Our second attempt for setting the new microwave transmission world record distance was unsuccessful. We did not reach our goal to connect the antennas over the world record distance, but we gained a lot of experience and valuable information from the data we gathered.
It is not easy to set a world record – indeed.
FECORE made a 2nd attempt to set a new 5GHz microwave transmission world record in Hungary, on the 20th – 21st of September 2019.
The two positions were the same as last time, giving a 307 km distance between the two towers:
3 people were on the Pecs side, Stefan P, Gabor and Attila a Hungarian microwave professional,
2 people were on the Debrecen side, Sandor Szekely with Viktor a Hungarian microwave professional.
On Friday 20th September we installed the new AF5XHD radios with the larger antennas on the towers by 11 AM and set the vertical alignment to 0°. We aligned the antennas to our reference line, heading 241° on the Debrecen side and 58° on the Pecs side according to google data. The heading was confirmed by using landmarks from Google Earth and with a compass on the ground.
We used the Ubiquiti spectrum analyzer software to monitor the frequencies and the Ubiquiti alignment tool.
We started with the Pecs antenna as the slave monitoring the spectrum analyzer and the Debrecen antenna as the master broadcasting the beacon signal. Below is a picture of Stefan at the Pecs location.
By turning one antenna at a time, we first swept the antenna at Debrecen in small increments of about 1 degree from our established reference line. The antenna at the Debrecen side was then swept a total of 20° from the far left (at 231° SW) to the far right (at 251° SW) whilst transmitting the beacon signal and the Pecs side scanning with the spectrum analyzer. The Pecs side then rotated the antenna to a new heading by turning approximately one degree, and the Debrecen side started to sweep the antenna in small increments backwards from the far right to the far left position.
After testing several headings we switched the master to the Pecs side transmitting the beacon signal and Debrecen became the slave. Again we used the spectrum analyzer software and the built-in alignment tool.
None of the configurations resulted in a connection. We tried several different frequencies, bandwidths, and RX transmission rate settings without a successful connection. At 3:30PM we came down from the towers as the wind was increasing and the temperature was cold.
We set up an FECORE meeting to brainstorm on the experiences of the day and to look for new ideas to develop for the next day.
Our focus was on the horizontal heading precision. The question arose, “Can we trust the google maps data?”
It showed a heading of 241° SW from Debrecen and 58° NE from the Pecs side.
Modelling the heading on the Azimuthal Equidistant (AE) map showed a few degrees difference to the google data, equaling 244° SW on the Debrecen side, and 61° NE on the Pecs side.
Red line – google heading, Blue line – AE heading, Yellow lines panning range
On Saturday we rotated the antennas with an electrical turntable pole controlled from ground level on both towers. The antennas were moved horizontally at approximately 2.5° increments and the relative heading was shown on the display of the turntable equipment.
We swept across the +-15° range from the new reference line using AE headings in small increments one antenna at a time just as we had done on the previous day.
We worked from 9 AM until 5 PM without being able to establish a link between the two locations.
After evaluating the test, we found the following possible reasons and discussed some potential solutions:
Is the AF5XHD unit with this antenna powerful enough for the 307km distance?
According to the manufacturer’s specifications, it is designed for such long distances.
We are going to test the unit on shorter distances of 10-20km, 50-60km and 100+km to measure power loss over distance. We can measure transmitting power and path loss over these distances to calculate dBm level for new the world record distance.
Was the horizontal aiming precise enough?
To establish the connection between the antennas, we need to be within about 3° LOS (Line of Sight) with a calculated signal loss of 3dB. We swept an area of 30 degrees from both locations several times within the 3° range. So we may conclude that the alignment was precise enough to establish the connection.
We found out from forum discussions on the topic that the Ubiquiti spectrum analyzer software may not be showing real time data and thus may not be quick enough to register alignment within the 2 minutes we paused at each heading position. It appears that it may be required to wait 6 minutes for Ubiquiti spectrum analyzer software display updated data.
We will test the Ubiquiti spectrum analyzer software for delays and check availability of analog scanning devices and a 5GHz signal generator to use for our next alignment attempts.
The electrical turntable was a great tool to set azimuth angles and made sweeping the area very easy. We are planning to design a new instrument we will call SAMAD for Super Accurate Microwave Aiming Device to automatically sweep a specified area defined from the reference line both in horizontal and vertical mode. SALAD was vital to establish extremely long distance laser observations. Perhaps a world record microwave connections will require the SAMAD.
Was the vertical aiming precise enough?
To establish the connection between the antennas we need to be within about 3 degrees of LOS (Line of Sight), both vertically and horizontally, with a calculated signal loss of 3dB. The Ubiquiti software has a 0.1° precision vertical alignment tool, that was confirmed by our microwave expert’s own measuring device. We set the vertical to near 0° both at the Debrecen and Pecs side.
The difference between the AE map headings and the Google Earth headings is 1.5°. That is within the 3° required precision range of the antenna.
The LOS propagation of microwaves is not necessarily a straight line, it may be deflected by atmospheric conditions, especially in the lower atmosphere near the ground.
The Pecs side is at an altitude of 626 meters MSL, and Debrecen is at 169 meters MSL. The beam on the Debrecen side is only 30 meters from ground level with forests along its path. 25km away from the Debrecen tower the ground level drops 30 meters to 100 meters MSL giving the microwave beam extra clearance from the ground.
The beam at Debrecen side was near the surface where optical density gradients increase with elevation. Leveling refraction may have caused the microwave beam to bend upward above our antenna. We observed the upwards refraction phenomenon on several occasions during the 2016 Balaton laser experiment. The picture below is from the 2016 Balaton observations.
As seen on the pictures of the Chicago skyline from the other side of Lake Michigan the bottom part of the buildings are hidden. We theorize that it is caused by the light bending upwards, due to leveling refraction. This lower elevation optical density profile may be an impenetrable layer for the microwave beam, similar to light. We will do peak to peak mountain tests to avoid lower elevation atmospheric effects in the future.
Did an obstruction in the path of the MW beam cause signal loss?
We checked the whole path of the beam carefully to look for any possible obstructions in the planning phase of the test, and the path of the beam is clear from any high rise buildings, or hills. The beam is crossing a low elevation agricultural area of Hungary, from 70 meter MSL to 115 meter MSL. It starts from Pecs at an altitude of 626 meter MSL and slowly descends with -0.08 degree angle downwards to Debrecen. The critical elevation starts 25km from the Debrecen tower, where the ground level rises to 130 meter MSL. This portion of the beam is passing south of the Debrecen airport with some farmhouses but mostly trees. The beam center clearance reduces to about 20 meter.
In conclusion FECORE will now improve the MWDT (Microwave Distant Targeting) method with the SAMAD and test the antennas over shorter distances to prepare for the next world record attempt.
We try, we fail, we learn, we succeed!
(Written by Sandor Szekely with editorial changes by Robert Scott)