Blogs

u-blox NEO-6P GPS chip

Posted by Antonie Kruger on November 16, 2012 at 12:30am

I3689487203?profile=original was just notified by a supplier of gps modules in RSA that this product has become available locally(RSA). I'm not sure how long this has been out in the rest of the world, but the following statement had my attention : NEO-6P GPS module for Precise Point Positioning....OK, what is PPP, and how accurate is it exactly. The guys at ublox claims positional accuracy of <1m. http://www.u-blox.com/en/gps-modules/neo-6p-/neo-6p.html
This is brillian for a L1 receiver. Then the small print - it can achieve and maintain this level of accuracy in static or slow moving applications. It is priced at around 1.5* the T chip's price and also has the raw data output capability that can be used in achieving a RTK fix (OpenRTK lib. ect).


Looking ahead in the quest for affordable, reliable 5cm accuracy receivers, I can not help but wonder when we will see a L1+L2 receiver from ublox at a price that does not involve selling a kidney. Are these guys going to stay with the mobile device position market? Will this market's demand for more accurate solutions push them to release technology they already posses (I'm speculating, but pretty sure). Can't wait for the NEO-6RTK.

Tags: gps, ublox
E-mail me when people leave their comments –
Follow

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • LanMark February 20, 2013 at 1:13pm

    Does the NEO-6P output NMEA when you use the break out / eval board?

  • heliboy February 20, 2013 at 1:11pm

    Does anybody knwo if APM expects GPS in NMEA format ? What baud ?

    Thanks

  • LanMark February 20, 2013 at 11:44am

    If you get the NEO-6P evaluation board.. what sort of output do you get..  I believe it is raw.. but not sure what that means and how easy it would be to get a percise cordinate..  I think that this chip would allow me to take pretty percise reference points for aerial surveying by using a pole mounted receiver... just not sure what I would do with the output from the NEO-6P evaluation board.

  • Mateusz Zarudzki January 18, 2013 at 12:32pm

    Great Vijay,

    please share Your results with us.

  • heliboy January 18, 2013 at 12:25pm

    Just come up on this discusssion. I just got a NEO-6P evaluation board from an electronics shop in Latvia. So far I have just unpacked it and got it work on ublox test software. I am hoping to get 1 m 2D accuracy on tests tomorrow. My application is ground based, slow moving vehicle so I have hoping this will give me the needed accuracy without the murderous monthly RTK charges.

    - Vijay in the USA

  • Hagen F. Piotraschke January 14, 2013 at 1:23pm

    Hello!

    My name is Hagen, I'm in Germany and I'm working a lot especially with the u-blox NEO-6P.

    Using for my own business (precision farming) as well as offering several devices based on this module, e.g. for RTK but for autonomous mobile applications, too.

    If you're interested, please take a look to www.optimalsystem.de, but let me first tell you this here:

    Yes, the NEO-6P does its job still well in motion. It's not required to have a stationary setup.

    Some customers are already using my carrier boards in UAV applications.

    You cannot compare this module with the LEA-6T (except for RTK by using only the carrier phase raw data output).

    When using the PPP (autonomous resp. SBAS) there's nothing like the NEO-6P (in this price range), neither by u-blox nor by any other competitor.

    Maybe you can compare with one of the Hemisphere OEM boards (SBAS enhanced GPS with carrier smoothing) but these boards are several times more expansive (compared with my small carrier boards with solder pads which I can offer for 100-120 EUR, depending on pieces and accessories).

    There's just a new publication by some Danish scientists (for a robotics conference in Pisa) with some results of the NEO-6P. Not too bad, but I can tell you that it's possible to achieve more precise results (I think they didn't use the best antenna-receiver combination) due to my a bit better experiences ... :-)

    Best regards from Saxony,

    Hagen

     

  • Mateusz Zarudzki December 6, 2012 at 7:15am

    So finally has anyone tried it already? I could not find any material on the internet that says about PPP usage in aviation. Can anyone share his/her experience in using PPP GPS module?

    If not I am going to buy this Ublox NEO-6P module and compare it with LEA-6. 

  • Billy Chan November 17, 2012 at 1:18pm

    I should also mention that the caveat to ambiguity resolution is that:

    • The receiver needs to resolve the ambiguities for each satellite
    • If the receiver looses lock on a satellite, the ambiguity terms for that satelite has to be recomputed/estimated.
    • In an environment where there is very frequent loss of lock, you can't solve for the ambiguities

     

    How can you use PPP and ambiguity resolution in kinematic applications? Easiest way is to turn on the receiver and let it sit as a static receiver for a few minutes so that the receiver can have enough time to resolve its ambiguity terms. Once this is done, you can fly. Unless the receiver looses lock on satellites, you don't need to re-resolve your ambiguity terms - i.e. this works best when you have open-sky and don't fly under trees or buildings.

  • Billy Chan November 17, 2012 at 1:12pm

    Depending on how PPP is implemented, you might be able to get decimeter level accuracy with PPP for kinematic application. It all depends on the secret sauce used for carrier phase ambiguity resolution.

     

    1. For the uninitiated, PPP is a means to get sub-metre level accuracy without needing differential GPS. To get sub-metre level accuracy, you need to do a few things:
      • Get reduce as much GPS errors as possible, including satellite clock biases, satelite orbit errors, tropospheric delay, ionospheric errors. To do this, you can:
      • Get more precide orbit and clock information from external data sources (IGS)
      • Improve tropospheric delay modelling with weather information to better model out the wet component (which makes up 20% of the total delay)
      • Improve ionospheric error modelling using external sources of ionosphere (total electron content) information from such places as NOAA.

        Normally most of the errors listed above are cancelled out using between receiver single differencing techniques (aka. DGPS)
    2. Once you have a more accurate position, you can do other things such as carrier phase smoothing whereby you use the carrier phase observations to smooth out the pseudorange code measurements. By smoothing out the pseudorange measurements, you get more precise positioning solution that allows you to use the pseudorange meausrements, which is usually quite noisy (metre level), and get less noisy (decimetre level) measurements that does not require ambiguity resolution.
    3. Once you have a more accurate position, you can start solving for the carrier phase ambiguities.The carrier phase ambiguity is the number full wave cycles between the satellite and receiver. Since the tracking loop in a GPS receiver can only measure the phase of the incoming signal, it can measure the length (phase) of the last cycle but not the number of full cycles between the receiver and satellite. You have to estimate the number of full cycles using some nifty estimation techniques. To visualize this problem, it helps imagine that you have a very long chain between a satellite and the receiver that allows you to determine the distance between the two. However, you can only measure the fractional length of the last chain on the receiver end but not the number of full links in between. Carrier phase ambiguity resolution tries to estimate the number of full chains. Since by definition, there has to be an integer number of full chains, if you observe for long enough and/or have precise enough estimates of the receiver position, you can solve for the integer ambiguity which provides cm level accuracy. Before this happens, the estimator inside the receiver generally treats the number of full cycles as real numbers (double/float in programming terms). Until you solve for the integer ambiguity (aka. fixed integer ambiguity), you'll have decimetre level (40-70cm) accuracy but not cm level accuracy.

    All in all, PPP mainly performs Step 1 and RTK invovles Step 2-3. You can perform RTK using DGPS but if you only have one receiver, you can't get around to doing RTK without PPP.

  • Andras Schaffer November 17, 2012 at 4:13am
    Ppp has no real use for moving vecicles. it helps to achieve perfect position in stationary applications.
    you can read it in the ublox docs too : positioning improvement can only be expected in an environment with unobstructed sky view during a period on the order of minutes.
This reply was deleted.