Re: Good Deal? SURE !!

Hello Nick,

You wrote:
> No way is that a liniac supply (the current is too high and the
>voltage too low.  If it was intended to drive a voltage multiplier it
>would have an ac output).

Sorry, but I HAVE built a Marx bank (multiplier) with a DC input.
Secondly, it depends on the size of the linear accelerator.
A university near us had a very small linear accelerator which
they dismanteld. This was powered by a 20kV transformer.
Also, after Bob mailed me that he had discovered the Leybold
stickers, I wrote (in my last email) that it probably was designed
for vacuum sputtering usage.

Nick wrote:
> If you want to
>use it remember that your local powerco will only give you a 3 phase
>supply if you register as an industrial user.  If you're really unlucky
> you might have to pay for your own substation.  Also the
>transformer will not be shunt wound
>like a neon so you will need an old welder transformer or one hell of a
>saturable inductor to limit the current.

In USA there are almost NO residential areas (speaking for Southern
California here) that have three phase power. The power company
probably would be more than willing to get you a 3 phase connection
(and a LONG term contract) if you pay for the whole wiring route (which
might be mighty long and expensive.....) As Bobīs unit is a three
phase Y machine, how would you go about current limiting it with
a welder?!? You would need at least three on a common shaft.
A much easier way to do this would be to vary the current of the
secondary side (reactor and a variable rotary spark gap).

Nick wrote:
>If you are very smart you could come up with a system which double
>rectifies the output from a mains isolation transfomer then re-
>modulates it to give you the 3 phases -at- 380-440V.(Warm
> MOSFETs)  There are some comercial inverter kits
>available in the uk (maplin part no. VF62s) which you may be able to
>modify to do the job.  You will need to put some linear mos driver
>chips in to take the gate current and a lot of big  and very
>expensive MOSFETs (I'd recommend BUZ61s $3 each) in a push
>pull configuration off +200 -200 supply rails.  For a highly inductive
>load such as transformer you will also need all the smoothing and
>snubbing netwroks you can get your hands on.  Some zeners to
>protect the FETs would be a good idea as well.
>The hard part will be getting the 3 sine wave signals 120 degrees out
>of phase.  The easy but expensive way to do this is to program 3
>eproms with an 8bit sine wave then scan throught the addresses
>with scaled decade counters 120 degrees out of phase and feed
>the outputs to 3 8bit D/As.

Geez, just how complicated to you want to go? That would cost
Bob a fortune !! It would be easier (and much more reliable!!) for
him to visit T&R (e.g.) and pick up a single phase pole pig or a
PT for much less than the cost of the inverter itself (not to
mention the size and weight of the inverter).

Coiler greets from Germany,