Starting on my First (official) DRSSTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Starting on my First (official) DRSSTC

1 2 next

Move Thread

LAN_403

SparkLabs

Mon Oct 05 2009, 02:11AM

Registered Member #2418 Joined: Mon Oct 05 2009, 12:24AM
Location: Indianapolis, Indiana U.S.A.
Posts: 11

Hi all, This is my very first post and I have a few questions before I atttempt to build my first (official) Tesla Coil. I have already built tesla coils before, but all of them were just merely prototypes and I never really attempted to build a powerful one, but now I feel that I'm ready to build a decent Tesla Coil for once. I have already designed the driver, designed the controller, and built the actual Tesla Coil (primary and secondary). I have attached an image of the layout of the H-Bridge that I am going to build soon.

Is it considered OK to parallel 2 of the same IGBTs?
Would this driver setup have a very low stray inductance?
Is there any concern for the specifications I posted?

Design Specifications:
My Tesla Coil's Resonant Frequency is 467Khz which is a bit high for a DRSSTC.
Secondary turns: approximately 1500 turns.
Secondary wire gauge: 40AWG.
Secondary Coil Length: 5 Inches.
Secondary Coil Diameter: 2 Inches
Topload Capacitance: approximately 2.4pF.
Primary Coil Turns: 3 turns.
Primary Coil Wire: 10x 26AWG wire wrapped steel cable style.
Primary Capacitor: 940 Series 0.15uF 1.6KV
Coupling Factor: Approximately 0.2

Voltage to H-Bridge: 120-240VDC no more than 300VDC.
H-Bridge Capacitors: 1x 150uF 330V Photo Flash Capacitor and 1x 2.2uF Film Capacitor.

Power Supply: 50W 12V to 150V Homemade Inverter.
I am not planning to operate directly from the mains with this Tesla Coil.

I am planning on using the D2PAK versions of these HGTP3N60A4D IGBTS for the driver. These IGBTs are the best ones I can find for my small Tesla Coil. I have successfully accomplished 4" sparks with a prototype driver made with a full bridge of these IGBTS, but I am now trying to make a more powerful version capable of 6 inches or better. I am going to provide each IGBT with its own gate resistor and protection diodes, but the collectors and emitters are paralled.

I don't have any images of my old setup because I had no way to take pictures, but now I have a digital camera so I will provide pictures in the future. Any advice will be helpful since I am basically starting from scratch.

Jason

Z28Fistergod

Mon Oct 05 2009, 02:49AM

Registered Member #2040 Joined: Fri Mar 20 2009, 10:13PM
Location: Fairfax VA
Posts: 180

Hey there, welcome.

467 KHz is pretty high. I didn't bother looking up the datasheet for your igbt, but I've never read about any that could work effectively at a frequency that high.

Because the DRSSTC is switched at the resonant frequency, and hopefully when the current through the circuit is passing through zero, there shouldn't be any problems with current sharing during the turn on period. That being said I would avoid paralleling IGBTs.

I'm sure you've heard of KISS. It's not just a design guideline, it's a rule to live by! Sure, you could do it, and probably make it work well, but I think it would be easier to use four instead of eight. That's half as many to replace if the whole bridge eats it.

I don't know what the power dissipation is of that package but I would think it is lower that some of the bigger TO-247 sized packages, maybe you should think about that also.

Good luck and let us know how it's coming along.

SparkLabs

Mon Oct 05 2009, 03:08AM

Registered Member #2418 Joined: Mon Oct 05 2009, 12:24AM
Location: Indianapolis, Indiana U.S.A.
Posts: 11

Hi, Thanks for the info. These IGBTS do have lower power dissipation than a TO247. I am not concerened with power dissipation in the bridge because I am not going to operate this TC from the mains. I am going to power the bridge using one of my homemade 12V to 150V 50W inverters. (I modified the post above to show PSU info.)

Yes it would be easier to use 4 IGBTS, but my resoning is that by paralling 2 IGBTS, I can double the current handling for my Full Bridge. If you look at the datasheet of my IGBT, you would see why I would like to do this. (44A pulsed rating is uncomfotably low for me)

Actually, I am not familiar with KISS.

Jason

Mads Barnkob

Mon Oct 05 2009, 04:43AM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

KISS explained:

What is your secondary diameter? I agree with Dalus that you should aim for a lower resonance frequency, I think one of the highests I know about is Wards 300kHz DRSSTC.

Paralleling IGBTs does not give you double current handling, but gives you the ability to drive them right to the edge of one IGBTs ratings, and a little above, due to uneven sharing you will never achieve double current rating without a complicated driver and control network.

As its your first DRSSTC, you should cut down on the complexity and go for something really basic.

So keep it simple, stupid. :)

Wolfram

Mon Oct 05 2009, 07:53AM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

Check this out

if you haven't already. I don't think you need to parallel IGBTs to get enough power for your coil, just go with a fullbridge.

A.M.

SparkLabs

Mon Oct 05 2009, 01:50PM

Registered Member #2418 Joined: Mon Oct 05 2009, 12:24AM
Location: Indianapolis, Indiana U.S.A.
Posts: 11

Thanks for the links guys.

So looks like I just need to 4 IGBTs instead of 8 IGBTs. Ok, that will do. So looks like I need to design the power supply to output 240V instead of 150V to get the longer sparks that I need.
I have already mentioned in my first post that I have already built a prototype that generated 4 inch sparks using the exact same Tesla Coil above. (DRSSTC) So I have experience with this setup. This is not my first DRSSTC that I have built, but it is the first DRSSTC that I am "officialy" building. (If you know what I mean..)

The Secondary Coil is about 2 Inches diameter. Also I incorrectly specified the height so I changed the specified height on my first post to the correct value now. (it is 5 Inches)

Jason

SparkLabs

Thu Oct 08 2009, 04:22AM

Registered Member #2418 Joined: Mon Oct 05 2009, 12:24AM
Location: Indianapolis, Indiana U.S.A.
Posts: 11

Hi all, I got my birthday present for this year. I got a box of parts from Digi-Key to build my new Tesla Coil Driver!

I got these following parts:
20x of the 3N60A4D IGBTs. 10 of them are the D2Pak versions
1x of 940C .15uF 1600VDC capacitor
1x 2mil 3x4.5" Unetched PC Board
25x of the MUR105 fast recovery diodes.
20x of the ultra fast recovery 1N4150 diodes
10x Bipolar ESD diodes rated for 30V
10x of 390V 600A Pk TVS
10x of 20 Ohm non-nductive wirewound resistors
5x of the HEF4046 PLL Chip

With a grand total of $55! I will post pictures as soon I can of the construction of my Tesla Coil, but I will very busy this week so I will not be able to update frequently. Any questions or comments about my current setup will be apreciated.

-Jason

Z28Fistergod

Thu Oct 08 2009, 12:41PM

Registered Member #2040 Joined: Fri Mar 20 2009, 10:13PM
Location: Fairfax VA
Posts: 180

So your planning on going the PLL route with your driver?

SparkLabs

Thu Oct 08 2009, 09:19PM

Registered Member #2418 Joined: Mon Oct 05 2009, 12:24AM
Location: Indianapolis, Indiana U.S.A.
Posts: 11

Yes I am going to use a PLL just like the one I used in my prototype coil controller, but this time I am going to use a CT on primary wire instead of a CT on the secondary wire. I can't wait to see how the new Extremely Low Inductance driver performs! I have to build it first, but it has to wait until later this week.

Jason

Z28Fistergod

Thu Oct 08 2009, 10:35PM

Registered Member #2040 Joined: Fri Mar 20 2009, 10:13PM
Location: Fairfax VA
Posts: 180

The only problem with the PLL is that it probably wont get a chance to lock onto the resonant frequency during a single burst. Keep in mind there are usually only a few (about ten or so) RF oscillations per burst.

You might say that you will tune the PLL oscillator so close to the resonant frequency that it wont have to adjust very far, and it will be able to lock on quickly. Well, if you are using the type 1 phase detector, there is a 90 degree phase shift between the incoming signal and the output, at the center frequency. The phase shift is exactly what we don't want.

The type two phase detector doesn't have this phase shift, maybe you could design your circuit using it.

An easier, tried and tested method is Steve Ward's driver, or some variation thereof. It uses a buffer to convert the analog CT signal to digital, and drive the bridge with that. In this way the only thing that determines the drive frequency is the actual resonant frequency of the primary and MMC. Meaning that your drive circuit is self tuning, and it soft switches the bridge during the current zeros. This makes the whole system much more robust, and easier to tune.

1 2 next

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