[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: "D.C. Cox" <resonance@xxxxxxxxxx>




With all the accumulated knowledge of experimenters (thanks to Chip & Terry's great Tesla list) the math reverts to simple algebra and even less --- just simple arithmetic!

Step 1: Determine the approx dia coilform you would like to use. Example: 6 inches.

Step 2: Multiply the outside dia x 4.5 to get the correct winding length for the coil. Allow 1 extra inch on top and usually 2-3 extra inches on the bottom to obtain total coilform length. In this example, 6.3 x 4.5 = 28.35 inches, so let's round it up to 29 inches.

Step 3: Divide 29 inches by 1,400 turns (a great and very effective no. of turns for almost all Tesla sec coils). 29/1000 = .029 inches which is 29 mils. This is the correct outside diameter to determine the AWG of wire to use. With very large coils, ie, 20 inch plus coilform diameters, we prefer to use 1,200 turns as an effective value.

At around 2,000 turns the resistance starts detracting from the performance of the system so stick with 1,400 and you won't go wrong.

Step 4: From any Beldon (or other) wire distributor, obtain a chart that shows the ODs for Heavy Polythermalize magnet wire (this is 200 degree C. wire and is the preferred type to use due to its excellent extra insulation build --- prevents breakdown when pushed a bit). Look up the .029 inch od on the chart and read off the correct magnet wire AWG size. In this case it was 28 AWG.

You can drive this combination either with a 30 or 60 mA NST running at 9, 12, or 15 kV. We usually use 12 kV as these are common and easy to find. The 15s only give an extra 2 inches of spark output and usually not worth the extra cost and searching efforts.

You don't really need to do heavy electrostatic analysis or differential equations to get it right the first time!!

A few more tips:

Using a 4 inch dia. sec form allow 1.5 inches above the horizontal plane of the pri for the start of the sec winding. With a 6 inch (ID) form use 3.5 inches for sec start (again above the horizontal plane of the pri winding). With an 18 inch dia form the correct start height is 6.75 inches. With a 24 inch dia. form its 9 inches. I've discovered this relationship is almost perfectly linear!!! and hope to expound more on that in my next book on TC design. Overcoupling and poor quenching are the two biggest mistakes beginners make when building NST coils.

We used a small remote control drive motor to elevate various sizes of coils while they were running to gather all this information. We were not concerned with the actual value of coef. of coupling --- we just tuned with elevation for maximum spark length off the sec toroid.

Use a toroid with an outer dia in the range of 2 to 4.5 times the dia. of the sec coilform (OD).

Setup a scrap wire primary and tune for resonant point. Then, adjust your MMC cap bank value until you can get a good resonance around 9-10 turns --- this provides a nice loaded impedance match for the cap bank and NST driving xmfr.

Now --- you're done and it will all work right the first time without wasting a lot of your time and effort. Replace the scrap wire primary with a nice copper tubing pri and you are finished. We usually add 2 extra turns beyond 10 to allow for small tuning variations.



Dr. Resonance
----- Original Message ----- From: "Tesla list" <tesla@xxxxxxxxxx>
To: <tesla@xxxxxxxxxx>
Sent: Thursday, December 15, 2005 1:53 PM
Subject: A LUA file for BELA (a help to build electrostatic model of TC)


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

All,

Dmitry showed an electrostatic analysis of his TC, made by a program called FEMM4

I thought it would be cool to be able to automate the construction of the model, because FEMM is not a fast application to work with.

The fun part is designing the grounded shield and the dielectric stuff.

FEMM4 is a combination of a magnetic and an electrostatic calculator. The electrostatic module was formerly called BELA, but is now a part of FEMM4.

Over on Yahoo groups, I met a guy, Senthil Ram, who not only offered to- but actually _did_ write a script to build the model of the secondary coil. With that script as a model, I was able to add the code for the remainder of the coil, as well as position the coils relative to Y=0.

It only does cylindrical coils, not pancake and not conical, but that is because I`m too lazy to do the code. Perhaps a job for you?

If you would like to model a coil of yours, download the file from here:

http://home5.inet.tele.dk/f-hammer/DRSSTC.lua

Then get FEMM 4.01 from here:

http://femm.foster-miller.net/Archives/bin/

At the bottom of the files menu: "load LUA script" load DRSSTC.lua and follow the prompts, to describe your coil and work out the electrostatic stress imposed on the primary by the secondary.

Imput is mm, and the inputs are meant to be self explanatory, perhaps except
no. 5  "Maximum Segment Degree"

This determines how many segments a circle is divided into.
If you make the input =1 then each circle will consist og 360 segments, and it also means that the sourrounding mesh that will be generated to define the matrix that calculations are carried out on, will become very fine meshed.

And that calculation time will increase, as well as memory use will increase.
I think that a value around 20 is a good value to use but YMMV here.
input no.13, "triangle size (o.25-3)"
Has a similar aim, namely to controll the size of the grid. Try with 3 to start with, go lower if you feel better resolution is in order. I went out and got a 1meg memory module for my TravelMate 240 laptop, and as long as I don"t try to model all 3000 secondary turns, I`m alright. Try a couple houndred turns to start with.

I hope it may be of some help to someone apart from me.

Cheers, Finn Hammer