Drsstc 3

[oneKone]

time to get started on this project again!

in the time since i last posted i haven't achieved much. the bridge and the mmc are pretty much completed with some minor tweaks needed.

For the bridge i ended up using 8mm aluminium plate for the bus bars which has an added bonus of making the assembly more rigid, although i don't know if its a wise decision to be relying on the terminals of the bricks to take the load of the capacitors?  As for the mmc it was pretty simple. 3mm x 15mm coper flat bar was used along with m6 cap heads to keep everything in place.

[oneKone]

For about a month now I've been put off on this project, I really couldn't be bothered making any progress and I couldn't figure out why..... Then it hit me! I hate the look of it 😢

Basically I had a 450x450x350mm "cube" for the base of the coil, to me that's pretty unappealing. One look I do like are round bases both a helical primary. Apart from Mads drsstc 1 I haven't really seen bigger drsstc with helical primaries at all, unless having a pancake primary was more of the norm 5-10 years ago?

So my question is, are there any drawbacks from having a helical primary at this size I can't think of?

[Mads Barnkob]

For about a month now I've been put off on this project, I really couldn't be bothered making any progress and I couldn't figure out why..... Then it hit me! I hate the look of it 😢

Basically I had a 450x450x350mm "cube" for the base of the coil, to me that's pretty unappealing. One look I do like are round bases both a helical primary. Apart from Mads drsstc 1 I haven't really seen bigger drsstc with helical primaries at all, unless having a pancake primary was more of the norm 5-10 years ago?

So my question is, are there any drawbacks from having a helical primary at this size I can't think of?

Good old Dr. Spark, which seems to have closed down his website, primarily built helical primary coils and they performed great!

Now I built a helical coil on my first DRSSTC for several reasons:
- To make the coil take up less space, smaller diameter
- I thought it was easier to make primary coil holders for this type compared to spiral
- Tapping was easier with cable just going up the side

I will avoid helical coils in the future for the following reasons:
- It has a high coupling that is harder to adjust/control
- It is much closer to the topload, so there is more strikes to ground rail
- It was a pain to wind the coil through the 4 sheets with too small holes, it took me hours to move it centimeters at a time as it kept locking itself in place
- There is no performance gain over spiral, for a DRSSTC

I would however still make round platforms with spiral primary coils, as I think they look great, but when you get up in size you really need a bigger base to stabilize the thing with wheels out in the corners of a box.

You could also go inbetween spiral and helical and make a upsidedown U shape primary if you want it to have a smaller diameter overall.

Looking forward to see what you decide on

[oneKone]

Cheers for the guidance Mads. Is having a upsidedown U primary the same calculation as a flat pancake?  I had a quick look at your drsstc that uses one, looks extremely nice.

[Mads Barnkob]

Cheers for the guidance Mads. Is having a upsidedown U primary the same calculation as a flat pancake?  I had a quick look at your drsstc that uses one, looks extremely nice.

I think I just went with a median value based on spiral coil and coupling as if it was a helical, but as you can see from the article on my website, I did a lot of measurements instead, as I could not calculate its specifications.

I guess you have to use something like FEMM in order to get a correct simulation as spiral/helical calculators and javatc does not support that primary form.

[oneKone]

i installed femm and straight away i was overwhelmed, i'm just too basic! so i'm going to do what i love doing.... trial and error!

i made small progress by cutting out the new base for the coil, i like it so much more than the square frame i put together! for the primary i've drawn up a basic shape in aspire and have given roughly 10mm spacing. at this point i'm trying to decide if i should have my mmc at 330nf or 165nf. It's been a while since i've had to recall but the lower capacitance means higher number of primary turns which would result in a lower achievable apk? i will have to re-read a few threads again to get my understanding back!

this is what the coil looks like now and the proposed primary, which is 50mm high and 100mm long. on the weekend i'll make a better effort to get some more in depth progress photos.

   

[Mads Barnkob]

Looking good

I guess you can get away with just estimating that as a flat primary in JavaTC and it will not be too far off.

The higher MMC capacitance, the lower system impedance to a given resonant frequency. So yes that means higher peak currents, shorter on-time due to faster current ramp up etc. But it also gets more expensive to build due to increased cost on parts that can withstand the higher voltage/current. Higher impedance is cheaper to build but comes at the expense of greater switching losses due to longer on-time needed to produce long sparks.

[oneKone]

unfortunately I've been dormant for quite some time! other projects got in the way of dedicated coiling but now they're over i can get back into the swing of things!   

in my last update i made a mdf base and went with plywood primary supports. At the time i knew the supports weren't the best idea and i still wasn't happy with the base... it was time i changed the base again! which also brings me to distraction number one, a new cnc router! to me its a tool that i needed for many of my hobbies and will serve a great deal in coiling.

i wish i took photos during the design, cad and cutting but i didn't think ahead... i redesigned the base to be hex and made it out of plywood, i also made the primary supports out of polyethylene. I think once the base has been given a light stain the white from the mmc and primary supports as well as the primary and secondary should be appealing to the eye, not the most critical part of a drsstc but a part none the less. during the week i should have the coil wired up to begin testing, which is extremely overdue.




 

[oneKone]

Not that much has happened. A simple coupler between the primary and bridge output was made, at least with this info ever need to I can remove the primary easily. I think I also fou a place for the feedback/OCD CTS. I think I'll just make a simple bracket to remove any mechanical load. One thing that I'm.not sure about is if I have the gdt leads too long. I guess when I start testing I can always check the gate waveforms?

[Mads Barnkob]

It looks very nice! Now that you went for the hexagon base, did you see the covering plates in the thread with the twin system? Where they sit into slots in the supports.

You should place your CTs at either output from the bridge, never between the primary coil and MMC, as you have a high voltage ringing between L and C.

Small progress is better than no progress! I am currently in a period of no progress, mostly due to starting on a new job and that takes all my mental capacity

[oneKone]

I love the idea of the covering plates and the simplicity of the wingnuts, I'm still unsure if I'm going to cover the electronics on mine though.

I think I have little time! I couldn't imagine your workload!

I got some light testing in tonight. One thing that I can't remember is switching times.. I'm going to have to ready your write-up again mads. Basically I'm getting turn on/off times of 1uS (1000ns?), Once I get better educated on switching times I can see if/what I'll have to change with the gdt. I'm also getting 5v overshoot, to me.this seems acceptable, but again it's something I'll have to read up on.

Another factor is that these tests were done without a topload so it will be interesting to see.if the lower frequency helps with the switching times.

Picture of gate waveforms for reference:

[Mads Barnkob]

You could properly need a slightly higher gate resistance to damp the overshoot, but generally that is a near perfect driving waveform for a IGBT gate. 10% overshoot would be perfect, 20% is borderline to too much, but okay!

[oneKone]

Cheers for.the suggestion mads, I might try add a resistor in series with the primary of the gdt, in the past it's worked well so hopefully it works here.

I've had some time this weekend to work on the coil. I rewound the inductor as it was off the mark for what I needed, I also made the (but not finished) the topload.... As well as first light! At the moment i haven't made any measurements besides setting the phase lead.

While running up.the coil to about 1/2the input voltage it struck my router and killed a stepper drive, so all tests will now be done outside.. pictures!

[Mads Barnkob]

Congratulations on first light!

A real shame about killing a stepper driver on your CNC router, get some metallic net/chicken fence and throw over it, grounded and it should be safe

Your topload looks really cool, Russian design inspired I see, but the pieces of pipe covering 180 degrees seems to be too little when you have a sharp edge on the "outside" of the virtual toroid, so you have a break out elsewhere than from your breakout point. You properly need another 20-30 degree inwards in each end, so you need 240 degree pieces of pipe.

[oneKone]

Congratulations on first light!

A real shame about killing a stepper driver on your CNC router, get some metallic net/chicken fence and throw over it, grounded and it should be safe

Your topload looks really cool, Russian design inspired I see, but the pieces of pipe covering 180 degrees seems to be too little when you have a sharp edge on the "outside" of the virtual toroid, so you have a break out elsewhere than from your breakout point. You properly need another 20-30 degree inwards in each end, so you need 240 degree pieces of pipe.

Cheers, and yeah.. I've learnt my lesson with bigger coils now!

As for the topload I'm hoping I can cut out another radius and just bend it by hand with the t-pieces soldered on. I know it was another group that first put the topload out there, but it was Finn hammer that inspired me to do it.

In other news I ran the coil successfully tonight! Managed 1.6m streamers even with the topload situation. MMC was cool to the touch. The bridge was ever so slightly warm so I might just add active cooling just to be safe.

[oneKone]

Small update. (Basically just to remind myself for the future!)

I got a 20a capable outlet installed for possible future builds, but this also means I can easily run a a higher current limit. I've now verified the bridge with 900a @ 100us on time. One major thing I must fix is the toroid. I'm going to cut out a concave radius and try to push each segment in to try and form a new radius, if that fails I'll have to unsolder the joins. Hopefully this weekend I'll get some decent pictures of my setup and some arcs.... I'm truly Hoping for near 2m

[fnordest]

Wow! Really nice, those sparks look amazing and the hexagonal base suites your coil very well

[oneKone]

Wow! Really nice, those sparks look amazing and the hexagonal base suites your coil very well

Cheers,
 I'm pretty happy I went with a hexagonal base, definitely something I'll be using on future coils.

Another small update;
The ixys 40A 1200v bridge rectifier is dead! I ran the coil Hoping to get a few pictures and a video for the forum but unfortunately that couldn't happen. I'm not sure as to why it failed as the heatsink it's mounted to was barely luke warm (same heatsink as igbts). I've ordered a 150a 1600v bridge rectifier to replace it. When the fuse first blew I thought it was the bridge itself so I'm taking this as a win!

[Mads Barnkob]

I must admit that I never put much thought into the bridge rectifers I have used, since they are mostly just grossly overrated, but when I have used some at a more proper rating, they have went up in smoke a few times. Just a quick guess that they might not the be fastest diodes and have a reverse recovery time that is too slow for DRSSTC transients?

You mentioned that your IGBTs heat sink got slightly warm on the first light run, how about your primary coil?

[oneKone]

Tbh I always just treated bridge rectifiers as "basic linear" devices, but I think you're onto it about transients. But there is also one more thing, the primary was struck by a streamer..... It actively avoided the strike ring!

As for the igbt heatsink getting warm, it's simply not good enough of me saying "slightly warm" or "barely luke warm". I'll get a probe on both next time I run it.