Re: wanted high voltage resistors

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

Original poster: Jim Lux <jimlux-at-earthlink-dot-net> 

At 07:55 AM 11/18/2003 -0700, you wrote:
>Original poster: Matthew Smith <matt-at-kbc-dot-net.au>
>Tesla list wrote:
>>Original poster: "colin.heath4" <colin.heath4-at-ntlworld-dot-com>
>>hi all,
>>         im looking for some high voltage resistors for my marx bank. its all
>>i need to finish. i hope to get 20kv approx 300k ohm
>--SNIP--
>
>What actually makes a high voltage resistor different - is it just a 
>stronger glaze?  What I'm getting at is: can you increase the voltage 
>rating of a resistor by potting it or dropping it in oil?

it kind of depends on what voltage you're talking about.  Physical size is 
probably the most important difference.  You don't want a flashover across 
the surface of the resistor.  For instance, if you use the usual (3xfree 
air distance) guideline, a 30 kV resistor would need to be at least 3 cm 
long (and that would be pushing it).  As you say, though, putting it in oil 
would help.

Another important aspect of HV resistors is their power/energy handling 
ability.  It's easy to dissipate too much power in a moderate value HV 
resistor.  I learned this the hard way when building a voltage divider to 
measure 20 kV kinds of potentials, and forgot that 1 mA at 20 kV is 20 
Watts. I had 40 470K  resistors in series, but, oops, they were 1/4 watt, 
and I noticed that funny smell that tells you that you have made a serious 
design error.

Energy handling and peak power becomes important in pulsed power kinds of 
applications. A wirewound resistor may be rated at 100W average power, but 
may not have a peak power rating much more than 3 or 4 times that (because 
the wire acts like a fuse).  Put it into a circuit with a 1% duty cycle 
(i.e. 10 kW, 1% of the time) and it may fail somewhat spectacularly (gosh, 
I wish I had a video camera for that!).  Pulsed power resistors tend to be 
bulk resistive media (e.g. units from Carborundum,Cesiwid,Globar,Kanthal, 
etc.) or, as in the Maxwell resistors, a wide card with resistive media on 
it, to spread the energy out.

Finally, there's a signficant issue of Voltage Coefficient of Resistance, 
depending on the application.  Partly this is due to temperature 
coefficient (current heats the resistor changing it's value), partly this 
is due to the resistive medium changing resistance in high electric 
fields.  If you spread the field out over a longer physical space, then the 
field is lower.


Potting for HV is a funny thing.. It can actually make things worse.  Small 
voids in the potting may increase the field in the void (or at the edge of 
the void) because the potting material and air have different dielectric 
constant(epsilon). Potting can also change the thermal properties, 
depending on the potting material.  Potting for high performance HV 
assemblies is very much a proprietary art, with careful selection of the 
filler material and the surrouding matrix for good electrical, thermal, and 
mechanical properties, and then careful process control for the 
voids.  It's a far cry from sticking the parts in a cooking tray and 
pouring in the 2 part RTV.

If I had to say what the *single* most troublesome aspect of HV things that 
I deal with at work are, it is potting and related areas.  Components, 
mechanical design, circuit topology are all fairly straightforward and you 
have datasheets and test data to use in the process.