Re: Energy storage in primary?

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

Original poster: "Terry Fritz" <teslalist-at-qwest-dot-net>

Hi Jolyon,

At 12:58 PM 1/28/2003 +0000, you wrote:
>Is it possible to build a TC where energy is stored initially as high 
>current in an inductor (the primary) rather than high voltage in a capacitor?

It is theoretically sound and possible.  But not easy ;-))  In a way, my 
OLTC does this but it still uses capacitors to drive large primary 
currents.  It really could use high current power supplies too.  They would 
have to be 700V 5000 amps supplies...  But that would eliminate the 
caps.  One could also just make a very high current CW supply and run CW.

>I am thinking of a setup where current ramps up slowly through the 
>inductor before being abruptly switched off (by semiconductor switch or 
>similar) after a predetermined current or period of time has been 
>exceeded; the current in the primary rising and falling as "saw-tooth" 
>waveform.

Much like a CW coil.

>
>As primary input power for this would be largely determined be current 
>rather than the voltage of the PSU
>how high would the current have to be/ how low could the voltage be for 
>decent spark output say, a minimum of 6 inches or more?

Using John's formula, a 6 inch spark needs about 15 watts of power.  At 
120BPS we are talking .125 joule per bang.  That enrergy would have to be 
stored in the inductor where.

0.125 = 1/2 L I^2

Normally, L would have to be a very low value (like in my OLTC) otherwise 
the current in the inductor will drive very high voltages at the switch.

>
>For the control logic would this likely need an exotic switch-mode power 
>supply IC with PWM and dead-time control or could a simple astable like a 
>555 do the job?

Nothing exotic at all, just a simple 555 timer, TLP250 isolator/drive chip, 
and a really good IGBT.  A few other details about not blowing things up, 
but that is basically all it takes to switch such things ( <1000V, >1000 
amps) at your whim.  Since we are using low duty cycle pulses and low loss 
IGBTs, heat on the IGBTs is very low.

>
>For the high-current, high-speed switch would bipolar transistors (e.g.. 
>TV line-output power devices) or MOSFETS be suitable or would IGBTs be 
>necessary?

IGBTs...  Cheaper to use smaller ones in parallel.

>
>Would it not be necessary to connect a capacitor across the switch to 
>absorb/slow down the high-voltage transient produced when the switch 
>opens/ would necessary voltage rating of switch and capacitor be 
>comparable to that of the primary capacitor in a conventional spark-gap TC?


In that case, it would be exactly like a typical coil.  Remember, you can 
either charge the cap to say 20kV and close the gap. Or, you can drive 1000 
amps through the primary circuit and open the gap.  Either way works and 
they are theoretically identical.  It's just hard to run 1000 amps and then 
have the current source withstand 20kV of RF voltage.  We are talking a 
20MVA switch there!!  Spark gaps do that easily for voltage drive.  Current 
drive is not so easy.

We are so used to voltage drive that we never think about current 
drive.  However, today we have IGBTs and other new devices that may open 
things up for us!  Super high voltage is also not such a big requirement 
since very high current but lower voltage devices are so easy to get now.

Cheers,

         Terry