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

Re: 5MV, 312kW



Original poster: Greg Leyh <lod-at-pacbell-dot-net> 

5MV along 250" of coil length equals 2o,oooV per inch *average* 
gradient.  I'd venture to say that such a gradient is unrealistic by about 
a factor of 3 for a pampered indoor coil, and a factor of 5 for an 
outdoor-rated coil structure [where a coil of such dimensions would 
probably end up].
For the ALF, I'm designing for 4.8MV, supported by an 1152" coilform.  This 
might seem a bit conservative, but failures of the tower structure could be 
quite costly.


>Original poster: Finn Hammer <f-h-at-c.dk>
>All,
>
>To arrive at 5MV on the top, the toroid has to be the right shape and size 
>to suppress breakout at lesser voltages.
>It _could_ look like this in INCA:
>* 5MVtoroid
>
>Lcenter terminal 20 0 2 1.17 2
>Cedge terminal 20 2 2 0.83 0 180
>Cedge terminal 20 2 2 0.83 0 -180
>
>and would then be :
>Minor diametre: 1.66 metre (65")
>Major diametre: 5.00 metre (196")
>Chip in with 250pF of topload and
>  3125 Joules of energy........
>
>Running the coil below it, at 100BPS, would require 312.5kW
>
>Let`s be bold, without being utterly rediculous, and charge the primary 
>capacitor to 50kV, it will then have to be 2.5µF
>
>Pairing this cap up in a tank with a primary coil that has 2 turns, 
>starting at 70" diametre, 10" spacing and 15.6µH will make it ring at 25kHz.
>
>All we need is a secondary that is 50" in diametre, 250"long wound at 3 
>turns per inch.
>
>Assuming 2nd. notch quench, and a coupling of 0.122, the energy will be 
>transferred after 13 cycles/break, or a total of 0.052secs/sec. Ringing at 
>5.2% of the time, there should be ample time to charge the cap btwn. bangs.
>
>I`ve used INCA, ballpark numbers, and wintesla (Yeah, that`s right: the 
>reg`d copy), for this brief design exercise which untill someone points 
>out an order of magnitude error, suggests that a 5MV coil is not at all 
>that unrealistic, not being concerned with the aspect of funding.
>
>Cheers, Finn Hammer