Re: Maximum input voltage for cockraft walton voltage multipliers.

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

Original poster: Hauke Johannknecht <ash-at-ash.de> 

On Sat, 6 Mar 2004, Tesla list wrote:

 > The diodes must have a minimum rating of Erms x 2.8.   All caps should have
 > a minimum rating of Erms x 1.414.  Practical applications usually use diodes
 > rated at Erms x 3.6 and caps rated at Erms x 2.2.

there are various modifications to these (indeed good)
requirements out there.

        | HVDCout
   |-|<-|
   |    =C4
   |->|-|
C3=    |
   |-|<-|
   |    =C2
   |->|-|
C1=    |
   |    |--GND
    ACin

a) sometimes C1 has half the voltage rating of the other caps
b) sometimes the C1/C3/.. string gets higher capacitance than the
    C2/C4/.. string.
c) sometimes the caps in the stages closer to the ACin end get a
    higher capacitance than those higher up.

and from the "dont try this at home unless you have a huge bag
of spare rectifiers" division:
d) Mr Barros using diodes with Erms x 1.95 rating at twice
    their rated current and wondering why he keeps blowing them.

a thing i have been curious about lately:
how fast should the rectifiers be for which ACin frequencies?
is there some rule-of-thumb for that question?
how complex would it be to calculate (or estimate) the losses due
to diode slowness for given input/diode characteristics?

assuming a "100ns" rectifier and 20kHz feed frequency, the 100ns
translate to 0.4% of the AC halfwave it is supposed to block.
that number puts my uneducated mind at ease without even looking
at the diode profile, because it just seems so small.

for setups like Hydro18s MOT-cascade and other mains-frequency
operated cascades the speed of the rectifiers seems to be
somewhat irrelevant as long as they arent steam-powered.

i would really like to get some real-world feedback on these
questions.

MfG,
   Hauke

-- 
Hauke Johannknecht        Berlin / Germany        HJ422-RIPE
Use PGP ! -> lynx -dump http://www.ash.de/ash.asc | pgp -kaf