How to get envelope from AM signal without phase shift
in [DSP]
|
Prev: How Rune Allnor can attend the COMP.DSP Conference
Next: Inertial navigation (anyone familiar with this stuff here?)
From: Eric Jacobsen on 30 Mar 2010 14:40 On 3/30/2010 9:56 AM, WWalker wrote: > Rune, > > You are using insults and ridicule as arguments. Are you afraid to discuss > this topic rationally? You obviously know a lot and I am sure you can > contribute. Why don't you join us in the discussion. > > I told you I have measured the nonlinear phase responce of a magnetic > dipole antenna using a RF Network analyser and it matches very well with > the curve in figure 9 of my paper, also in the Sten Paper a NEC simulation > shows the same results (Figure 3). There is no need to doubt the theory if: > theory, simmulation, and experiment match. > > William I'm curious as to how you characterized the antenna with an RF NA. I'd think that would be an extremely difficult thing to do properly. Antenna characterization takes some careful work and usually a special facility. Even then it's difficult. -- Eric Jacobsen Minister of Algorithms Abineau Communications http://www.abineau.com
From: Rune Allnor on 30 Mar 2010 23:48 On 30 Mar, 18:56, "WWalker" <william.walker(a)n_o_s_p_a_m.imtek.de> wrote: > Rune, > > You are using insults and ridicule as arguments. I am stating facts. If you find the facts insulting, you might want to change them - e.g. by learning the bascis of the subjects you pretend to discuss. > Are you afraid to discuss > this topic rationally? A rational discussion requires a rational party. Somebody who 1) States he can make information travel faster than the speed of light and only provides a numerical simulation as evidence 2) Claims, apparently in earnest, that he thinks the dipole is "difficult material" doesn't exactly qualify as simultaneously 'competent' wrt the subjects discussed, and 'rational'. > You obviously know a lot and I am sure you can > contribute. Why don't you join us in the discussion. "Us"? Every single poster here have pointed at at least one of your flaws, blunders and errors. My list of such approaches two dozens. Unless you make serious efforts to address such remarks, no one will 'discuss' with you. > I told you I have measured the nonlinear phase responce of a magnetic > dipole antenna using a RF Network analyser and it matches very well with > the curve in figure 9 of my paper, also in the Sten Paper a NEC simulation > shows the same results (Figure 3). There is no need to doubt the theory if: > theory, simmulation, and experiment match. Your theory is nonsense. Do the derivations of the exact solutions from scratch (I have told you how elsewhere), and you will find that there are no "faster-than-c" effects at all. Again, all this is trivial wave theory 101 material that any sane, competenet student will figure out in a couple of days or weeks. Rune
From: WWalker on 31 Mar 2010 11:09 Eric, I measured the nearfield magnetic dipole dispersion curve by using a 4 port 24GHz vector network analyser and measuring the transmission coefficient between two 1.5cm dia. magnetic dipole antennas separted 20cm apart. The network analyser was 4 port calibrated up to the antennas and the electrical length of the antennas was measured using the network analyser, the electrical characteristics of the antenna where calculated and both the electrical length and electrical characteristics were subtracted from the result, yielding a phase vs frequency curve that matches very well theory. The magnetic antennas are simply a solid shield coax cable, bent in a loop with the outer shields sodered together after the loop, and up to the cable connectors. The solid shield was severed in the center of the loop making the outer conductor a shield for the electric field but enabling the magnetic field to pass. The plane of the loops were parallel during the measurement. The measurement was made indoors but several meters away from metal objects. Placing metal plates a meter away did not affect the shape of the curve, only the observed noise of the curve in the farfield, not the nearfield. When the weather gets better here I will repeat the measurement outside. This should improve the SNR of the curve in the farfield. William >On 3/30/2010 9:56 AM, WWalker wrote: >> Rune, >> >> You are using insults and ridicule as arguments. Are you afraid to discuss >> this topic rationally? You obviously know a lot and I am sure you can >> contribute. Why don't you join us in the discussion. >> >> I told you I have measured the nonlinear phase responce of a magnetic >> dipole antenna using a RF Network analyser and it matches very well with >> the curve in figure 9 of my paper, also in the Sten Paper a NEC simulation >> shows the same results (Figure 3). There is no need to doubt the theory if: >> theory, simmulation, and experiment match. >> >> William > >I'm curious as to how you characterized the antenna with an RF NA. I'd >think that would be an extremely difficult thing to do properly. >Antenna characterization takes some careful work and usually a special >facility. Even then it's difficult. > > >-- >Eric Jacobsen >Minister of Algorithms >Abineau Communications >http://www.abineau.com >
From: Eric Jacobsen on 31 Mar 2010 11:25 On 3/31/2010 8:09 AM, WWalker wrote: > Eric, > > I measured the nearfield magnetic dipole dispersion curve by using a 4 port > 24GHz vector network analyser and measuring the transmission coefficient > between two 1.5cm dia. magnetic dipole antennas separted 20cm apart. The > network analyser was 4 port calibrated up to the antennas and the > electrical length of the antennas was measured using the network analyser, > the electrical characteristics of the antenna where calculated and both the > electrical length and electrical characteristics were subtracted from the > result, yielding a phase vs frequency curve that matches very well theory. Some sensor must have been used to pick up the magnetic field for the NA. How was that sensor calibrated? In situations like this it's often difficult to separate measurement of the Tx, channel, and Rx antennas. When claiming that the Tx antenna was measured, one has to be certain the effects of the channel and the Rx antenna were removed. The NA can calibrate the cables out by removing the antennas and connecting the cables together, but there is ambiguity between the Tx, channel, and Rx antenna, as they are difficult to separate. How did you do this? > The magnetic antennas are simply a solid shield coax cable, bent in a loop > with the outer shields sodered together after the loop, and up to the cable > connectors. The solid shield was severed in the center of the loop making > the outer conductor a shield for the electric field but enabling the > magnetic field to pass. The plane of the loops were parallel during the > measurement. > The measurement was made indoors but several meters away from metal > objects. Placing metal plates a meter away did not affect the shape of the > curve, only the observed noise of the curve in the farfield, not the > nearfield. When the weather gets better here I will repeat the measurement > outside. This should improve the SNR of the curve in the farfield. > > William How long was the NA sweep? There are ways to calibrate out the channel, but they're very difficult and time consuming. If there were known reflectors within range I'd think that'd be problematic for electric coupling, but perhaps not with magnetic coupling. Regardless, again, be careful that you're measuring what you think you're measuring. -- Eric Jacobsen Minister of Algorithms Abineau Communications http://www.abineau.com
From: glen herrmannsfeldt on 31 Mar 2010 14:52
WWalker <william.walker(a)n_o_s_p_a_m.imtek.de> wrote: > I measured the nearfield magnetic dipole dispersion curve by using a 4 port (snip) Last night I read chapter 21 of Feynman Lectures on Physics, Vol. 2. I would recommend that everyone following this discussion read it. The goal of that chapter is to connect the formula for the field from a moving charge to Maxwell's equations. Feynman claims to almost, but not completely, do that as, at one point, the math gets too complicated to fit into a book. He suggests that advanced students get out a lot of paper to go through that part. Among others that you can get from that chapter are the potentials and fields from a charge moving at a constant velocity. That will be pretty close to near field for a slowly moving charge, yet there are some non-obvious results. Consider this case: A charge is moving along the Z-axis with position (0,0,vt). That is, velocity v going through the origin at t=0. For an observer along the X-axis, at what time is the potential (or field) maximum observed? Two choices: t=0, or t=x/c (x being the position of the observer on the X axis.) As a hint, note that the Lorentz transformation was not derived to fit special relativity, but to fit Maxwell's equations. (Maybe that is why it is named after Lorentz and not Einstein.) -- glen |