Re: Marx Generator

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

Original poster: "Kevin Ottalini by way of Terry Fritz <twftesla-at-qwest-dot-net>" <ottalini-at-mindspring-dot-com>

Richard:
        There doesn't seem to be a perfect value ... it depends on what sort
of pulse you want out of the system.  I tried a variety of values from 50K
up to 1Mohm and found that I liked the 300K resistors.

My HV bible: "High Voltage Engineering" by Naidu and Kamaraju has a
little bit of info (some 6 pages) about Marx Generators summarized by
the following 5 items, with comments by me in parenthesis:

1.  The charging resistor Rs (the one between the source supply and the
        first stage of the stack) is chosen (typically) to limit the
charging
        current to 50ma or 100ma (I don't have one in my design ... and
        should).

2.   The total capacitance Ct (all the caps in parallel) is chosen such
        that the product CtRs is typically on the order of 10seconds to
        one minute (this is certainly not a requirement ... I chose one
        about two or three seconds).

3.  The usual gap spacing is chosen such that the breakdown voltage
        of the gap is greater than the charging voltage.  (Again, this is
not
        a requirement and in fact I use a gap spacing that is slightly less
        than the maximum charging voltage ... at least one gap in the
        stack that is).

4.  The discharge time constant CsR1/n (n=number of stages, R1 =
        the value of one of the resistors in a stage) will be very very
small:
        I.E.: microseconds compared to the charge time, so no (significant)
        discharge takes place through the charging resistors (see below).

5.  Impulse generators are nominally rated by the total voltage (nominal),
      the number of stages and the gross energy stored.  The nominal
      output voltage is the number of stages multiplied by the charging
      voltage.  The nominal energy stored is given by 0.5CV^2. (either
      the stages in parallel and the charging voltage, or the stages
      in series and the output voltage).

My little unit has 10stages at 1000pf each stage and a charging
voltage of 60kv, so the stored energy is about 18 joules
(18watt seconds).  I use a 10watt supply, so it takes at least
a couple of seconds to reach the gap breakdown voltage.

If you use a lower value resistor, your charge time will be faster,
but then you need a lot more power (and larger wattage resistors).

If you get the resistor value too low, then the resistors start to
represent a significant loss of energy during the discharge.  In
my case, the system sees  300Kohms*10 (3.0Mohms) at 600kv.

The charging resistor RC is therefore 3Mohms*100pf or about
300uS, but the actual discharge time is at least 1/100 that
(or less) so the loss of energy through the charging
resistors is insignificant.

The actual discharge time could be calculated if I know what the
effective resistance of all the gaps was during the discharge, (and
the resistance and capacitance of the main discharge gap and all
the parasitic capacitance) but I suspect it would be easier to
measure the discharge time and calculate backwards.  I'm just
a little reluctant to blow my O'scope, so I may leave this task
undone for a while.

(This was a subject on the list a while ago ... very useful stuff indeed!)

Variations of the circuit design are usually for waveshaping, cost and
space requirements., but there are a number of variations that deal with
control ... important when you are doing specific sorts of testing.

You can also generate pulses of either polarity, and I suspect that
the system could be possibly driven by AC as well (if you just want
a lot of sparks that is!).  I plan on trying this in a little while ... I'll
let
you know what happens.

Best,
            Kevin

----- Original Message -----
From: "Tesla list" <tesla-at-pupman-dot-com>
To: <tesla-at-pupman-dot-com>
Sent: Wednesday, December 26, 2001 8:11 PM
Subject: Marx Generator


> Original poster: "by way of Terry Fritz <twftesla-at-qwest-dot-net>"
<Robynsaddiction-at-aol-dot-com>
>
> Hello List,
>
>
> I know that this veers from Tesla Coils, but I was wondering if anyone
> could direct me towards a web page, or perhaps a book, that gives the
> formulas needed to make a Marx Generator. That is, so that I can have some
> method of knowing what value of resistors would work best for my other
> given values. Also, where online can one purchase the high voltage ceramic
> type resistors required for a Marx Generator.
>
> I have been to a few web sites that discuss the theories of operation for
> their Marx generators, but none of these sites told how they determined
the
> values of the components for their Marx generators. Any help would be much
> appreciated.
>
>                Thanks, Richard Harris
> Robynsaddiction-at-aol-dot-com
>
>