The Beginnings of a DRSSTC / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

The Beginnings of a DRSSTC

Move Thread

LAN_403

Avalanche

Sun Aug 04 2013, 04:45PM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

Hi all,

Here you see the beginnings of my first DRSSTC project started this weekend

So far the H bridge is assembled, so most of the metal work done.

This will be my first 'proper' coil, if all goes well. I thought it's about time I did something - the only previous experience I have with Tesla coils are some mini SSTCs I built back in 2006/07. I'm using whatever I can obtain easily for this coil, I have most things already (I bought the perspex tube back in 2005 to build an SSTC!)

Spec wise, nothing set in stone yet, as I'm still playing around with the various TC modelling programs on the web. From memory, (my notes are at work) the tube for the secondary is 130mm diameter, and I'm aiming for a winding length of 650mm, and ~2400 turns. Aiming for a resonant frequency of <120kHz without toroid, and hoping for 6-7' of spark. I'll put some proper specs up when I've spent a couple more lunchtimes thinking about it.

Layout was done in Visio, then cut out and stuck onto copper sheet for cutting and drilling. I like doing things this way, because it's easier than trying to figure it out in real time! And I can do the layout over lunchtime at work

The IGBTs are 1200V devices, so I should be good for up to a maximum 800V on the DC bus (limited by the capacitors, assuming the IGBTs stay within their SOA's at 800V)

I'm having to series some 400V 6800uF caps, which isn't ideal, but I've tried to keep the inductance as low as possible!

The insulation is two sheets of overhead transparency. As you can probably guess, I haven't tested it yet for breakdown!

Everything assembled

The capacitors seem to dominate - and it seems happier this way up... so I'll probably mount it this way up, then the output will be closer to the primary as well. I'm reluctant to have the heatsink too close to the primary though, due to massive eddy losses it will create!

Next up is to flash test the bus bar assembly, [rebuild the whole thing if it blows itself to bits

] and then design some seriously over-the-top gate drivers. More about that when I've ordered the parts and found out if my circuit works properly or not!

101111

Mon Aug 05 2013, 02:12AM

Registered Member #575 Joined: Sun Mar 11 2007, 04:00AM
Location: Norway
Posts: 263

That is one great looking bridge!
How thick is the bus bar?

Avalanche

Mon Aug 05 2013, 08:57AM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

Thanks

The copper is 1.6mm, so it's sturdy enough to hold the capacitors in place without bending.

I've dug out my notes, and the design I settled for is below:

Wire Diameter = 0.25mm
2600 Turns (winding length 650mm)
Fres with Toroid = 84kHz
Primary L = 16.8uH
Secondary L = 166.8mH
Mutual L = 665.1uH
Coupling K = 0.396

This is using a conical primary, made with 10mm copper tube, which is what they have in my local DIY store. I might downsize the tube to make it easier to bend! I'm thinking the coupling is too high in terms of flash-over, but the primary is conical to avoid flashover at the top. I assume most of the coupling is coming from the lower part of the primary so it might be ok? Feel free to chime in if anything looks way out

Steve Conner

Mon Aug 05 2013, 09:50AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Looks good, except the busbar design gives me the creeps.

How do you stop it from shorting out if one of the capacitors twists slightly? I'd suggest running some nylon screws right through both plates so they can't move relative to each other.

Avalanche

Mon Aug 05 2013, 07:43PM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

I was relying entirely upon the single screw to stop it twisting, but now you've mentioned it I think I will sort it out properly before trying to power it up as I still haven't dared try it

It's almost impossible to see the washer when it's all assembled so it might have already twisted out of alignment...

Steve Conner

Tue Aug 06 2013, 09:52AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I scrapped my previous H-bridge design completely for similar reasons, I couldn't get creepage and clearance distances that I was happy with. There's so much stored energy that if it arcs over it will probably blow a hole in itself.

Good luck! Those IGBTs look familiar, the ones I tested seem to perform very well and make a nice cup of tea.

The coils of 10mm copper pipe you can get from DIY stores should be in a soft annealed state. Much softer than the straight lengths of plumbing pipe and very easy to bend by hand as long as you get it right first time. The more you bend it back and forth, the more it "work hardens" until it is a complete mess. I'm going for 12mm for mine.

Avalanche

Mon Jun 23 2014, 07:57PM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

Design revision and a bit of progress!

I decided to change the primary to use 8mm copper tube - as I thought this was more in proportion to the size of the secondary tube that I already have.

I re-modeled the coil as follows:

2600 Turns (0.25mm wire)
fRes = 82.65kHz (with toroid)
Total Topload Capacitance (Ctop): 18.55713 pF
Primary Inductance (L1): 31.06074 uH
Secondary Inductance (L2): 161.91815 mH
Mutual Inductance (L1,L2): 484.40702 uH
Coupling Coefficient(K): 0.216001

It's going to be a medium-high impedance coil, but I'm a DRSSTC noob and just want something to play around with.

I did the design in Visio, which produced all the templates for cutting. The primary was a pain to wind. I had to practically shape it into the correct cone-spiral first, and then fit it onto the former!

The inverter uses Semikron SKM200GB128D's and my controller will be a PLL of some sort. Current feedback will be primary current feedback, using a zero-crossing detector and lots of experimentation (an area I've wanted to explore in the past). Inverter voltage feedback will be via a voltage transformer connected across the inverter output. I've been playing around building the meanest gate driver ever, using TO220 devices in the output stage which themselves are driven by fast gate drive ICs. One disadvantage of this method is that I limit myself to an output voltage of +15/-12V because the gate drive ICs have to share the same rails, but the reason I'm doing it this way is that I want to use the same driver to drive some very large traction IGBTs I have (they have about 50nF input capacitance).

Anyway, some pics... obviously the driver schematic isn't finished. Can't decide whether to drive the input with optoisolators or a tiny signal transformer.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob