I have been theorizing on how to build a system for the transmission of longitudinal waves, after I found this article wherein a succesfull practical proof of concept is described:
Since it was reported by Dollard that the propagation speed of longitudinal waves is a factor pi/2 (1.57) larger than the propagation speed of transversal waves, I figured a demonstration of longitudinal moon bouncing would be THE final chapter for Einstein's relativity nonsense, so I started a thread at the EF to see how far we can come with that:
Most important conclusion so far is that your sphere has to have an n * 1/4 lambda radius in order for it to resonate like a dipole antenna, since with a n * 1/4 lambda radius, you basically have an infinite array of 1/2 wave dipoles....
Now if you want to calculate the wavelength for the frequency you are designing your transmitter for, you can simply calculate the corresponding transversal frequency by dividing your longitudinal frequency by pi/2 (1.57). I calculated that for the values reported in the paper:
"They used a frequency of 433.59 MHz, with an equivalent EM frequency of 276 MHz. When we feed that in a wavelength calculator ( http://www.csgnetwork.com/freqwavelengthcalc.html
), we get a wavelength of about 1.1 m or 1/4 lambda of 27 cm, while they used a sphere with a radius of 30 mm, which would be about 10% more than 1/4 lambda."
And apparantly that works pretty well. Interesting detail is that they feed their sphere from the centre, where you have a current node, just as what you have with a normal 1/4 lambda dipole, so you can drive it with a normal transmitter.
If you drive it from the outside, you drive it at a voltage node, which means you drive it with high voltage, low current. Dollard used capacitive coupling in his longitudinal experiment, so that is probaly the way to go if you want to feed your sphere at a point at the outside. It may be a good idea to use a trimmer cap between your coil and your sphere, so you can tune the whole setup.
-- Arend --