Welcome to 3DRPilots.com

Sign up and join the largest 3D Robotics drone community!

  1. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    If you follow the FB Groups you probably already saw these. I have been doing some testing on some of the popular antenna combos.

    This is ground based testing and far from scientific.
    Solo Antenna Study - Google Drive

    This second study is a little more viable as it was done under a more controlled environment in actual flying conditions.

    Phase one I aquired a set of base lines using the stock leg antennas. My results are tabulated in a pdf- goo.gl/UV63fj

    Phase two will be after replacing the leg antennas with the rubber ducks off the controller to repeat the same test sequence to try and determine if there is any appreciable signal gain.
     
  2. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    Interesting, I prefer seeing numbers as opposed to word of mouth. My advisor would always say "I want numbers not adjectives"

    This is an okay start. Now you just need to repeat the measurements to be sure you're getting consistent values assuming you have control on all other variables (or their effects are negligible). At a minimum you can bracket signal strength for each case.
     
  3. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    - so it looks like you did repeat some measurements that were suspect. You should repeat every measurement a handful of times to be thorough and then take the average and compute standard deviation.

    If the uncertainty/error in measurement is greater than the difference between your test cases then all you've learned is you need a more precise and accurate way to measure signal strength.
     
    Damien and Wetstone like this.
  4. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    Thanks for the good guidence. Research is not anything that I have ever attempted and your advice is well taken.

    As to that so called "Ground Study" I suspect that repetition and averaging would help legitimize the results but factoring in ground interference and my use of a free analyzer on my mobile, I never expected much in the way of accuracy.

    The airborne study has more merit as it is based on a RSSI provided by the Solo. Accurate or not the data accumulated is derived from the same source and therefore a constant.

    I didn't elaborate but I took mesurements at each point after a minute or so of loitering and after the displayed value held steady. I did randomly recheck the signal strength values when returning back to the base at the same location and found them equal each time.

    Granted this was no NASA study and my outcomes should take my lack of experience and testing criteria into account.

    All things considered the chart does not provide any big surprises. I'll be anxious to see if there will be any positive (or negative) effect replacing the professionally engineered leg antennas with the ducks intended to be used on the controller.

    Besides providing accessible connections for possible additional antenna configurations such as the CP biquad mentioned in several threads, it might also help to insure that moving the antennas out of the legs so they can be replaced with retracts might prove useful.

    Thanks again for your input!
     
    backcountryice likes this.
  5. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    After your confession of more specifics, you are on the right path. Observing the signal for a time interval while loitering was another step in the right direction.

    If you really wanted to get into it, you could just pull the recent flight logs and analyze slices of the recorded signal data (where you hold position). I haven't done this, but from some digging the RADIO_STATUS: remrssi in the tlog is thought to be the controller radio signal. Find it, slice it, then directly compute the average over your "hold" interval.

    Note: There is also a 3dr-wifi.log but I haven't learned how to interpret it yet... there could be other logs for the radio as well.

    Another route is to make a plot of signal vs. distance which would be the most useful and easiest to interpret/compare directly for your test cases. I haven't come across a "distance-from-home" log, however, you could calculate this from the telemetry data. I may try this for fun because I am a data nerd.

    Most importantly, your preliminary results look as one would expect: signal weakens as distance increases and the higher power antennae show less attenuation.

    I love that the logs have so much data!
     
    Coastwise and Wetstone like this.
  6. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    Again, I appreciate your excellent input. Alas most of them as they apply to the logs are unfortunately above my understanding. I have experimented with reading the basic log data but not to any real understanding. That said, my albeit basic study, will hopefully expose my prime intention to see if replacing the stock leg antennas with the Ducks provided with the controller will have any appreciable positive or negative effect. The first phase of the study did not expose any surprises.
     
  7. mikediggs

    Joined:
    May 10, 2017
    Messages:
    43
    Likes Received:
    5
    What about omni directional antennas
     
  8. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    So I took a test flight yesterday for my spare 2.0 cube on open solo and then decided to also gather some distance vs. signal data (plots attached).

    I flew out from home point to 1500 m (5000 ft) at about 45-50 ft agl and returned at 80 ft agl (my RTL was set to 80 ft). I first flew out with the stock 3DR omni antenna and then repeated the flight using Alpha antenna. The tlogs were downloaded using WinSCP, exported to a text file using Mission Planner, then I parsed and pulled only the data I needed (simple process from a utility, I'll find the link for it). The distance values were calculated using a 3D distance formula:
    d=√[(x2-x1)^2+(y2-y1)^2+(z2-z1)^2]

    It was pretty straight forward since I found that the dx, dx, dz values were recorded in a mavlink parameter which meant I didn't have to convert lat/lon into decimal degrees and the home point was referenced to (0,0,0) which simplified the distance formula to:
    d=√[(x2)^2+(y2)^2+(z2)^2]

    I believe 'remrssi' is the remote rssi from what I've found on the interwebs. There is a conversion to go from the unlabeled (this is really really annoying to me) rrmrssi values to dBm but I have not discovered it as of yet. Supposedly there is documentation somewhere.

    In both test flights I never lost radio connection with my controller, however, video did cut out before reaching my maximum distance. I attribute losing video to a loose HDMI cable connection on the board (I took the solo apart the night before and didn't double check). When I reached 1500 m (5000') out, I hit RTL button and let the solo find it's way back to me. I should note that when I hit RTL the solo stayed in nearly it's same orientation as when I flew out.

    Nothing surprising in the results in regards to the relationship between signal attenuation and distance. There appears to be an anomalous dip between 600-900 m on the return flight for both cases which could be local noise. At 900-1000 m rssi appears to flat line, which could mean that it's dropped below a measurable level, radio is dropping more packets, and/or signal to noise is less than 1. The flat lining might be responsible for an "apparent" dip Not certain on this, still trying to understand it. There could be natural interference sources in play as well (haven't taken a magnetometer out there yet).

    Next test is to fly at a higher altitude where I suspect I will get less natural interference from the ground as well as multipathing.

    -I apologize for the Excel (cringe) plots. I was at home having breakfast while I pulled data so blame it on convenience and tools available. I won't do it again ... Unless it's another pre-work, breakfast analysis! tripall.jpeg [​IMG]
     
    Michael Lione likes this.
  9. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    Blue- Alpha antenna
    Red -stock solo antenna
     
  10. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    Well that sure puts my silly study to shame. I switched out the OEM Leg antennas for the ducks off the controller and ran through my test parameters to see how that configuration stacked up to the results I got last time out. Compared to what you did it's pretty lame. Bottom line, which was really the primary focus of the study is that there is no appreciable difference between the stock leg antennas and the ducks when used on the Solo. I suppose that's good to know for any that are adding retracts. I live in the hills and rarely fly beyond 1000'. Today I took the Solo out to 1500' one time and that was for me BLOS. All said and done I will default to the 6dbi Whips. The controller fits in the backpack with them installed and they provide decent connectivity. Here's Part Two of my study if your interested. Definitely nothing very exciting. Antenna Propagation Study - Google Drive
     
  11. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    Mike, that's interesting that there's no apparent gain and more importantly, not much (if any) loss of signal strength when you mount the ducks on the solo. Wonder if a different omni antenna would give better results.

    My test was arguably the lazy/easy way as all I did was fly out and back and watch a screen. I could do the same plot with your flight logs, no sweat. In fact, I found another parameter that records distance so there's not even a need to compute distances. Just throw them in your project folder with a note as to which log is which antenna configuration.

    Going forward, the orientation of the solo should be varied so that you'll mimic actual flying conditions. The easy way to do this is just fly around Willy nilly making sure you get to your max distance a handful of times and then just pull the flight data and plot it. In your methodology, you'd loiter at your distance interval and the perhaps fly around in a circle or few figure 8's.
     
  12. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    All but the Alphas in my study are omni directional.
     
  13. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    By the looks of your testing lab you definitely have an awesome enviornment to collect data. Wow!

    I hear you about moving around. That said, considering I don't fly beyond 1000' I doubt there will be any appreciable change in signal strength.

    Thanks for your offer to chart my tests. There will be 8 flights in all. I'll get busy pulling off the tlogs. If you can, plotting the OEM leg antennas with the ducks (each assigned different colors) for each combination would probably expose my focus which compares the two antennas on the Solo.

    One obvious result of my project is I now have outboard antenna connections to expieriment with different combinations. Am I correct that the antennas connected to the radio be of the same gain to maintain ballance? I purchased the 6dbi whips with all intention of using them together on the controller. Since the paddles are 7dbi I'm thinking I would have made a better choice to have found 7dbi whips to better match the Alpha when used in the mixed configuration.

    Finally one last question. What say you using the CP biquad on the Solo that has been discussed several times in other threads? It seems those that use it fly with one CP and one controller duck. Am I correct in my understanding that both the stock leg and ducks are 2dbi? Sorry, that was two questions. ;-)
     
  14. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    Plots are no problem
    I actually want to look more closely at the utility that parses the data to be 100% confident it's preserving the data correctly.

    I believe for antennae you want to up the power on both ends since you're only boosting transmission. Also keep in mind that Tx power is separate from Rx sensitivity and there's a balance to strike between the two.

    As far as in-depth antenna knowledge, I'm more of an observer and learning as I go so I can't comment on the CP setups or a hybrid combination. I have read a few threads a while back about CP antennae.

    My next test flight will be at a higher AGL (200' or above) for the full duration in hopes to mitigate any possible local interference.
     
    Michael Lione likes this.
  15. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    Just a heads up that I added the tlogs to the folder. I.m embarrassed to admit that hard as I tried I can't make head nor tails out of the logs so as to attempt to align them with which antenna configuration they represent. Additionally going by the .BIN files there are only two that were not overwritten from the first phase (June 7th) where I ran the test on the OEM leg antennas. Phase two (June 12th) was where I ran the same antenna configurations after swapping out the leg antennas with the ducks off the controller. So much for your plots. Bummer.
     
  16. Ralph E. Johnson

    Joined:
    Aug 21, 2015
    Messages:
    212
    Likes Received:
    34
    Hi guys!! Dumb question but have you tried the fplvr antennas?
     
  17. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    Hey Ralph. The main focus of my study was to compare signal propagation between the OEM leg with outboard mounted ducks used on the controller. I used a set of $7 6dbi omnis and the 7dbi patch (Alphas) in different combinations to come to the conclusion that there is no appreciable difference with the duck over the stock antennas.

    The fplvr setup is just one of several other alternative antennas that have gotten good reviews by several folks. I don't fly in heavy RF contention or BVLOS so I haven't any need for those solutions.

    As backcountryice graphically provided the Solo goes pretty far right out of the box.
     
  18. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    No worries Michael. I'll plot what you have available and we can see if we can sort it out from there. Which folder did you put the tlogs in? They aren't showing up for me, maybe check share permission?
     
  19. Michael Lione

    Joined:
    Jul 12, 2016
    Messages:
    144
    Likes Received:
    41
    Location:
    Tunkhannock, PA
    Howdy! (Not sure if you want your name listed here so I didn't) The tlogs are in a folder inside the main Propagation Study folder along with my table and photos. I just checked and sharing is enabled for the tlog folder the same as the others. Thanks again for doing this!
     
  20. backcountryice

    Joined:
    Apr 18, 2018
    Messages:
    29
    Likes Received:
    9
    Okay so the Drive mobile app is being stupid. I see them with a proper web browser now

    -Christian (or backcountryice)
     
    Michael Lione likes this.