|
|
Post by leviw on Aug 16, 2016 14:36:05 GMT -7
We're considering a site that looks promising but may cause our payload to drift a rather large distance. One of our estimates was 150 km away from the ground station (while at max altitude) which sounds like it would probably be too far away to maintain a connection. There are no mountains in the way and we can set up on a hillside with no obstructions. It would be impractical to set up in the middle because it drifts over a lake.
I vaguely recall the MSU team (or someone) saying they had connected to the payload from ~90 km away. Is that accurate, and how well have other people been able to connect to their payloads across long distances?
Trying to get a feel for how far is `too far`.
|
|
|
|
Post by Skylar MSGC on Aug 18, 2016 11:26:58 GMT -7
Well I don't think I ever said that considering the farthest I've had on connection on one of our flights is 72km. This was a test in South Dakota where we traveled downrange of the balloon and setup. A year and half before that We tested in New Mexico just point to point and got connection at 40 miles(~64km). During this test we hand pointed the system and traveled to another point downrange on the side of a large hill. This was before the tracking system had been conceived.
All in all I it depends on how calibrated the tracking system is on startup, but I wouldn't try to setup at a position farther than 75km from the balloon.
|
|
|
|
Post by David MSGC on Sept 11, 2016 14:42:23 GMT -7
LSU had recently (last week of Aug. 1st week of Sept.) flown the Ubiquity video system from HASP and they were receiving video out to 50 miles/80Km. Someone from LSU could probably give more details and were preparing a report as of Sept. 8th.
|
|
|
|
Post by DougGranger LSU on Oct 3, 2016 12:56:13 GMT -7
I am waiting on the final report from the students. We did get a signal at about 60 miles distance over ground. It was spotty but we did get something.
|
|
|
|
Post by colinrunnion on Oct 21, 2016 9:14:47 GMT -7
DougGranger
I was also told that you all used a grounding rod. What did this ground rod look like and how was it made? I have a launch tomorrow and would like to implement this.
|
|
|
|
Post by David MSGC on Oct 21, 2016 12:31:25 GMT -7
We have used a ground on our antenna that was a 7/16" or so diameter copper rod about 2 feet long beat into the ground. Then a wire was attached to the rod with a pipe clamp and the other end of the wire went to the antenna dish. The longer the rod the better, lol. The AC ground on most homes are about 6 -7 feet long but that is for high voltage AC grounding, the antenna you are just trying to get an RF ground so it does not need to be 7 feet into the ground. I would try at least 18-24" in the ground though. (you can get copper ground rods from Home Depot, Lowes etc)
|
|
|
|
Post by parkerswanson on Feb 11, 2017 12:53:26 GMT -7
Our ground station dish antenna may be subject to a good deal of rotation around the axis of its feed stalk. We appreciate that the Ubiquiti system has Horizontal and Vertical polarized "chains". In terrestrial applications the antenna would be oriented to match the Horizontal and Vertical polarizations at the other end of the link. But for our balloon-borne application, we expect that the signal from the payload will not be exactly polarized, since its Fat Shark antennas won't be exactly horizontal or vertical. My question is: should we expect some signal degradation as our dish antenna rotates? Does anyone have experience with this? Thanks!!
|
|
|
|
Post by David MSGC on Feb 11, 2017 13:54:47 GMT -7
I would imagine the payload would be rotating a lot more than your ground station antenna and the polarization of the fat shark antennas will be changing way more than that of the dish polarization. The dish is very directional while the fat shark antennas are omni-directional. Considering all this, my opinion would be the changing polarization and degradation from the dish moving would be negligible.
On the RFD system you can watch the received signal strength of both radios of a pair simultaneously using the 3dr program. I have watched the signal strength change while rotating the antenna laterally on one radio while the other was stationary and you can see the signal strength dip and rise with the matching and "unmatching" of the polarization of the antennas. The "degradation" in signal when unmatched was about the same as increasing the distance by a few miles. I have not done any similar experiments with the Ubiquity radios.
|
|
|
|
Post by Parker Swanson on Apr 7, 2017 21:15:53 GMT -7
Hi David,
I guess I didn't make myself very clear. Our ground station antenna really will be rotating quite a bit, because it will be on a ship at sea subject to yaw, pitch, and roll. (And BTW our payload will be rotationally stabilized per the excellent suggestions of the UMN team.) I know that in the usual terrestrial application of Ubiquiti Rocket-M systems, a good deal of attention is paid to aligning polarizations of the 2 data streams (H and V). But as you say the pair of Fat Shark drone-FPV antennas would be expected to be pretty much omnidirectional and of unpredictable polarization. We personally don't use the RFD 900 MHz system because it didn't add appreciable value to our project. We have managed to extract a relative signal strength indication of the 5.8 GHz Ubiquiti signal from its HTML output. We will continue to investigate. Thanks.
|
|
|
|
Post by ParkerWSwanson on Apr 7, 2017 21:27:35 GMT -7
Here's a general question:
We're trying to find out what performance we should expect from our video downlink 5.8 GHz system. We aren't satisfied with what we've done so far. I see on this board several somewhat vague reports of video reception, but we would VERY much appreciate more concrete and solid reports! So, please:
We're only interested in video from the balloon in flight to the "kit" ground station supplied at MSU as of July 2016. (So, we're not interested in ground-to-ground tests.)
We're only interested in cases where the ground station antenna was pointed entirely by the automatic tracking system provided. (So, we're not interested in hand-pointing or tweaking.)
We're only interested in tests in which your ground station received broadcast-quality video for an appreciable period of time (how long?).
With those interests in mind, we would like to know what range you have achieved. Thanks for your help!
|
|
|
|
Post by David MSGC on Apr 10, 2017 16:08:52 GMT -7
Hi David, I guess I didn't make myself very clear. Our ground station antenna really will be rotating quite a bit, because it will be on a ship at sea subject to yaw, pitch, and roll. (And BTW our payload will be rotationally stabilized per the excellent suggestions of the UMN team.) I know that in the usual terrestrial application of Ubiquiti Rocket-M systems, a good deal of attention is paid to aligning polarizations of the 2 data streams (H and V). But as you say the pair of Fat Shark drone-FPV antennas would be expected to be pretty much omnidirectional and of unpredictable polarization. We personally don't use the RFD 900 MHz system because it didn't add appreciable value to our project. We have managed to extract a relative signal strength indication of the 5.8 GHz Ubiquiti signal from its HTML output. We will continue to investigate. Thanks. If the antenna moves in any plane in which the horizontal or vertical line to the balloon changes by 2 degrees or more you will lose at least 3dB of gain, 2 degree change will result in half the signal strength received and this is best case assuming the calibration was perfect. The calibration of the IMU assumes the antenna will be stationary and its' location is only calculated once. The IMU is not used for pointing, only for initial calibration. If the antenna moves it would have to be re calibrated. You will not be able to use the system as is if the antenna is moving. You would have to use a closed loop control system to point the antenna from constant IMU and GPS data. |
|
|
|
Post by parkerswanson on Apr 10, 2017 17:22:39 GMT -7
Hi David, I guess I didn't make myself very clear. Our ground station antenna really will be rotating quite a bit, because it will be on a ship at sea subject to yaw, pitch, and roll. (And BTW our payload will be rotationally stabilized per the excellent suggestions of the UMN team.) I know that in the usual terrestrial application of Ubiquiti Rocket-M systems, a good deal of attention is paid to aligning polarizations of the 2 data streams (H and V). But as you say the pair of Fat Shark drone-FPV antennas would be expected to be pretty much omnidirectional and of unpredictable polarization. We personally don't use the RFD 900 MHz system because it didn't add appreciable value to our project. We have managed to extract a relative signal strength indication of the 5.8 GHz Ubiquiti signal from its HTML output. We will continue to investigate. Thanks. If the antenna moves in any plane in which the horizontal or vertical line to the balloon changes by 2 degrees or more you will lose at least 3dB of gain, 2 degree change will result in half the signal strength received and this is best case assuming the calibration was perfect. The calibration of the IMU assumes the antenna will be stationary and its' location is only calculated once. The IMU is not used for pointing, only for initial calibration. If the antenna moves it would have to be re calibrated. You will not be able to use the system as is if the antenna is moving. You would have to use a closed loop control system to point the antenna from constant IMU and GPS data. Hi Again David, Right - we know those things, such as the -3dB beamwidth of the Ubiquiti dish antenna. And, yes, we're modifiying the software so it takes IMU data frequently. Levi is in charge of that part of our system. When we're satisfied with our software we'll release it. (We're modifying a lot of other parts of it, also.) Thanks for keeping in touch! |
|
|
|
Post by David MSGC on Apr 11, 2017 8:17:47 GMT -7
If the antenna moves in any plane in which the horizontal or vertical line to the balloon changes by 2 degrees or more you will lose at least 3dB of gain, 2 degree change will result in half the signal strength received and this is best case assuming the calibration was perfect. The calibration of the IMU assumes the antenna will be stationary and its' location is only calculated once. The IMU is not used for pointing, only for initial calibration. If the antenna moves it would have to be re calibrated. You will not be able to use the system as is if the antenna is moving. You would have to use a closed loop control system to point the antenna from constant IMU and GPS data. Hi Again David, Right - we know those things, such as the -3dB beamwidth of the Ubiquiti dish antenna. And, yes, we're modifiying the software so it takes IMU data frequently. Levi is in charge of that part of our system. When we're satisfied with our software we'll release it. (We're modifying a lot of other parts of it, also.) Thanks for keeping in touch! Hello parkerswanson, I was unaware that you were working with levi and sorry if I repeated anything obvious, I was not trying to insult anyone's intelligence or knowledge. We will keep everyone informed of any changes/advancements. We are also working on pointing the antenna by using continuous IMU data but I don't think we will make much progress until sometime in May since all the work is done by students who are busy with school priorities until the first week of May. So right now it is kinda slow to progress. |
|
