Welcome to PhantomPilots.com

Sign up for a weekly email of the latest drone news & information

Reduced GPS Satellite Counts-Please summarize findings

Discussion in 'Phantom 2 Help' started by Case29247, Mar 19, 2015.

  1. Case29247

    Joined:
    May 7, 2014
    Messages:
    68
    Likes Received:
    17
    Location:
    Easton, MD USA
    First some background:

    Purchased Phantom 2 from B&H almost exactly one year ago. Phantom 2, 2-Axis Zemeus gimbal for Hero3, couple of extra flight batteries, props, carrying case, etc. Nice kit. A couple months later, added ImmersionRC video downlink and Black Pearl monitor. Very satisfied. All was well. Then in September, added iOSD Mini. All still okay. Love the flight telemetry.

    In all, based on the flight battery discharge cycles reported by the flight assistant software, I've flown about 100 times without "serious" incident. I have flown only in GPS and ATTI modes. I use my $30 toy quad toy for manual control mode playtime. The toy has taken quite a lickin', but it keeps on ticking. :eek:

    Come winter, I pretty much put the Phantom away. I didn't fly from early December through February. However, when I flew again in early March, I discovered that I was not routinely acquiring as many satellites compared to earlier flights. Where I easily acquired 7 or more satellites, and as many as 12 or even more, now I was getting 5 or 6, with the attendant flashing lights and OSD display warnings. I even had some automatic "return to home" activity...or at least that's what it seemed like while in GPS mode until I flipped to ATTI mode and took over. This continued for several attempts over several days.

    So, I turned to these forums and discovered I was not alone apparently. I studied the threads and took note of the information. Good stuff, a lot of it. Some nasty fits and turns, but mostly decent information. THANKS!

    I've taken the suggestions seriously and have performed some of my own repairs based on them. I've tried to be systematic about things, and have tested the results reasonably carefully.

    First, I opened up the housing again and carefully checked the GPS connector on the main board. It appeared to be completely seated and reasonably "tight." No problem there. I also inspected the male and female parts carefully.....Mine seem to be gold plated. That's good. Even though somebody knocked gold as a "lousy conductor", it's my metal of choice. Gold is in fact and excellent conductor. Not the best of the best (silver still holds the record), but at 2.5 x 10^-8 ohms/m, it's resistance is almost negligible (jezzzzz....10^-8, for gosh sakes), and when you're dealing with only a couple thousands of an inch of plating, it's as close to zero as you can get. And not that an ohm or two would make any difference. Now even a slight impedance mismatch.....that could make a difference with digital communication.

    So, I don't think my problem is the connector (resistance, impedance, or intermittent's). However, I did take the advice of some contributors, and I have jammed a small piece of tie-wrap end into the connector to further secure the two parts. I have also bent pins 2 and 4 ever so slightly out of alignment with the other two in order to add some tension to the assembly. Okay, so far so good. But still no improvement in satellite reception.

    Next, I tucked the lead wires between the GPS module and the main board underneath the factory shielding. Doing this has added 1 or 2 satellites. Great.

    Next, I covered the factory shielding with aluminum HVAC duct tape. Once again I got an improvement in satellite reception. Now I'm up to 8 in my driveway, and 10 or 12 under open sky. I'm actually starting to feel pretty good about things.

    (I live in a house with a mostly wooded yard, including some 100' oaks and poplars, which are now bare of leaves but still obscure a lot of sky. So each time I test, I first power up in my driveway, then drive a quarter mile to an open fallow corn field, then back to the driveway. I always get a couple more satellites in the open compared to the driveway. I have also done my modifications and tests as quickly as possible to minimize the variation in the number of satellites visible. I have tested the results of each modification, one at a time. I don't have props on the bird, but I cycle the motors anyway. I also turn off and on the Hero3, wi-fi, and start and stop video recording during each test (doesn't seem to make any difference in satellites). The method isn't perfect, but it's pretty good I think.)

    UPS brought me a sheet of adhesive-coated copper foil this afternoon. I plan to replace the aluminum shield with the copper. Personally, I don't expect this to make a difference v. aluminum, but we'll see. I also plan to cut a couple of strips of copper and encase the wire between the GPS module and main board in an extra layer of shielding. I will report on the results.

    I'm interested to see how the copper does. I'm trying to be sensitive to adding additional weight. Frankly, if I find the copper is no more effective than the aluminum, I'll probably go back to the aluminum.

    I have not installed a more sensitive antenna yet, even though some have reported to be helpful. However, I am really tempted. It looks really easy to remove the factory antenna and install the new one. We'll see. If anything, it's my ace in the hole if I need it.

    MY QUESTION: Can anybody with more experience than me provide a good summary of what the other threads have said? I've been all over the place reading, including some pretty long and convoluted threads. I think I've filtered a lot of good from the bad or useless, but I'm not fooling myself into thinking I've got it all.

    Thanks. I hope that what I've written will be useful. It is still a mystery why my satellite count has gone down. Maybe we'll never know....But if I can get the number up again, maybe whatever caused the change won't be a future operating factor anyway.
     
  2. jason

    Joined:
    Oct 12, 2014
    Messages:
    4,104
    Likes Received:
    404
    Location:
    Commerce Twp,MI
    What most haven't mentioned is in order to seat 4mm GPS antenna correctly you'd would have to remove most of the support ribs under the antenna. The only other modifications would be a variable tilt control and a yaw mod for very slow pans on the RC controller.
     
  3. Meta4

    Meta4 Moderator
    Staff Member

    Joined:
    Jan 8, 2014
    Messages:
    7,729
    Likes Received:
    3,448
    It's not clear from the OP that you understand that the number of satellites visible from any point is not constant.
    It varies quite a bit throughout the day because of the earth's rotation and the orbits of the satellites.
    The number visible can change by 4 or 5 sats throughout the day in the same spot so it's not possible to say that some modification added satellites.
    Then factor in the masking of part of the sky by terrain and vegetation or buildings.
    [​IMG]

    A handleld GPS or a GPS app on your phone will give a good display of the sats that are visible in your sky and you can see how this changes with time.
    This one is from the Android app GPS Status
    [​IMG]
     
  4. Case29247

    Joined:
    May 7, 2014
    Messages:
    68
    Likes Received:
    17
    Location:
    Easton, MD USA
    While I do not pretend to be an expert, I have used GPS systems for many years and understand that the network of satellites is in medium high orbits, which means that they are not "stationary" over any place and the numbers of "visible" satellites varies over time. I also am aware of the satellite predictors available, and also have apps on my mobile devices (one for instance that looks exactly like the one above on an Android phone).

    Having said this, in my description of what I have done, I made every effort to do them quickly and conduct successive tests quickly one after the other. I know this isn't, strictly speaking, covering all the bases, but I'm pretty comfortable with the results.

    Also, as an update----this evening I removed the aluminum shield that I had previously installed and fashioned a copper foil shield replacing the aluminum. I covered the foil with clear packaging tape for insulation.

    In testing, I got 10 satellites within about a minute in my driveway with all the woods, and 12 in the nearby field under a completely open sky. This is higher than I have seen since I first had my reduced satellite problem. From 4-5/6-7 to 10 and 12.

    I elected NOT to fashion a separate shield for the wire between the GPS module and main board, because of the complication possible contact between the bare copper foil and the circuitry on the pc board.

    I will also conduct further testing over time and keep all informed. I will be rather surprised and really disappointed if the number of satellites goes down into "problem numbers." But I do know there's a possibility that this evening's test was just a fluke with more than the average number of satellites being visible. Time will tell.

    I gave more thought to the concept that impedance mismatches might be contributing to the problems some people are having. Just like light reflecting off a pane of glass, any impedance mismatch at connectors, for instance, will reflect part of the signal power back toward the source. With digital data, constructive and destructive interference between the forward signal and backward reflections might cause real confusion....even turning one's into zero's. When this happens, you can guess what happens to the integrity of the data flowing through the wire. Anybody with experience with LAN's "back in the day" when all the computer were connected serially to a single coax line will remember just how important impedance matching and even the length of each cable segment between computers was. Those pesky standing waves and reflection interference.

    As for the installation of the more sensitive antenna, I am aware that you must remove some of the structural ribs on the housing to make clearance for the thicker antenna. That's actually the main reason why I haven't just gone ahead and installed the new antenna.

    Also, FWIW, my bird now weighs 1304 grams with gimbal, camera, and the other stuff I've added. That's getting pretty close to DJI's maximum recommended weight.

    I'll get out my phone or tablet in subsequent tests and keep track of the satellites. I'll keep you informed of anything new I discover.
     
  5. RichWest

    Joined:
    Nov 1, 2014
    Messages:
    405
    Likes Received:
    43
    Location:
    Houston, Texas
    Meta4's post is something to heed in basing your results, you'll drive yourself crazy otherwise.

    If you're on a mission to cure the GPS issue, then you got the right process - keep going.

    I think you'll find copper to work best over aluminum. I don't see the reason to insulate the exposed surface however, at least at this point. It's unneeded weight in my opinion.

    Really interesting to read your steps and results, thanks for sharing.
     
  6. IflyinWY

    Joined:
    Sep 12, 2014
    Messages:
    3,470
    Likes Received:
    939
    Location:
    Where the deer and the antelope play
    Hey Case,

    Good posts, bad to better news, and it sounds like you're on the right track. But why the heck are the sats more difficult for your bird to see. Meta4 offered one possible answer which shouldn't be discounted, not that I thought you did. But WTH?

    I can't imagine you left your bird out back in the shed for the winter, so what changed. Connections corroded? Dust or film on the exterior skin of your Phantom? I doubt it.

    As I understand it, your most recent upgrade/mod (prior to having satellite issues) was the IOSD. If that is correct, I would suggest finding a way of shielding that unit so it can't cause interference.

    I would like to know just what bird you are flying, I know it's a P2. You folks amaze me with your smarts. I really enjoying reading intelligent discussions, sorry if I've been one of those to post "Some nasty fits and turns,".

    I'll be watching and wish you the best.
     
  7. dirkclod

    dirkclod Moderator
    Staff Member

    Joined:
    May 9, 2014
    Messages:
    10,633
    Likes Received:
    5,698
    Location:
    Amory Mississippi
    Case29247 very little of the plastic has to be removed from the ribs to install the new antenna If ya decide to do it .
    I just used an xacto blade to trim where the outsides of the antenna sits and was no problem .
    Here is some reading on it viewtopic.php?f=27&t=32311#p295495
     
  8. J.James

    Joined:
    Apr 21, 2014
    Messages:
    1,882
    Likes Received:
    162
    Location:
    Upstate NY
    I thought the same thing about the insulating the foil being unnecessary intill a few weeks ago on a p2 I just got done building that i had added a layer of adhesive backed copper foil with out adding a patch of tape to the sticky side just for extra protection and every thing was working fine just like always BUT when I went to close the shell up some of the foil got pushed up right against the bottom of the GPS puck and some thing on the gps's board scratched threw the thin layer of adhesive and shorted out the GPS and my compass. and I was getting compass errors AND even worse my compass on the leg was getting really hot with in a few seconds any time i powered on the phantom. luckily It didn't burn any thing out.

    But peeling back the foil and seen it was the 5vdc shorting out in to the other connections that the plug were the compass goes in to the gps. But putting a layer of thick packing take over the spot were it could contact the gps fixed the problem.

    and luckily any other ones I own or ever repaired all have the added layer of the stock foam and copper tape as the first layer.


    and Im glad it happened on the bench and not while in the air and flying. That would not of been good.
     
  9. saltire

    Joined:
    Jan 24, 2014
    Messages:
    241
    Likes Received:
    6
    I tried a 4mm antenna and made no difference to the sat count or the speed it found them .Having tested an Iris+ today a little experiment was done and the Iris + within the blink of an eye had 10 sats and stable and my phanton 2v also got 10 but 3 mins later and and would constantly drop to 7 or 8 then back up to 10 where as the Iris + just stayed stable on 10 might be that other company's are using better Gps antennas ? .
    Case29247 When you have sats locked have you seen any difference when switching on the video recording?
     
  10. yorlik

    Joined:
    Feb 9, 2015
    Messages:
    547
    Likes Received:
    91
    I often contribute to fits & turns, so apologize up front for making some of your extra reading... Even though I am an electrical engineer, extra class ham, FCC first class licensee, I do not have a lot of experience at GPS wavelengths - just enough to get into trouble. That I did, by suggesting clouds and precip affected count. I have since learned these do not have MAJOR affect on count, but rather they have SOME affect. That is why I blamed them for MY count issues; when I had extremely marginal GPS reception, with my OEM 2mm antenna, these small weather effects indeed frustratingly reduced my lock counts. Then I learned that my beloved 11 year sun spot cycle is tracked today as KP index, and it too has potentially major affect on count lock in. So with your marginal lock count, you too may have been effected some by clouds & the high K index happening recently? Now that I have 4mm antenna, my count is no longer as marginal, and even the high K index does not seem capable of lowering my count. Something to consider why you should go to this antenna.

    I'm with IflyinWY: what changed?
    I often ask exactly that of my customers with broken machines. I tell them "the last thing they changed, or happened, most often is the cause of the failure." You say you had good sat count, stored it for 2 months, and now counts is down. Seems like something had to change. Any chance the kids dropped the snow shovel on top of it when getting it down off the hook and that put a hairline crack in the 2mm thick fragile GPS antenna? Something else?

    The 4mm upgrade antenna absolutely does NOT require any rib cutting or changes to the housing - if installed at proper 45 degree angle on PCB. Picture attached for reference. Did I leave mine that way? NO, since I was not 100% convinced it may not play 100% with the pcb ground plane design.

    The 4mm upgrade antenna absolutely does not require but ONE small rib cutting if installed in same orientation as original. Picture attached for reference. Only a very small amount of ONE rib trimmed back with dikes.

    Last idea: Have you followed the various threads where people swear their GPS count is now significantly LESS - due to firmware update from 3.08 to 3.12? Did you? Might THIS be THE CHANGE?
     

    Attached Files:

  11. Case29247

    Joined:
    May 7, 2014
    Messages:
    68
    Likes Received:
    17
    Location:
    Easton, MD USA
    I don't really have a major aversion to removing the internal ribs on the cover to make room for a thicker antenna module. I also saw pictures showing rotating it 45°to miss the ribs. I just thought it best to take things a step at a time, and for no particular reason, I just put a new antenna at the bottom of the list.

    I too know just enough about radio wave propagation at various wavelengths to be dangerous, but I do know that we are totally bathed in them, and just like the choppy surface of a swimming pool, the electromagnetic field we're bathed is in constant flux. I have tried to be careful to run my driveway and open sky tests at each location within inches of each other to minimize the effects of nearby reflecting and absorbing factors. Still, I know my procedures are really quite loosey-goosey relative to real research. No PhD thesis will come out of what I'm doing. :roll:

    As for insulating the copper foil with that 2" wide, clear packaging tape we're probably all familiar with, I just thought it a good trade-off between caution versus weight (maybe a gram or two?). Having had a career as a chemical engineer and working with some pretty hairy stuff at times, I tend to be a "belt and suspenders" type of designer.

    As for what happened over the winter, that's a mystery. Inside and in a hermetically sealed case all winter. Perhaps I just didn't notice the problem or anything that hinted that the problem was looming earlier. I wish I could hit rewind and replay past flights again.....

    I installed the iOSD Mini module at the end of the summer and flew quite a few times with it. I admit that I fretted at the time about about mounting the iOSD module close to other circuitry, plus running gobs of little wires all over the place. I even unmounted the CAN-bus connector outside on one of the landing legs and brought it inside the housing to make the overall installation a little neater and compact. I did not shield anything from anything. But it's on my mind if problems persist.

    No, I see no significant difference in satellite numbers when with the Hero3 off/on, recording/not recording, or wi-fi off/on. Nor with the flight motors off, running up, or idling, etc.

    I also saw reports suggesting that firmware upgrades might be a factor. If all else fails, I'm keeping the idea of a rollback in mind.

    FWIW, here's a shot of what the iOSD installation looks like. Note the Can-Bus connector inside near the lower-right hand corner of the photo. Also, the wad of little wires at 12 o'clock tucked down along the side of the main board. It's pretty crowded in there.

    [​IMG]
     
  12. Case29247

    Joined:
    May 7, 2014
    Messages:
    68
    Likes Received:
    17
    Location:
    Easton, MD USA
    This morning, I went out and actually flew the bird for the first time since I started my investigations and installation of the copper shielding and rerouting the cable between the GPS board and main control board under the shielding. Initially, I was disappointed because I was only getting 6-7 satellites compared to a couple more that I had gotten during the ground testing I conducted while doing the work.

    The numbers were generally 6-7, but they also dipped to 5 briefly (with the attendant red flash on the flight lights), and I also saw 8 now and then. Operating the Hero3 and wi-fi didn't seem to affect the numbers, which was good.

    What I did differently this time (and I am really stupid not to have been doing all along) is that I fired up my Android phone and monitored the satellites it reported along with the P2. The phone reported 8/8 and 7/8 satellites.

    So, on the one hand, it's good that the P2 was receiving numbers in the same range as the phone was reporting. On the other hand, it's disappointing that the P2 generally received lower numbers. But at least I had a benchmark against which to compare the P2 numbers.....something I should have been doing all along.

    I'll fly with 6-7 solid locks any day, but I was still disappointed to see the numbers down from my ground testing a couple days ago. I definitely didn't like seeing 5 again, even if only for a couple seconds. I'll keep on and report what I am finding with the phone satellite numbers for comparison. Maybe it's time to order the more sensitive 4mm antenna.

    I appreciate feedback and additional tips. Maybe I'm just expecting too much. Maybe not.
     
  13. Case29247

    Joined:
    May 7, 2014
    Messages:
    68
    Likes Received:
    17
    Location:
    Easton, MD USA
    I just went out and did another quick test because my Android phone GPS monitor was showing 10/10 satellites. The diagram showed five almost overhead, and five low on the horizon.

    In my driveway, the P2 picked up 5-7 satellites, mostly 6's and 7's. In the field with the wide-open sky, the P2 locked in on all 10 satellites, and even flashed 11 for a period. So all seems quite okay for the satellites.

    But I had a strange experience out in the field. I decided to fly a little while. All was normal for a few minutes, but then the controls started acting strange.

    The flight started normally for a minute or three. Up and down and yaw worked fine. Right and left worked fine. But after a short time of normal, forward and backward on the sticks started reacting as right and left.

    I guided the bird home and landed. I didn't shut down the bird, but I cycled the power to the controller (the standard white controller that comes in the kit), and then flew again. The controls returned to normal. Flew around for about ten minutes...all okay.

    I'm going to post this to a new thread.
     
  14. J.James

    Joined:
    Apr 21, 2014
    Messages:
    1,882
    Likes Received:
    162
    Location:
    Upstate NY
    The last few days with the super high k index even my phantoms with the up graded 4dbi gps antennas are taking a few minutes to get even just 6 or 7 satellites and even when flying are only picking up about 9 or so instead of 12 or 14

    Also even with an upgraded gps if its been more then 30 days sence one is turned on its got to do whats known as a cold start and has to down load all the data from the gps satellites for there position and also there projected times and locations that are stored in the ublock chip that when stored and updated makes it easier to get a lock being it knows the projected locations already when doing whats called a hot start. Some times a gps devise thats doing a complete cold start and has to down down all the data can take up to as long as 12 minutes for it to get the entire file that has all the data for all 24 gps sats locations and projected locations at any giving time in the future for were they should be in orbit in the next 30 days.

    If that tiny hearing aid battery on the gps puck completely dies its got to do a cold start every time you turn on the phantom being it cant store the data on the chip when you power off the phantom if that battery is dead
     
  15. kirk2579

    Joined:
    Jan 13, 2015
    Messages:
    1,336
    Likes Received:
    262
    Location:
    Girard. ohio
     
  16. yorlik

    Joined:
    Feb 9, 2015
    Messages:
    547
    Likes Received:
    91
    its still a mystery to me. they say heavy cloud cover doesn't effect being able to lock on, but I tried to fly tuesday 3pm from same spot on driveway as many times before where I had 7-8 sats with my 4mm antenna, my sat app said 10-11 easily seen, yours above says 10 (@10degree which should be good for my tree covered horizons), and I could not even get past 5 hovering up at 30ft in ATTI waiting on them. Kp was low. Had to abort. But those darn thick clouds are said not to bother it....
     
  17. kirk2579

    Joined:
    Jan 13, 2015
    Messages:
    1,336
    Likes Received:
    262
    Location:
    Girard. ohio
    I actually have the opposite issue it seems at times. give me a cloudy, snowy day and boom ! lock I was out almost all of feb and march in snow and clouds. didn't know what sun looked like.

    yesterday was 65 and sunny clear as can be. I had to put my plus away till evening , no lock--- however my p2v1 with 3.08 did lock in few minutes but went in and out and finally stayed locked so I flew it.
    The plus (3.12) would not lock for more than a moment or 2.

    both have 4mm upgraded antenna.

    not being happy about the gps I opened it back up and redid my alum shield tape under cover, this time I went all down arms to screw post etc.

    Thinking that I recently did the wifi fix and cut the tape around module I added even more tape around module seams and even at plug areas on side.
    Looking at it I decided to add more shield tape to the gps lead wire all the way to black plug.

    then I cut a small piece of shield tape and placed around the actual black female gps plug. slid in between compass plug and around completely.

    this leaves almost no part of gps feed to naza unshielded.

    put it in drive at test spot (same spot for gps test always) hit 6 in a minute or so and 10 shortly after that.

    It appears that my shielding was a bit weak after my wifi fix.
     
  18. RichWest

    Joined:
    Nov 1, 2014
    Messages:
    405
    Likes Received:
    43
    Location:
    Houston, Texas
    Just for the sake of shared information, I direct you to a solution that worked for me...I still have the 2MM puck...and lock 10+ in very little time.

    viewtopic.php?f=27&t=33453
     
  19. Mike_in_Letcombe

    Joined:
    Mar 25, 2015
    Messages:
    75
    Likes Received:
    29
    Location:
    Letcombe Regis, Oxon, England
    I've actually had the opposite; updated my V3.0 to the latest version (3.12) and have now found I am routinely getting 9 or 10 satellites, compared to only 7 (if lucky and stood around long enough) before. Now feel a lot more confident in sending the thing up in the air and that it might, if needed, actually return to me if signals lost. Standard kit, I haven't opened mine at all as I believe this will invalidate my warranty and I had to send the first one back, due to range issues (was getting around 40 metres max!)

    I'm in Southern England, whereas most on here appear to be from the US.
     
  20. dirkclod

    dirkclod Moderator
    Staff Member

    Joined:
    May 9, 2014
    Messages:
    10,633
    Likes Received:
    5,698
    Location:
    Amory Mississippi
    Look at the clouds in this screen shot and then the Sat count . Clouds don't effect what ya pick up .
     

    Attached Files: