From: Bert Hickman
Date: Fri, 31 May 2002 00:09:19 -0500
Subject: Re: Top V Probe Design (was RE: [TSSP] Topload breakoutpotentials)
Mark, Terry, Paul, After more thinking about the problems plus Paul's analysis of the parasitic capacitance effects in the compensated divider, I am forced to agree with Paul - the design Terry is proposing looks to be an elegant approach that should solve most problems. While calibration may be a bit of a challenge, it should be manageable. One other challenge still remains: we need to insure that we prevent any induced EMF's or circulating eddy currents from raising havoc with the measurements. This means that we'd likely NOT want to use a circular "bases" for the pickup and ground cylinders, since these would introduce undesirable circulating currents (not to mention losses!). The facing edges of the slotted cylinders will have an induced EMF of hundreds or even thousands of volts. To prevent circulating currents and balance the e-field contributions, we should pick off the signal and ground signals from a _single_ point on each cylinder that's located 180 degrees opposite from the slit. BTW, I had a thought as to how we might reduce eddy current losses. Suppose we wind the pickup and ground shield "cylinders" as continuous windings using magnet wire, holding the windings in place with varnish or polyurethane. All of the wires would then cut along a vertical line up the side of the winding using a small handheld grinder. Then, carefully strip the wire's insulation along a vertical line 180 degrees opposite from the vertical slit just created. Finally, a wire would then be soldered all along the stripped length of the winding to provide a common contact point for the "cylinder" - each wire would be connected only at one point to prevent unwanted current loops. The result would be virtually perfect electrodes (from an electrostatic/capacitive standpoint) which also keep eddy currents and EMF fields at bay. Best regards, -- Bert -- -- Bert Hickman Stoneridge Engineering Coins Shrunk Electromagnetically! http://www.teslamania.com "Mark S. Rzeszotarski, Ph.D." wrote: > > Greetings Bert and All: > At 07:46 AM 05/29/2002 -0500, you wrote: > >Hi Malcolm, > > > >This could work in principle, but I think a larger amount of capacitance > >would be desirable in practice to reduce AC ratio errors introduced by > >the stray capacitance between voltage grading toroids/rings and to the > >surrounding secondary winding. I'd think that we'd want to keep the > >capacitance between successive stages at least 100X as large as the > >parasitic capacitance to reduce the effects of capacitive coupling from > >the surrounding secondary. >> >> It occurred to me last night that if the central > >> divider chain were to be a compensated RC ladder, it might be > >> possible to make a resistive chain self-compensating by using the > >> stray capacitance between resistors. Essentially, by arranging the > >> resistors in the following fashion: > >> > >> | > >> ---///-- > >> | > >> ---///-- > >> | > >> ---///-- > >> | > >> ---///-- > >> | > >> > >> etc. and suitably sizing the resistors such that their resistance > >> matched the parasitic capacitive reactance, we might obtain such a > >> result. > > Looking at my High Voltage Engineering text last night, I see that > the big boys are using deionized water doped with CuSO4 in PVC tubing as the > resistor with the tubing spiral coiled to distribute the capacitance and > corona effects. They claim bandwidths on the order of 100 MHz with their > systems (used to look at impulse/step responses similar to our pinger > experiments). Also, 1000:1 is not near enough ratio for a live coil. I > would shoot for 500,000 top volts dropped to 5 volts or a 100,000:1 divider > ratio. > Regards, > Mark S. Rzeszotarski, Ph.D., Assistant Professor of Radiology and Biomedical > Engineering, Case Western Reserve University, Cleveland OH USA
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.