinks are however pretty expensive from new, but it did spark me to start up on a heat sink and cooling chapter for the DRSSTC design guide, as I did have some notes here and there on selecting a heat sink and some measurements from Hydron.

If it wasn't so darn expensive to ship 5 kg half around the globe, I would have sent you a heat sink from my stack

oneKone · « **Reply #15 on:** May 24, 2018, 12:00:05 PM »

haha i knew i was dreaming about the heatsink! one thing i've thought of is just getting another and having them side by side with a module one each heatsink. Honestly cheers for the thought of the heatsink, i'll have a few months to think about my options...... which brings me to this,

I'm going to have to put the build on hold. i ended up cracking two teeth in my sleep, one half of one tooth was missing and the other is cracked! so now instead of buying the parts for a coil i'm going to have to save money for crowns.... such is life

oneKone · « **Reply #16 on:** May 28, 2018, 04:52:49 AM »

@Mads,
Assuming you could be bothered dealing with shipping a heatsink. I don't suppose you'd have a ballpark figure of what you'd want for the heatsink plus shipping?

Mads Barnkob · « **Reply #17 on:** May 29, 2018, 07:13:27 AM »

5 kg is something stupid like 75€ to ship to Australia, you could properly get a brand new from RS with that money. Maybe try to find a local aussie with some spare heatsinks on eevblog.com forum?

oneKone · « **Reply #18 on:** June 16, 2018, 09:45:09 AM »

you were correct mads, A heatsink from rs components isn't badly priced.

I think for the mmc i'm going to use 940C20P47K-F with 3 in series and 3 in parallel, for the heatsink i'm going to go with this: https://au.rs-online.com/web/p/heatsinks/1898404/ . its a bit off putting being only 30mm fin depth but with 2 x 140mm pwm pc fans i think it will stay under control. as for the bus capacitors i think i'll be going for 2 x 4700uf 450vdc in a voltage doubler configuration.

The other day i scored 2 x cm150du-24f bricks for $30 delivered. Initial tests show them to be working perfectly, so i think for a smaller coil for my next build these will be a goto.

edit:
i was looking at these capacitors and was wondering if anyone has seen them in a drsstc before?
https://au.rs-online.com/web/p/polypropylene-film-capacitors/1235134/

Mads Barnkob · « **Reply #19 on:** June 18, 2018, 11:13:39 AM »

Without doing the numbers, that MMC sounds way too small for a CM300 coil. The snubber capacitors are good pulse rated MKP capacitors and others have used them in a MMC before. Works just fine!

The heat sink will do good, noise from high air flow is not really a concern in a Tesla coil, so just give that plate plenty of air flow and the fin length will not matter, the superior cooling comes from high air flow and many fins!

DC bus capacitors sound fine, add some MKP snubbers aswell, across the IGBTs.

Neat with some smaller bricks, looking forward to drsstc4

High Voltage Forum · « **Reply #19 on:** June 18, 2018, 11:13:39 AM »

Author Topic: Drsstc 3 (Read 34546 times)

High Voltage Forum

Recent Topics and Posts