30MHz Spark Gap Testing - Is this real??
From: terryf-at-verinet-dot-com [SMTP:terryf-at-verinet-dot-com]
Sent: Tuesday, April 07, 1998 6:31 PM
Subject: Re: 30MHz Spark Gap Testing - Is this real??
At 01:18 AM 4/7/98 -0500, you wrote:
>From: Greg Leyh [SMTP:lod-at-pacbell-dot-net]
>Sent: Monday, April 06, 1998 4:13 AM
>To: Tesla List
>Subject: Re: 30MHz Spark Gap Testing - Is this real??
>> What are the implications if this is true? If the primary
circuit is doing
>> all of its work at ~50MHz and at hundreds, if not thousands, of amps at that
>> frequency, everything changes! The conductors must be short, wide, copper
>> strips. The capacitors must be able to withstand even greater stresses than
>> we ever imagined. And the spark gaps... who knows? This would imply that a
>> much better spark gap or other switching system may give much better
>> performance. EMI becomes a major concern.
>The sharp, high spikes in your waveforms look familiar;
>I have seen those on most of my scope msmts in and around
>the primary vault, even when measuring points on the 60Hz
>circuits. I had been attributing this to switching noise
>capacitively coupled from the rotary gap, since my scope
>sees them even with the input leads shorted! Wrapping the
>scope (Fluke 123) in Al foil had helped somewhat.
Thanks for your response. Are the spikes you see occurring as one powerful
spike at the beginning for the firing cycle and then a series of other lower
spike at the peaks of the fundamental waveform?
My Tek scope seems very immune to radiated noise but I am not dealing with
the giant system you are! I need to work more on verifying the noise
shielding and such. I have a bunch of die cast enclosures on order which
should vastly reduce any noise getting in.
>What type of sensor are you using for the current msmt?
>Wideband CT? Shunt?
My sensor is a 0.01 ohm array of ten 0.100 ohm 1 watt surface mount thick
film resistors in parallel. I copper strapped them up to have good RF
performance. I work with high power RF so I have a good idea how to do
this. This shunt is connected to a simple LED driver circuit. The
expensive fiber-optic LED has a rise and fall time of 3 nS. I cannot
verify the actual HF performance with the equipment I have, but it seems to
work well. More testing needed here.
>One possible verification of the signal fidelity is to
>short the scope input _at the sensor_ and see if the
>signal persists. You probably have already done this.
If I short things out and do other simple noise checks all looks good
except my fiber-optic voltage probe which is obviously being messed over
hard by the spikes. Hopefully, the enclosures will help this. this is why
I turned to the use of the antennas.
>The main reason that I am skeptical is my present coil
>seems to work at about 70% efficiency, with primary ckt
>plumbing that is expecting most of the power to be at
>38 kHz. My primary would be hopelessly sloppy at 50MHz.
I suspect the frequency of the bursts varies widely from system to system.
Also the spark gap may have high frequency bursts but these may get
attenuated down by the time they reach the primary inductor and such. We
normally think current is equal in the whole primary system. However, at
very high frequency that may not be true. All that wiring starts looking
like transmission lines and such. More testing to be done here to. It is
well known that the primary should have good high current and high frequency
design despite the low frequencies the coil should be working at. If the
primary is working into the MHz region at huge currents, this behavior would
be well explained. More testing here as well.
I will continue testing and with others help we should be able to figure
all this out pretty fast. I haven't found any great mistakes yet and it
still seems like a very real phenomenon.