[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Terry's New Plane Wave Antenna

Original poster: Terry Fritz <vardin@xxxxxxxxxxxxxxxxxxxxxxx>

Hi,

At 12:11 PM 12/22/2005, you wrote:

> Original poster: Terry Fritz <vardin@xxxxxxxxxxxxxxxxxxxxxxx>

>>ready for the bricks :-)

> If you know how to do it better, do it!

he-he - a nice one, but not to me :-) i`d like to say right away, that
the purpose of my posts is not to  show this antenna is bad, the purpose
is to know - if i understand its working correct ;-)
i don`t know what is better or worse - how about Marco`s liquid
resistive divider? unfortunately  most of people are too small for it %-)
Bart and Grishka are thinking of just using a bare sheet of metal without the back plane structure at all. So there make indeed be "better" ways to make it!! Indeed, a simple wire antenna with a load capacitor might be simpler still. The load capacitor (and resistor) really are the keys to making the response flat and compensation for the 50ohm cable to the scope. The actual "antenna" part could probably be just about anything. The problem with the old "just a bare wire antenna" was that it was full of harmonics and wild non-linearities. But simply the load capacitor, resistor, and coax "fix" that problem.
Since the coil's fields are orders of magnitude higher than anything else, shielding and directionality (?) are not really needed. As I mentioned before, the antenna just sort of "turned out" as it did for no great reason... Mostly due to parts I had right there in the junk box and I wanted to make it mount on a tripod. Feel free to change it if you want. 



>>the most interesting - to see the voltage changing at the start (and
>>after) the discharge appears, but the antenna will work in its
>>worst way in this case - coz the discharge carries the area  of the
>>strong field in its head. even if the discharge is not directed in the
>>antenna side - it should  change the field configuration a lot anyway.

> One could test for that with just a common metal wire shaped like a
> streamer.  That tends to increase the secondary capacitance buy the
> usual 5 to 8%.  But the system voltage would drop with this load.  I
> suppose one could "calibrate" the antenna with a wire streamer
> simulator too...

can you implement the "wire streamer" into E-tesla to run a "real"
simulation? or maybe there is  something like "Bela 3D" here?
coz i can`t fire coils now - only computers :-)))
Use the second terminal. Imagine the wire streamer going straight up from the center of the coils top terminal. The 2nd terminal is like a sphere say 1mm in diameter. the program automatically connects a wire to the terminal. So you can add a wire just fine with the present program. 



>>imo this antenna can`t be used for any measurements - only for
>>observing waveforms, and even then with some restrictions. but
>>waveforms of what? indeed not of the secondary voltage itself, but
>>summ of it with voltage changes caused by other processes.

> It is a question of accuracy.  I think the antenna does really well
> compared to the next best thing, which does not seem to exist ;-))

you take its design from industrial e-field antennas at your work? but
they are truly _field_  antennas, and your device is not, as you have
said many times.
so output of your antenna is in proportion not with the voltage
between two points in the field at arbitrary distance from the toroid,
and not with the field strength/voltage at one point - it`s
proportional to the voltage on the capacitance 20nf, so if i`m not
mistaken, the antenna is not sensitive to the field  distortion,
caused by presence of some heavy things near (not necessarily conductors).
It is a capacitive divider. I assume, for better or worse, that the structure and nature of the toroid to antenna capacitance does not change at all. 


but how do you think - does the capacitance between the toroid and
front plane of the antenna depend from the strength/form of the field?
coz the field between the toroid and the antenna is not uniform, and
this field is not completely contained between them.
As long at it does not "change", the shape and distortions do not matter. For a give set of fields, the voltage should be linear and accurate. If you have it all setup and calibrated but then place a chair between the toroid and antenna, then everything will be changed by the chair. 



you advice to calibrate antenna by giving a known high voltage at the
toroid - say, from the nst. but  the field of working coil is much
stronger - imo the charges in the area "toroid-plane of the
antenna" will be located in other way, so the capacitance of this area
will change and the calibration will be not precise, no?
It is really best to resonant the coil from a signal generator. Then the top voltage is the input voltage multiplied by the "Q" of the secondary system. Then the fields from say the secondary coil itself are more realistic. But the NST is really easy and simple to do. It depends on how close you really want to get. In many cases, the absolute voltage really does not matter because you are looking for notches or other things rather than pure voltages. Probably 90% of the time I don't care about calibration since I really just want to "see" the waveform's shape, timing, duration, etc.. 



p.s.

"Testing:
Predicted waveform
[...]
Actual waveform
[...]
Almost a perfect match ;-)"

if these waveforms are for the case without sparks? but you can`t
predict waveforms in case of the  sparks presence, so how could you
know if it is perfect or not perfect match in that case? you`ll just
have nothing to compare with ;-)
One directs the sparks to the far side away from the antenna. Obviously, if the sparks are say 4 meters long, that is not easy to do ;-)) But if the sparks are not too big, the toroid shields them well. If someone finds out that the waveforms really don't look like that at all, let me know :o)) The antenna might be telling all lies!!! But the vast majority of the evidence seems to agree with it... 



p.p.s.

"In order to get the signals from the antennas to the oscilloscope
input, a coaxial cable is used. This cable represents a transmission
line and the antenna system must be designed to take the cable
impedance into account.
[...]
R1 is fairly critical and should be a small 1/8W carbon resistor
whose value is selected to be very near 50 ohms to match the impedance
of the cable."

you have matched the transmission line from one end - from the side
of the generator, but at the load end there`s no any matching -
wouldn`t it be the reason of having high reflected signal (or ringing)?
It works fine. Gerry also asked that and went through it all... Unlike an RF power thing, we have a signal source going into a very high impedance load. Thus, you only have to impedance match one end. That damps the system (Q=0.7071..) so it does not ring or distort at the frequencies we are working in. There is a ring at about 300MHz as the 2M line and ground shield hit the propagation delay (speed of light thing) but that is far higher than we are concerned with. 

Cheers,

        Terry



-----
Let the bass kick! =:-D