Re: HV Divider Schematic

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

Original poster: "Jim Lux by way of Terry Fritz <teslalist-at-qwest-dot-net>" <jimlux-at-earthlink-dot-net>

The basic design concept shown in Dan's design is basically sound, but I
have a few questions.

At 60 kV, you'll have almost 10 kV across each of the 7 sets of divider
resistors.  That seems awfully high, unless you're using special high
voltage resistors.  I assume the resistors are something non-inductive? Most
HV probes I've seen that are made with multiple resistors use a lot more
resistors (which admittedly makes design a bit trickier, especially for HF
response). This probe will also draw about a milliamp of current, which is
quite a lot (70 Meg -at- 60 kV>> 50W?) Are those 5 Watt resistors? (using the
"dissipate no more than 1/2 the rating" rule of thumb, and for good
temperature/metering stability, one would want more like dissipate no more
than 1/10th)

At 10 MHz, the 160 pF in each stage is only 99 ohms impedance, for a total
string impedance (neglecting the bottom end of the string) of about 700
ohms. That's awfully low compared to your resistance, and will certainly
dominate your design. At your 60 kV rating, you're going to have 60000/700 =
almost 100 Amps of RF flowing through the string. At 60 Hz, the 160 pF is 16
Meg or so, a more reasonable match to the 10 Meg resistance.  Even if the
design isn't actually for measuring 60 kV at 10 MHz, but is more designed to
have 10 MHz BW on transients, that really low impedance is going to perturb
the circuit you are measuring (unless it's something like a 100 uF pulse cap
discharging, where the 20 pF equivalent series C is insignificant).  I'd
also question the RF/transient performance of the design without seeing the
physical layout. Most HV gear is physically large (60 kV rated stuff is
usually more than a foot tall, more like 18-24") and the inductance of even
a 18" wire (about 0.5 uH) is significant at 10 MHz (3 ohms reactance).

Most precision HV dividers I've seen (as from Ross Engineering, etc.) use
much smaller stage capacitance (a few pF), usually formed by the
equipotential shields or the top toroid to ground, for instance.

Just goes to show that making high quality pulse measurements at HV is
non-trivial...
(people can and do write entire books about it, and devote their lives to
it)
----- Original Message -----
From: "Tesla list" <tesla-at-pupman-dot-com>
To: <tesla-at-pupman-dot-com>
Sent: Tuesday, February 04, 2003 7:31 PM
Subject: HV Divider Schematic


 > Original poster: "by way of Terry Fritz <teslalist-at-qwest-dot-net>"
<dhmccauley-at-spacecatlighting-dot-com>
 >
 > Sorry, the last link didn't work.
 >
 > But anyways, here is a schematic of a recent high voltage divider I
designed
 > which was rated for up to 60kV, and a 50 ohm termination.  It is a good
 > example of a very well designed
 > high voltage divider and should give you an idea of how these actually
look
 > like and how the compensation networks are applied.  This particular one
 > works extremely well up and beyond
 > 10 MHz.
 >
 > But of course, if you are only planning on using one of these to hook up
to
 > a digital multimeter to measure your primary voltage (60Hz), then you
don't
 > need anything this elaborate.
 >
 > http://www.spacecatlighting-dot-com/hv_divider.pdf
 >
 > Dan
 >
 >