oneTesla - a small single-board DRSSTC / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

oneTesla - a small single-board DRSSTC

first previous 2 3 4 5 next last

Move Thread

LAN_403

nixie

Sun Jul 29 2012, 03:06PM

Registered Member #3908 Joined: Tue May 24 2011, 09:40PM
Location: Gilbert, Arizona USA
Posts: 68

Hi There,
I bought one of the boards. And received the known items through Mouser/Digikey.

A few questions on materials and assembly. These may be common, and might help others wanting to build it.

(a) Heat Sink; Who is the supplier/Part number?

(b) GDT Core; a spec, or recommended supplier/part number?

(c) GDT windings; how many turns? I'm guessing 20 perhaps?

Thanks, Jeff

Dr. ISOTOP

Sun Jul 29 2012, 09:20PM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

nixie wrote ...

Hi There,
I bought one of the boards. And received the known items through Mouser/Digikey.

A few questions on materials and assembly. These may be common, and might help others wanting to build it.

(a) Heat Sink; Who is the supplier/Part number?

(b) GDT Core; a spec, or recommended supplier/part number?

(c) GDT windings; how many turns? I'm guessing 20 perhaps?

Thanks, Jeff

1) Mine are currently pulled from ATX power supplies; I'm still seeking a supplier that will do small quantities.
2) I used a Sanlin SL5 core about an inch in diameter
3) 15-20 turns seems about right.

Ben Solon

Sun Jul 29 2012, 11:55PM

Registered Member #3900 Joined: Thu May 19 2011, 08:28PM
Location:
Posts: 600

Dr. ISOTOP wrote ...

Heatsink USA will sell relatively cheap heatsink by the inch^2, but I don't know if they do price breaks. They will also drill and tap for you(probably extra).

Really, any ferrite with a high permeability and good transfer at the desired frequency will work. Then just calculate the # of turns=(V*T)/(B*Ae).

I just may have to make/buy one of these myself after my big drsstc is done! Good work bwang and Kramnik, this is an interesting project you've got going here.

Dr. ISOTOP

Mon Jul 30 2012, 03:06AM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

ben123324 wrote ...

Heatsink USA will sell relatively cheap heatsink by the inch^2, but I don't know if they do price breaks. They will also drill and tap for you(probably extra).

It needs to be the right shape of extrusion (for board-mounting).
Really, the cheapest way might just be getting surplus Dell PSU's off eBay and gutting them

Killa-X

Mon Jul 30 2012, 05:01AM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

I got my heatsinks off ebay, had to really hack them up to fit.
I use the ferrites electronic goldmine USE to sell.
I'm using 8 turns, which is my typical DRSSTC setup.

Seems to work great, may do a 120V test with IGBTs today, get more sparks you know? :D

Dr. ISOTOP

Tue Jul 31 2012, 10:36PM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

So Killa-X told me via PM that the IGBT's died at 300A doing burst mode operation. Kramnik also killed a bridge doing music with really long pulsewidths, so I guess the conclusion to be drawn here is that the coil doesn't like super-high duty cycles.

Ben Solon

Tue Jul 31 2012, 10:54PM

Registered Member #3900 Joined: Thu May 19 2011, 08:28PM
Location:
Posts: 600

Raise be impedence of the system a couple times. I would like to see the results of that. It might not increase spark length or primary current, but the longer on-times might make it a bit more user friendly for people who buy one without the knowledge of exactly how they work.

What exactly have the limits been? 300A by Kramnik, has anyone else blown a bridge and measured the conditions? The irg4pc50Ud might be a decent upgrade...

Killa-X

Tue Jul 31 2012, 11:08PM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

I had to been running 50 primary current cycles...quiet a bit heh...plus i was burst-mode.

Once i replace the bridge later on i will leave everything the same and attempt to count the cycles..and measure the on vs off time, of when my coil went pop..

Most DRSSTC bridges have a snubber capacitor on them, is it not required on yours for some reason..?

976UNl

not to mention i had 1 flash over before the video, i lifted the coil up by an inch, fixed the issues.

Dr. ISOTOP

Tue Jul 31 2012, 11:16PM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

ben123324 wrote ...

Raise be impedence of the system a couple times. I would like to see the results of that. It might not increase spark length or primary current, but the longer on-times might make it a bit more user friendly for people who buy one without the knowledge of exactly how they work.

What exactly have the limits been? 300A by Kramnik, has anyone else blown a bridge and measured the conditions? The irg4pc50Ud might be a decent upgrade...

IMO high-impedance primaries are actually less user-friendly - they require long on-times and run the transistors hotter.
I haven't tried the IRG4PC50UD yet...they seem roughly equivalent to the 60N60's I've been using and have a higher thermal impedance, plus they are a bit slower.

Killa-X wrote ...

I had to been running 50 primary current cycles...quiet a bit heh...plus i was burst-mode.

not to mention i had 1 flash over before the video, i lifted the coil up by an inch, fixed the issues.

Wow, nice work. 50 cycles is a lot...that's a couple hundred microseconds at your resonant frequency right? If you were variac'ing the bridge while keeping pulsewidths and such constant, have you tried holding the bridge at a constant voltage and raising pulsewidth until you find performance satisfactory?
My original excuse for not using a snubber was that i didn't have the board space for one (back when the board was etched on to 4x6" PCB stock from eBay). The current design uses a super-low-inductance laminated bus, so I haven't really found a need for a film cap yet. A film cap will most likely improve performance a bit - you can try soldering one directly to the capacitor terminals if you want.
As an FYI for everyone who has bought a board so far: the bleeders are meant to be soldered directly to the buscap terminals on the bottom of the board! Use bleeders!

Killa-X

Tue Jul 31 2012, 11:53PM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

I can see, i have a few caps but they maybe too small for snubber use. Its a 400V 1uF capacitor, 0.007ohm ESR.

I only messed with the duty cycle for making the arcs have some meat to them, and not be thin strains of hair, and its roughly around that 200Khz zone.

first previous 2 3 4 5 next last

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