Testing report
Yesterday, I was able to test my antenna's using a vector network analyzer. The results were surprising and therefore interesting.
First, we tested the naked antenna for which we already knew it had a nice SWR ratio at the designed frequency. After some tuning, we got an almost perfect SWR at the desired 1296 MHz frequency, and we made a plot of the measurement:
At the left of the image, you see the nice dip at the designed frequency, while you see another, wider dip at around 1750 MHz. When we divide 1296 by 1,57, we should get the transverse resonance frequency of our antenna, which computes to 825 MHz, so the first higher harmonic should be at about 1650 MHz. If you woud consider the reflector plane as a capacative load, one would expect the resonance frequency to become lower, so it is interesting that the first transverse harmonic appears to be at a higher frequency than expected.
After we tested this antenna, we went over to the design with wave guide. Of course, Murphey came along and it turned out that my connector was NFG and that the solder between the dipole and the reflector plane was broken, so I first had to heat up my soldering iron and fix that. Fortunately I took my soldering iron along and there was a guy selling the right connectors for a nice price. It were used ones, but much better and cheaper than the new ones I had.
When we finally were able to test the antenna with wave guide, we found out to my surprise that it did not come even close to any decent SWR at 1296 MHz. Hmm. Maybe I had to add some top-plate? Let's try that. I quickly pulled a 20 cents coin out of my wallet and soldered that on top of the whip, to make another try. NFG again.
Then, more or less by accident, we found out that when we pushed the dipole antenna out of its socket into the wave guide, we DID get a dip.
So, I completely removed the reflector plate and we did another measurement with the naked antenna. Again to my surprise, we did NOT see a dip in the SWR at 1296 Mhz.
Then we moved the antenna into the waveguide, from the bottom. Like this:
High-res version: http://www.tuks.nl/img/Lamare_Waveguide_probe.jpg
This time, when we moved the probe/antenna into the waveguide, we DID see a dip, although it was very hard to tune the thing to 1296 MHz, because the lower part of the dipole was too short without the socket, and the socket was hanging loose on there and tended to slip away, etc. All in all, it was a bit touchy.
Still, it was very interesting to watch the screen when moving the antenna in and out the hole in the bottom of the waveguide, and it seemed that the length of the part of the dipole inside the waveguide determined the frequency of the dip. The more of the dipole that stuck out at the bottom, the higher the frequency of the dip.
Another observation was that the length of the wave-guide did not seem to have any noticable influence on the size/place of the dip. This is confirmed by the "microwave antenna book":
http://www.qsl.net/n1bwt/chap6-3.pdf
BTW, the diameter of my waveguide was 10,5 - 11 cm, which means it should NOT support any normal waveguide mode at 1296 MHz:
According to this, at 1296 MHz, our waveguide should have a diameter of AT LEAST 300/1296 * 0.59 = 13,7 cm. So, the fact that we DID get a nice dip at around 1296 MHz is very interesting...
Of course, we did not have a proper mounting of the probe or anything, but we did manage to make a nice plot of the situation skeched above:
Interestingly, when I moved the antenna without reflector plane into the outer shell of the waveguide, just a hollow tube, we did NOT see a dip around 1296 MHz.
So, it seems that the reflector is necessary in order to get the longitudinal wave to radiate, because the antenna without ground plane did not seem to produce anything near a decent SWR, which was measured by measuring the impedance of at the connector of the antenna, btw.
It could very well be that the antenna without reflector DOES resonate, but NOT radiate, something I was worried about before:
Yesterday, I was able to test my antenna's using a vector network analyzer. The results were surprising and therefore interesting.
First, we tested the naked antenna for which we already knew it had a nice SWR ratio at the designed frequency. After some tuning, we got an almost perfect SWR at the desired 1296 MHz frequency, and we made a plot of the measurement:
At the left of the image, you see the nice dip at the designed frequency, while you see another, wider dip at around 1750 MHz. When we divide 1296 by 1,57, we should get the transverse resonance frequency of our antenna, which computes to 825 MHz, so the first higher harmonic should be at about 1650 MHz. If you woud consider the reflector plane as a capacative load, one would expect the resonance frequency to become lower, so it is interesting that the first transverse harmonic appears to be at a higher frequency than expected.
After we tested this antenna, we went over to the design with wave guide. Of course, Murphey came along and it turned out that my connector was NFG and that the solder between the dipole and the reflector plane was broken, so I first had to heat up my soldering iron and fix that. Fortunately I took my soldering iron along and there was a guy selling the right connectors for a nice price. It were used ones, but much better and cheaper than the new ones I had.
When we finally were able to test the antenna with wave guide, we found out to my surprise that it did not come even close to any decent SWR at 1296 MHz. Hmm. Maybe I had to add some top-plate? Let's try that. I quickly pulled a 20 cents coin out of my wallet and soldered that on top of the whip, to make another try. NFG again.
Then, more or less by accident, we found out that when we pushed the dipole antenna out of its socket into the wave guide, we DID get a dip.
So, I completely removed the reflector plate and we did another measurement with the naked antenna. Again to my surprise, we did NOT see a dip in the SWR at 1296 Mhz.
Then we moved the antenna into the waveguide, from the bottom. Like this:
High-res version: http://www.tuks.nl/img/Lamare_Waveguide_probe.jpg
This time, when we moved the probe/antenna into the waveguide, we DID see a dip, although it was very hard to tune the thing to 1296 MHz, because the lower part of the dipole was too short without the socket, and the socket was hanging loose on there and tended to slip away, etc. All in all, it was a bit touchy.
Still, it was very interesting to watch the screen when moving the antenna in and out the hole in the bottom of the waveguide, and it seemed that the length of the part of the dipole inside the waveguide determined the frequency of the dip. The more of the dipole that stuck out at the bottom, the higher the frequency of the dip.
Another observation was that the length of the wave-guide did not seem to have any noticable influence on the size/place of the dip. This is confirmed by the "microwave antenna book":
http://www.qsl.net/n1bwt/chap6-3.pdf
The attraction of a coffee-can feed is that it is so simple to make, a matter of finding the right diameter can and soldering a coax connector with probe to one side. Horn length is not critical, but should be more than one waveguide wavelength to eliminate any stray modes launched at the probe transition. My experience with probes in circular waveguide is that they are touchy to adjust and get right; the horn impedance varies with changes in both horn diameter and horn length, so any modification can upset the VSWR.
The horn must be large enough for the lowest order waveguide mode, the TE_11 mode, to propagate. The cutoff wavelength for this mode is 1.706 × Diameter so the minimum waveguide diameter is 0.59 λ.
Of course, we did not have a proper mounting of the probe or anything, but we did manage to make a nice plot of the situation skeched above:
Interestingly, when I moved the antenna without reflector plane into the outer shell of the waveguide, just a hollow tube, we did NOT see a dip around 1296 MHz.
So, it seems that the reflector is necessary in order to get the longitudinal wave to radiate, because the antenna without ground plane did not seem to produce anything near a decent SWR, which was measured by measuring the impedance of at the connector of the antenna, btw.
It could very well be that the antenna without reflector DOES resonate, but NOT radiate, something I was worried about before:
Originally posted by lamare
View Post
Comment