RE: differential probes

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

Original poster: "Malcolm Watts by way of Terry Fritz <teslalist-at-qwest-dot-net>" <m.j.watts-at-massey.ac.nz>

Hi Jimmy, Terry,

On 15 Jul 2003, at 12:32, Tesla list wrote:

 > Original poster: "FRITZ, TERRY by way of Terry Fritz 
<teslalist-at-qwest-dot-net>" <TERRYF-at-aei-dot-com>
 >
 > Hi Jimmy,
 >
 > I quickly looked it over and had these thoughts:
 >
 > Out 1 and 2 are connected to the ground legs of the BNC connectors.  They
 > should go the the center pins of the scope connectors.
 >
 > Holes for LED31 and B31 and B32 wires may need to be larger?  Hard for me to
 > judge from the pictures...
 >
 > C12 and C21 should all have their "own" vias going to the ground plane (high
 > current at 150MHz!).
 >
 > There should be three "very" separate grounds.
 >
 > 	The low voltage electronics will have it's own ground plane but that
 > will end on the low voltage side of the 1Meg resistors.  There should be
 > "no" ground plans under the HV resistor areas.  The HV resistors from a very
 > wide line of isolation separating the low voltage stuff from the high
 > voltage stuff.  Only the resistors should cross that line.  The grounds are
 > also "very" separate since under a fault they will be far from zero volts!
 >
 > 	The MOV sets each have their own "very high" current (like 2000 amp
 > pulse) grounds that go out to ground studs.  The MOV grounds are meant to
 > take giant faults and ground them separately away from the other grounds.
 >
 > The Official "CE" clearances for 2000Vp on a PCB are 1 inch!
 > "Theoretically", the high voltage side stuff should all be spaced 1 inch
 > apart from anything else.  2000V will easily arc 1/4 inch between sharp
 > points over a surface.  So you may need far more spacing on the high voltage
 > stuff.

I haven't downloaded the file as I don't have a suitable program to
open it with available on the NW. I do have a tip to deal with the
problem mentioned in the previous paragraph: the way to increase the
isolation barrier without consuming a lot of board space is to
physically cut a slot in the board across the isolation barrier. This
gets rid of the problem of surface tracking. I presume fibreglass is
the material of choice for this board? <paxolin or anything like it
is OUT>.

Malcolm
<snip>