[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: A LUA file for BELA (a help to build electrostatic model of TC)



Original poster: "Mike" <induction@xxxxxxxxxxx>

Hi Finn,
I sent some things off list for speed. Also you can find some great Java models, including electrostatic at http://web.mit.edu/8.02t/www/802TEAL3D/ a lot of good work in making these many types of models. I have had a couple of requests for the pictures / data on the RF transformers so I took what I had here and zipped the pictures into one file www.hot-streamer.com/mike2004/RF-XMFR.zip so they are available
there, a total about half a meg, not bad.
I don't have an IGBT machine here (yet) but there are some secondary coils and various balls up to 3 feet diameter, left over from small prior coils, maybe I'll hook up the 5 kW Lepel induction heater (~ 450 KC), dump the output into a 15 to 1 RF transformer, using it backwards, wire one end of the 15 turn secondary to ground, the other to the base of the TC secondary and tune the coil with top load. Could also simply dump a big pulse cap into the primary but not easy to hear or see the discharge of secondary against the flash and noise of the cap dump for a one shot. I've driven coils before with a 2.5 Lepel but not through a transformer, this should be better, though it will be CW. If I did have an existing IGBT driver and system, I would be hooking up a one turn (slotted) wide sheet primary right now. Maybe one of the IGBT equipped people will beat everybody to the punch, after all they can quick pulse test the thing. What OD or ID are you thinking to use, how high a primary and how many turns within it are you going to pot. I'll get the potting material access name to you Monday when I get back to work.
You can see some more at http://www.eastcoastind.com/products4.asp
and if you click on "view list" you can see different models.
The 8:1 I spoke of data is at http://www.eastcoastind.com/products4_model.asp?Mode=E169 IMHO a potted RF transformer run backwards with it's secondary grounded on one end and the other wired to the TC secondary would drive it really nice. If the secondary wire with more turns per length were in the Potted even better.
Mike





----- Original Message ----- From: "Tesla list" <tesla@xxxxxxxxxx>
To: <tesla@xxxxxxxxxx>
Sent: Saturday, December 17, 2005 8:14 PM
Subject: Re: A LUA file for BELA (a help to build electrostatic model of TC)


Original poster: Finn Hammer <f-h@xxxx>

Mike.

What you are writing is sweet music in my ears. Is it possible to access pictures, drawings or even samples s,ince I could save myself a lot of experimentation that way.
Even just a lead to that flexible rubbery potting material would be nice.

Cheers, Finn Hammer

Tesla list wrote:

Original poster: "Mike" <induction@xxxxxxxxxxx>

Hi Terry, List,
Well, if I am following this line of logic about tight / high coupling or LERT right for the DRSSTC, why are people not using the method that RF induction heating output RF transformers, very well proven, use. I can see where using an iron core is a problem on the RF units but using a copper sheet secondary works great. At starting frequencies of 450 kHz and down into the 150 kHz range, this system runs hundreds of kW all 3 shifts in the CW mode. So, in the DRSSTC, you could run it backwards. Your one turn copper sheet with slot becomes the primary and of course potted between the primary and secondary. We use, build and fix these things (when a line breaks outside and the water shorts it all out) so we sort of do what Finn is talking about, only we usually make the primary from tubing 18 through 8 turns on the inside of copper shell /secondary and the copper sheet with vertical slot, with connection block and water cooling, drives the 1 or 2 turn work coil for pipe welding, etc. Regardless if the machine is a tube or IGBT unit running in the 200 kHz range, the current is stepped up by the turns ratio and the voltage of course stepped down by that same ratio. The "secondary" can be from 4 inches high (two slightly spaced and stacked for a N:2 ratio) or 8 through 18 or more inches high, again of course with that slot to make two live ends. We even place a tab halfway on the sheet for earth ground connection. These are tightly coupled, 3/16 through 5/8 inch primary to secondary and all in the vacuum pumped potting rubber material. RF transformers work fine with the primary on the inside of the shell and they are also done with the primary on the outside, though this tends to make less room for the work coil connections, so it's usually inside. I've run them backwards before, works great. I've also dumped large HV caps into the 15 turn primary (tubing) and got a 15 fold current gain on the shell /copper sheet secondary. Also done it backwards, dumped a large HV cap into the one turn copper slotted "now backwards primary" and saw one impressive voltage gain off the 15 turns. Watching the discharge on a scope, even a cap dump had no rise time issues in the typical TC frequency range. I found the noise from the 25 kV 100 MFD (pulse rated) cap spark gap less than desirable but the RF transformer, a stock commercial shell in the 300 kW class, worked fine. So, why can't you use this wide one or two turn sheet as your DRSSTC tightly coupled primary?
Can't get much tighter than that.
Scoped ring time through the transformer was about the same as a shorted to cap body spark gap shot though I think the ~ 31,225 J blast harsh on the ears. So why can't an IGBT in TC service pump such a wide and tight coupled primary, after all, IGBT's do it all the time in induction machines, just driven CW Vs pulse.
Normal or backwards, a transformer is a transformer.
Mike


----- Original Message ----- From: "Tesla list" <tesla@xxxxxxxxxx>
To: <tesla@xxxxxxxxxx>
Sent: Friday, December 16, 2005 9:39 PM
Subject: Re: A LUA file for BELA (a help to build electrostatic model of TC)


Original poster: Terry Fritz <vardin@xxxxxxxxxxxxxxxxxxxxxxx>

Hi DC,

They are trying to get high coupling with DRSSTCs to reduce the number of cycles and thus time it takes to bring the secondary voltage up. This helps to decrease the load on the IGBTs and gets a better LERT.

LERT is Load Energy Rise Time. The DRSSTC folks have found that there is an optimal time to energized the secondary to get the best sparks. This time is fairly fast and demands a higher coupling.

ScanTesla (the computer program) says that coupling otherwise does not matter too much and couplings like 0.13 might be just fine. But that is before this fast rise time effect is taken into account. ScanTesla can calculate LERT, but it does not really judge what is good or bad. LERT is a little to unknown right now to make "equations" but it seems to be a real important thing.

LERT might also be why coils with low surge impedance seem to do well. We really don't know right now what it all means...

Cheers,

        Terry
[Snip]