Re: 60kV 50MHz HV Divider Completed

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

Original poster: "Bert Hickman by way of Terry Fritz <teslalist-at-qwest-dot-net>" <bert.hickman-at-aquila-dot-net>

Dan and Gary,

You're both right. It really depends on the type of ceramic dielectric in 
the cap. If the caps are made using Type I (Strontium Titanate, rated for 
RF and pulsed power applications, with lower K than type II or type III 
materials), the dC/dV is considerably less than for DC-rated caps (such as 
type II or III - Barium Titanate). A Type II capacitor can have as little 
as half it's initial capacitance at full rated voltage as it has near 0 
volts, so it can really mess up things in a divider circuit.

However, Type I ceramic capacitors are often used in electrostatic voltage 
dividers in the power industry for transmission type potential transformers 
(69 kV+). Capacitor manufacturers can tailor their proprietary type I 
blends so that the voltage coefficient is almost nonexistent.

An example of one having NO voltage coefficient is Aerovox's HT/HU line of 
Type I HV caps:
http://www.avxcorp-dot-com/docs/catalogs/hthu.pdf

Stacks of disks and rods of this material are sometimes used to "force" 
voltage grading within large bushings for transformers and circuit 
breakers. Caps made from this material are often used in HV potential 
transformers, AC dividers, and "snubbers" for HV vacuum  switches. Since 
you want lower capacitance anyway for the HC portion of the divider, these 
capacitors are ideal for the front-end of a compensated voltage divider 
application.

BTW, resistors can exhibit dR/dV (a Voltage Coefficient of Resistance or 
VCR) as well, especially at higher voltages, but that's another topic... :^)

Best regards,

-- Bert --
-- 
-----------------------------------------------------------
Unique Offerings - Scarce Technical Books, Coins Crushed by
Ultrastrong Magnetic Fields, and Captured Lightning Bolts!
All from Stoneridge Engineering - http://www.teslamania-dot-com
-----------------------------------------------------------


Tesla list wrote:
>Original poster: "Mccauley, Daniel H by way of Terry Fritz 
><teslalist-at-qwest-dot-net>" <daniel.h.mccauley-at-lmco-dot-com>
>  > Hi Dan,
>  >
>  > I couldn't tell from your schematic what type of caps were
>  > used for the 82
>  > pF compensation caps.  My assumption is that they are ceramic.
>Yes, that is correct.  Both Vishay and AVX makes these type of high voltage
>ceramic capacitors.  These particular
>capacitors are rated 82pF at 10kVDC each.
>
>  > I once attempted to build a purely capacitive HV divider, also using
>  > ceramic caps on the HV side.  When I attempted to scope the
>  > output of an
>  > NST, the resulting voltage was grossly different than what
>  > the capacitive
>  > ratio should have produced.  I mentioned this on this List,
>  > and I think it
>  > was Bert Hickman who pointed out that ceramic dielectrics have an
>  > unfortunate property of creating a capacitor that's voltage-dependant.
>Actually, ceramic capacitors are pretty much the industry standard for
>commercial high voltage dividers.  Almost all
>your potted high voltage dividers will use these same type ceramic
>capacitors.
>  > This effect may be less evident if the dominant division is
>  > achieved with
>  > R's rather than C's, just something to keep in mind (and
>  > something that
>  > simulation models don't account for).
>  > BTW, your Marx generator is BEAUTIFUL!  What did you use for the gap
>  > electrodes?  Are the support plates for the electrodes
>  > constructed so that
>  > they are hinged and may be adjusted in unison?
>Thanks.  The spark gap electrodes are the home depot special drawer knobs
>which are standard issue at almost any
>Home Depot and sell for price of $2.29 each.  (adds up considering i have 24
>of them), but well worth it. I was going to
>use standard one inch corona balls, but these seemed to mount much nicer in
>they have nice flat flange on the back end to butte up against the support.
>The spark gap assembly isn't hinged, but does have the bottom and top
>support plates which can either be replaced with
>longer plates, or spacers installed.  I could have made a hinged design, but
>wanted something a little more permanent.  Initially it will be pain to fine
>tune the gap widths, but in the long run it will prove worthwhile.
>I'm finding out by tomorrow whether I'll be going to Kwajalein Missile Range
>(Marshall Islands) for six months, but if I don't i'll definitely be
>bringing the marx generator up to Rochester when I see you in August.  Just
>beware, my minicoil is growing stronger every day!!!!
>The Captain
>
>
>
>  > Regards, Gary
>  >
>  >
>  >
>  >   -----Original Message-----
>  > From:     Tesla list [mailto:tesla-at-pupman-dot-com]
>  > Sent:    Thursday, May 15, 2003 1:45 AM
>  > To:    tesla-at-pupman-dot-com
>  > Subject:    60kV 50MHz HV Divider Completed
>  >
>  > Original poster: "by way of Terry Fritz <teslalist-at-qwest-dot-net>"
>  > <dhmccauley-at-spacecatlighting-dot-com>
>  >
>  > I finally got around to completing my 60kV High Voltage
>  > Divider.  The design
>  > works extremely well (from initial tests) and should serve as a good
>  > baseline for others wanting to build hv dividers capable of
>  > high frequency
>  > measurements.  The frontend features a combination
>  > resistive/capacitive
>  > divider and
>  > a additional box which attaches to the unit to allow fine
>  > adjustments of the
>  > gain/frequency compensation of the entire hv divider.
>  >
>  > http://www.spacecatlighting-dot-com/hv_divider.htm
>  >
>  > Specifications:
>  > Maximum voltage:  40kV-60kV (construction dependent)
>  > Bandwidth: 54MHz (simulated 3dB roll-off)
>  >                     50MHz (measured using impedance analyzer)
>  > Ratio: 1000:1
>  >
>  > Enjoy!
>  >
>  > The Captain
>  >
>  >
>.