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

Re: Do these numbers look right?



Original poster: JOSEPH CACCIATORE <jocatch-at-us.ibm-dot-com> 


Bart, John, Dr. R,

You cleared up my confusion regarding the the size of the capacitor and the 
NST. I thought I had to match the circuit so the cap resonated with the NST 
The doc didn't say that that was a bad thing 8-). (I am not using a safety 
gap, need to read up on them some more). Yea, if I can change the cap to 
.01uf or whatever, I can play around with the numbers more. I think I do 
want to increase the secondary turns (700 seems a bit low) and I would like 
to reduce the primary a few. I need to go and have fun with the spread 
sheet again! 8-)

You guys are great. Thanks for all the help!

JC

P.S. Got 50' of bare 8 ga wire yesterday to play around with (the only wire 
HD had here). Since it is bare I guess I can't just lay it out on a piece 
of cardboard, uh? How do you guys get such nice looking spirals?




"Tesla list" <tesla-at-pupman-dot-com>

02/17/2004 09:10 AM
To
tesla-at-pupman-dot-com
cc
Subject
Re: Do these numbers look right?




Original poster: Bart Anderson <classi6-at-classictesla-dot-com>

Hi JC,

I ran your specs though Javatc.

Two main main concerns:
1) length and turns vs. the wire size.
2) cap size

The bare wire diameter for #22 is nominal 0.0253". With single build
insulation, the total wire diameter would be:
.0253 + (2*.0013) = 0.0279". Note, the .0013" is the insulation thickness
for single build #22 (from surface of bare wire to outer surface of
insulation, thus, when in a single layer winding, both sides of the wire
see the insulation thickness).

What that all boils down to is 765 turns won't fit using #22. If you want
to keep the 5:1 h/d ratio, you could go to smaller gauge wire or decrease
the turns. Or if you don't mind a 5.5:1 h/d ratio, you could use #22 wound
to a length of 22" (780 turns).

With Ed's spreadsheet, the turns per inch value needs to decrease. For #22,
the minimum would 35.84 TPI, but that assumes zero spacing between turns
(not realistic). I think about 35 would be a little more "real". If you
were to use this 35 TPI in Ed's spreadsheet, the turns will show 679 -at-97%,
which is 700 turns at 100% (nice even number).

When I calc'd with your original numbers, I showed the sec Fres at 281 kHz
with the toroid. I also modeled this as though it were in a garage or a
room. This would have tuned in at 12.1 turns (and that's good). This is
something Javatc can do well.

If I assume you changed to 700 turns for the reasons mentioned above (35 TPI);
The sec Fres = 308 kHz and the primary tunes in at 11.14 turns. So the
primary will be fine and allow a couple extra turns for fine tuning.

There is an issue here you need to consider. The .007uF cap size you've
chosen for the 12/30 NST is the "resonant" value for the transformer
impedance. This can be troublesome with NST's (their death!). I would
highly recommend going to the Static Gap LTR value of 0.01uF.

Should you do that, the primary tune point will decrease (more C, less L).
It comes to 9.44 turns. Something else which is nice is you only need 26.5
feet of tubing. Still, add a couple more turns than you calc to ensure
human error, component error, program error doesn't get in the way. Home
Depot should have tubing in 50" lengths and less. If not, just keep
checking the local hardware stores in your area.

Take care,
Bart




Tesla list wrote:

 >Original poster: JOSEPH CACCIATORE <jocatch-at-us.ibm-dot-com>
 >
 >Hello again. I am well on my way to designing my first TC after giving up
 >trying to restore my old PE magazine coil from 1964. I checked and there
 >are a slew of programs and spreadsheets written to help design a TC.
 >Unfortunately some give different results for the same input conditions.
 >
 >The program I found to be extremely helpful is an Excel spreadsheet called
 >the Tesla Coil Designer 3.0 by Ed Sonderman and Corey Ruch. It is great
 >because you can adjust the secondary/primary parameters and watch for when
 >they come in sync.
 >
 >But never trusting any program fully (i.e., Tax Deluxe say I had an extra
 >$1800 coming back on my tax return, it was wrong!), I was hoping someone
 >could verify and look at these numbers to see if the coil really will
 >perform before I spend all this time, effort and money.
 >
 >If anyone is willing, here are the specs I came up with:
 >
 >NST = 12,000v, 30ma, 60hz
 >Secondary = #22 wire, 20" long, 4" diameter, 765 turns-at-97%
 >Primary = flat spiral, I.D. = 6", 15 turns, .25" pipe, .25" spacing,
 >width=7.25"
 >Capacitor = .007uf
 >Toroid = Dmin= 4", Dmaj= 24"
 >
 >The spreadsheet says secondary res F = 257kh, primary res F = 221khz.
 >
 >My concern is if I use any larger cap, I can't get the prim/sec parameters
 >to match good enough. Also, I don't have any material for a toroid here
 >that I can look at so I assume 4" almunin flex hose but I don't know if
 >24" is realistic. Lastly, the primary requires 53' of copper tubing. I am
 >not sure where to buy such a long length. I have seen only 25' at Home Depot.
 >
 >Thanks.
 >  JC
 >
 >