From: "Terrell W. Fritz"
Date: Fri, 01 Mar 2002 09:41:22 -0700
Subject: Re: [TSSP] Ion Cloud Loading
Hi Paul, I suddenly found as I walked out the door for work this morning that the whole town was socked in with snow. Weee!! Snow Day!! :-)))) I think the output impedances of the TCT and Bertan really killed the Q of the other test so I set it up differently. I set up the secondary driven from the HP33120a (Zout = 50 ohms, I could have used a low-z amp but I forgot) with the Tek3012 monitoring the signal with the antenna as is normally done. I set a second coil and terminal next to the first and charged them up with the Bertan: http://hot-streamer.com/TeslaCoils/Misc/IonCloudTest/NextTest.jpg The center and -3dB frequencies are: 263.84kHz 262.00kHz 260.30kHz Thus the Q is 74.86. I set the signal to 9.00 volts (I must have moved because it was 9.04 when I started the test) giving a resolution of 1 part in 900. I could detect the Fo frequency change 140Hz in this way or 1 part in 1870. So charging the right sphere up from 0 to 45 kVDC gave the following shift in the received RMS signal. kVDC Voltage 0 9.04 5 9.04 10 9.04 15 9.04 20 9.04 25 9.04 30 9.04 35 9.04 40 9.04 45 9.04 (corona) So having the coil close to a charged object seems to have no effect on the Fo frequency. If there is some tiny effect down there in the noise, it would not be detectable in normal situations unless something goes wildly non-linear at really high voltages. >Oh, and while we're putting kV of DC into the coil base, would >this give a nice and accurate method for estimating the total Cdc >of the secondary? Charge to a few kV, disconnect the Bertan, and >then time the voltage decay as it leaks back to ground via a >known shunt resistance? I don't think so. Switching a few kV tends to be "messy" and corona leakage may be a big factor. There is no reason such a test would not work just as well at only 5 volts were normal instruments and such are very sensitive and capable. I don't see any advantage of doing a discharge test for capacitance over a more sensitive and easy LCR meter as long as the sense signal's frequency was far away from the coil's Fo frequency. Cheers, Terry At 09:50 AM 3/1/2002 +0000, you wrote: >Further to this, > >It occured to me that it might be better/safer to drive the >coil from the TCT via a few well insulated primary turns, >otherwise fine tuning could be nerve racking. > >Terry wrote: >> I found no evidence that the coil's DC potential had any >> affect on the system's resonant frequency > >Oh, I see now. If fres varied, then you'd expect a lower >amplitude. > >> The "Q" of the coil setup was about 15.8 (253/16) > >So not a very sensitive test, I think. How about repeating >it with TCT into primary, freq counter, and a much smaller >topload (so that the 'ion C' has a greater relative effect). > >Greg wrote: >> I suspect that it's the additional capacitance of the streamer >> channels themselves which are actually responsible for any >> measured tuning shifts. > >And I'm guessing that Bert would concur. The streamer channels, >rather than a cloud of ionised gas molecules, would be the >major source of an 'upheld' charge. > >Oh, and while we're putting kV of DC into the coil base, would >this give a nice and accurate method for estimating the total Cdc >of the secondary? Charge to a few kV, disconnect the Bertan, and >then time the voltage decay as it leaks back to ground via a >known shunt resistance? >-- >Paul Nicholson, >--
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.