2kW DRSSTC Converter / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

2kW DRSSTC Converter

previous 1 2 3 next

Move Thread

LAN_403

HV Enthusiast

Thu Aug 03 2006, 11:40AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Thanks for all the info. Fortunately, I have been designing these types of power supplies for more than six years now for military and space applications, so I am up on all the nuances, and do's and don'ts.

As far as thermal interface goes, I use simple grafoil which is considerably better than any of the phase-change materials.

I'm not a big fan of phase change material and don't like the stuff that much. Whenever you go to remove a supply, it always leaves half the stuff on the coldplate, and the other half on the supply. Not the easiest stuff to remove or clean up.

robert

Thu Aug 03 2006, 12:30PM

Registered Member #188 Joined: Thu Feb 16 2006, 05:18PM
Location:
Posts: 67

Well, i was doing assembly of lets say sensitive electronics at some company (which i shouldnt realy talk about too much)...
All stuff had to be done in cleanroom environment and it had to be _perfect_.
The modules assembled with grease (luckily) were only demo units for testing, the real ones were assembled with phasechange film.
The module surface was milled flat (lets say, real flat, perfect mirror finish, shining a laser beam at it makes no visible spot viewed from the side) as was the mating heat-exchanger surface.Both had to be the same aluminium alloy for thermal expansion reasons.

Actually, i bought some of the modules that didnt pass the incoming inspction (eg had a scratch or a bubble under the label) and im planning to use them for some projects too.And im getting the right thermal film too.No way im using grease.

HV Enthusiast

Thu Aug 03 2006, 12:54PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Lets please keep this discussion on topic and not debate techniques which for obvious and economic reasons will not be employed in a low-cost commercial / amateur application which is being discussed here. These supplies will be used in low-cost commercial assemblies and will not be launched with the latest multi-billion dollar satellite. Thanks.

Steve Conner

Thu Aug 03 2006, 01:28PM

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

Aww. I was getting all excited for the launch of the DRSSTC Twin Killer Satellites sponsored by Lockheed Martin

Anyway, to keep on topic. From what you posted, I'm guessing that you got a commercial switched-mode power supply brick that used a resonant design, removed the step-down transformer and secondary rectifier, replaced the switching devices with those IXYS fast IGBTs, hacked the protection circuitry to boost the peak current limit, and hooked the output of the inverter up to the old DC output pins?

I like it

HV Enthusiast

Thu Aug 03 2006, 03:48PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

wrote ...

Anyway, to keep on topic. From what you posted, I'm guessing that you got a commercial switched-mode power supply brick that used a resonant design, removed the step-down transformer and secondary rectifier, replaced the switching devices with those IXYS fast IGBTs, hacked the protection circuitry to boost the peak current limit, and hooked the output of the inverter up to the old DC output pins?

You guessed wrong. The design is custom from the ground up. :P

Steve Conner

Thu Aug 03 2006, 11:42PM

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

Wow. I don't know what to say. I mean, it's really impressive, but how in the world are you going to recoup the development and tooling costs?

My PLL driver system is made with 1970s style through hole components and needs a fair amount of hand assembly. (The end user is expected to make their own H-bridge.) It looks pretty big and crude in comparison to that nice brick thing. But I think I should have made it even cruder and nastier if I'd wanted to sell lots of them

I'm adding modules to it and hopefully I'll end up with a kind of high frequency resonant lego set that can be used to make anything from a small induction heater to a giant DRSSTC running off 480V three-phase power.

Terry Fritz

Tue Aug 15 2006, 11:58PM

Registered Member #393 Joined: Tue Apr 18 2006, 12:30AM
Location:
Posts: 297

Oh My!!

Dave just posted this to the other list...

Looks like IR put the DRSSTC driverer on a "chip"!!

Data Sheet here:

Cheers,

Terry

HV Enthusiast

Wed Aug 16 2006, 01:32AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Thanks for posting. Looks good, but doesn't look the onboard gate drivers can really do much as far as driving big devices. Might be good for a micro coil. Also, the max useable (at least advertised) frequency is 100kHz, which kind of opposes the micro coil route.

Terry Fritz

Thu Aug 17 2006, 03:28AM

Registered Member #393 Joined: Tue Apr 18 2006, 12:30AM
Location:
Posts: 297

Yeah....

The current drive sucks...

But the technology "is there now"... I'll give IR a "little" time to drive that current up

))

But it is a "leap forward" in the H-drive situation... We can always "tickle" the RC timing to sync with the primary current easily :o)))

I wish I was still in the industry in a position to just "tell" them what to do next :o)))

Cheers,

Terry

Terry Fritz

Sat Aug 19 2006, 10:22PM

Registered Member #393 Joined: Tue Apr 18 2006, 12:30AM
Location:
Posts: 297

Hi,

Half bridge with 250mA current drive:

But what is really cool is check out the "load" they are using now

I don't know if they have a specific commerical application in mind or if the "coilers" that talk to IR design are "getting their way"

Chip designers struggle to find the coolest problems to solve and make chips for. "We" bend their ears whenever we can

) It worked for MMC caps

I have my "order" in for a 10 amp 30V drive version at 250kHz to be designed next

Aparently, the present chips do higher frequencies with no problem at all... Modern IGBTs can run pretty hard off just 15 volts of gate drive. But that 250mA is a big problem. It can be amplified as discused here:

But best if they just make it all on the chip too

) Maybe sometihg bigger than a SOIC package too

Worth the wait

Any updates on Dan's "real and now" DRSSTC driver ??

Cheers,

Terry

previous 1 2 3 next

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