GPS Receiver issues Discussion

edited August 2014 in RaceChrono for Android
I am sure many others have experienced issues. I wanted to start a discussion on GPS accuracy/position issues, and maybe everyone can share their problem (and solution!).

I currently have two GPS receivers, an 818XT & an XGPS160.

818XT
My 818XT worked fine for about 6 months, then it started to have issues:
-timing jump issues (where It would show me jump a section of the track and drop as much ~.25-2s off my laptime).
-Timing accuracy issues (No jumping, but erratic/choppy line with about .25-.75s off on my actual laptime).
-Horrible horrible position accuracy, where my position would be offset by some amount of distance to the actual position.

Here are two screen shots of the data as it shows up on RaceChrono:
image
image
image

A few more if you are interested:
http://s277.photobucket.com/user/ymgamal/library/GPS Screenshots


Here is a screenshot of my normal/good 818XT behaviour:
image

After a lot of harassing, QSTARZ support got back to me. They suggested I try removing the battery for about 5 minutes. This SOLVES most of this issue, but does not fix the erratic frequency that is common with 818XT (I have that issue also, but two of my friends have "perfectly" working 818XTs). I basically end up removing the battery after each session, then putting it back in right before we grid. Do note, this will mean its a cold start everytime so startup times can be long. It's not satisfactory, but it makes the 818XT useable now.

XGPS160
Next my XGPS160 has relatively bad positional accuracy, but its spot on for timing. I haven't tested timing as much as the 818XT, but of the 6 or so laps I have confirmed, its always within the margin of error.
My position data for the XGPS160 looks like this:
image

This is my second XGPS160, so I am not sure what to do at this point. I may return the XGPS160 (since its expensive) and stick with the 818XT until something more reliable or consistent comes along. If anyone else has experience with these, help would be greatly appreciated.
«1

Comments

  • The screenshot of the perfect QSTARZ trajectory is probably not realistic expectation. It will not happen like that lap-to-lap. Even on Dual XGPS160 working properly you will always have some off-course trajectories when on relatively fast racing car with fast corner speeds. I'm not talking about 100 meters off, but 10-15 meters on few places should be expected norm for good reception and good data.

    That said the XGPS screenshot attached has very bad looking data (esp. when taken account relatively low speeds between 37-60 mph).

    The bad QSTARZ data looks like how the QSTARZ data usually looks on challenging reception. Challenging reception meaning either the satellites' signal is not getting to the receiver due to a) incorrectly placed receiver b) vehicle is blocking the signal otherwise c) landscape obstructions, mountains, tall houses etc.
  • I agree, I am not necessarily expecting the 818XT data to be like that all the time, but its usually very close to that. A lot of times its got me on asphalt, but sometimes It cuts the corners.

    I am not really sure about the blocked signal thing, Pictures with QSTARZ are the same track, as you can see. Only difference between them is about 2 hours. Matter of fact, this is from the weekend where only took it out of the car at the end of the day.

    Where do you normally place your receiver? I could try moving it to the rear of the car instead? Do you normally attach your GPS to the exterior of the car?
  • edited August 2014
    Here's my normal location. It could be more to the center and more towards the glass, to be optimal. The receiver here is XGPS160. Rear windshield should be tested if front windshield doesn't work. We used to run 818XT outside of the car due to bad trajectories it gave us inside cars with roll cages. Trick should work on the new receivers too, but so far I haven't needed that.

  • aol,

    Do you use the "Allow Mock GPS locations" option in Android with the XGPS160?
  • RaceChrono connects directly to the Bluetooth devices, so the operating system's GPS settings do not have effect.
  • Ran some tests on the GLO & XGPS160. I did this in 2013 Elantra on a small circuit through a near-by neighborhood & highway. Before doing this test, I tried a few locations over the last few weeks to see which one gave me the best reception (left/right/center at the front windshield and center rear windshield). I ended up with center of the front windshield as that is the position the XGPS160 gave the best data on (doubt this mattered much). The GLO was generally position agnostic. I did 2 laps with each Receiver before swaping to the next. It took me about 30 minutes total.

    -The XGPS160 got more satellites (16 vs. 14 average) and generally the data collected by RaceChrono was better (DOP was lower, frequency is 10hz vs. 6.5-8.5 min/max for the GLO).
    -The actual position data however was not. Friday data was pretty bad for the XGPS160. Today's data was much better, but still not as accurate the GLO.
    -In a lot of the main roads, the GLO can place me in the center of the lane or at least in the lane. The XGPS usually has me on the border or outside the lane.

    GLO was much better in the "sweeper" style turns and tended to keep a "straighter" line when straight. Also, I noticed the GLO tended to work well even with low satellites,

    Here is a SS of the GLO with 9 satellites:
    image


    Here a few SS of the GLO (grayed out/transparent) overlayed on the XGPS (solid) data:
    image
    image
    image
    image

    The GLO just tended to be of equal or better accuracy than the XGPS160 in most cases. I would like to also note this was the best data I got with the XGPS160.
  • Good work, and good comparison! Looks like Garmin GLO *is* better in accuracy.
  • Hey! I had similar issues yesterday! Same track, same equipment and same device locations as before, but data quality is incompareble...
    Screensots will follow
  • MAVMAV
    edited September 2014
    So, here are the screenshots. Qstarz 818XT with latest available RC version at the moment on Galaxy 4 Active
    Here is result that I'm used to since using my old iBT 747A+ GPS receiver in 5 hz mode
    photo Screenshot_2014-09-01-17-06-23.png

    There are 10+ laps in this track, so I'm quite happy with quality. Ok, you can't really expect to see lines perfectly in turns on Google maps imagery, but this imagery was not supposed to be used for this. The lines look gteat and I used export to analyse lines in external software like Cirquit Tools.
  • MAVMAV
    edited September 2014
    But yestarday I got some issues. It started with some issues with the lines:
    photo Screenshot_2014-09-01-18-15-06.png
    I've noticed strange things with split times and live timing but lap times was rather realistic, so I saw this mess only after the session..
    I thought that this might be because receiver was laying on the shelf for couple of month and I should had leave it on for some time before the practice session...
    then was first tima attack session:
    photo Screenshot_2014-09-01-17-08-19.png
    After that i had some spare time finishing working with the car so I tried to fix GPS issues... I was confused with only 7-8 satelites fix, and floating update rates. I hever had such problems before so never saw that update rate wasn't consisent in old logs too. I found my old 747 receiver and number of sats was the same, I managed to update A-gps data in 818 and prepeared for second time attack session.
    When it started I couldn't understand why I don't see timer runing. Later in logs I saw why:
    photo Screenshot_2014-09-01-17-10-17.png
    I tried to re-start session couple of times, but it didn't helped. I have several track fragments recorded off the sector traps. Finally i managed to re-start both GPS and RC app on the phone and it started to work fine. Here is one of the laps from the second part of the session. Except of strange "shift" on the straight doing 150 km/h it's fine:
    photo Screenshot_2014-09-01-17-45-58.png

    But's it hard to use this lap (it's fastest lap of the sesion) to analyse lines because of this shift and some other issues that data might have.. It works great for lap timing, but with no consistency it's useless even for sector monitoring and predictive timing (got confused by this couple of times), not talking about analysing exported data in external software...
    So, what happened with my 818? Or (taking into account that my trusted 747 has same issues) what happened with GPS? Bad day and wrong satellites above the location? Or do I need to tweak some settings?
  • One more thing - I woldn't call "good" line example from first message not realistic expectation. I got used to this using 747 receiver and from my initial experience with 818XT. These issues that I had yesterday are rather "unpleasant surprise" for me... I got perfect line earlier with the same receiver location - on the rear view mirror. So, for me it looks like "something happened" recently and unfortunately it seriously affecting RC experience.. :(
  • I have had four Qstarz 818XT receivers, and all of them have started to give crappy data after some use.
    One did work only one month before giving crappy data. The best of those four did work ok about a year.
    I did sent a email about the issue to Qstarz info, but as you can imagine, they did not reply at all.

    One possible fix has been/ might be:
    Replace the battery with a direct power. I.e. use suitable voltage, and solder wires dictly to the "battery connectors".
    Or use the USB-connector to supply 5V-power directly to the reciever. Any USB-cigarette charger, 12V->5V, is just fine.
    I know a couple of fellow track day guys who have used the method above.

    And I also personally know about 30-40 different cases/units (I have spent 30-40 days at track per year/ RaceChrono+818XT is very popular combination), where Qstarz 818XT has started to give crappy data.

    So Qstarz has at least some hardware/battery/connector/power supply issues with 818XT.

    Try the fix, and reply if it worked for you.
  • My 818XT is at Qstarz at the moment. They informed they are replacing the PCB board and sending it back. I am suspicious that they will do anything really. If it works great again, then I guess they must have changed something...

    The guy at Qstars was about as intelligent as a bag of rocks. He had similar excuses to the GLO guys. Unfortunately, we are a pretty niche group, so I can't even get them to budge on possible fixes for anything by mentioning the community as a whole having problems with the GPS receivers.

    I will say keeping the battery fully charged and removing it for 5 minutes before using (ie clear any data kept for hot startup) has fixed most of my accuracy issues. The problem is I occasionally get a jump ~.2-.5 second on a lap. It sucks because you can only really tell in RC if you don't do any scaling. The jump will change your laptimes by that amount and it has never been corroborated by my camera data. I don't know what causes it, but I haven't been able to get rid of that yet.
  • So, I tried my 818XT with and without external power supply. It'sbasically the same. Left device overnight with battery removed. Today I tried it and track looks a bit better. need ot compare it with my old 5hz 747A+
  • edited September 2014
    MAV, you may have better position accuracy if you use the 818XT in 5hz mode with DGPS enabled. Which is basically the same as the 747A+.

    If anything I suspect the 818XT's MKII chipset only supports 5hz update and the 818XT is interpolating the other 5 Hz. dunno if that affects accuracy, but it makes it look smoother. I have noticed my lines with 818XT are always smooth at 10Hz compared to other receivers.
  • MAVMAV
    edited September 2014
    How should I switch it to 5hz? It has only off-1hz-10h...
    Drove around 30 km yesterday evening. Well, I tried it with and without power sullpy and also at some point removed battery for 10 min before taking further tests. From what I can see now in recorded sessions - tracks have visually same quality. Whatever problem it was - now it's gone. I would even say that it is "perfect" now...
    Lines are nice and smooth, I can clearly see which lane I'm driving in most of the cases.
    This is great, but I'm a bit disappointed since I don't know what happened.
    Update rates are 7,5-10Hz, DOP around 1.1-1.3(sometimes)
  • edited September 2014
    aol maybe able to chime in. I think RC used to reconfigure the 818XT to 5hz if you turn it to 1hz mode. Otherwise you can hook it up your computer, download the QSTARZ software and reconfigure it through there. should be permenant after that. Trackmaster has a How-To on their forum since they suggest you only use 5Hz BT transmitters.

    As a quick update. I have returned my Garmin GLO and should be receiving the 818XT in the mail anytime now. I will update you guys on how it performs (on the street) sometime in the next week. I really hope QStarz fixed it :/
  • Yep, I would also recommend running it 5 Hz mode due to DGPS but the difference is not that big though. RaceChrono used to configure 1 Hz mode to 5 Hz automatically, but I had to remove this feature due to Android 4.3-4.4 Bluetooth flow control issue that made the connection unstable if I sent anything to the device...

    PS. I've just made this discussion sticky. Great contributions!
  • edited September 2014
    I have tested both qstarz 818xt and dual xgps 160. A couple of quick observations:

    818xt -- much better accuracy at 5hz due to DGPS. However, once the battery was running low (or at random times?) the gps signal started looking like garbage. After usage, the battery was not holding that much charge and it became even more erratic. I replaced it with a nokia battery -- apparently they are the same as some of their old phone batteries. Then it worked better, but that is still until the battery was low on juice and then it went crazy. Very reliable device when it works, but I would lose one out of 10 sessions of track data on average due to garbage. The solution was to recharge in between sessions when I saw the erratic data.

    xgps 160 -- More accurate than 818xt and 10hz refresh rate. However, I had to update the firmware first. Ver 3 was better than 2. I have not tried 3.5 yet. I can tell exactly what line I was using on the track with the xgps 160 and it correlates with my video output. For some reason the g-force computation seems to be erratic and have steps from 1 to 1.1 to 1.2 instead of being smooth and having more precise fractions. 818xt seemed to give a smoother g-force line. Not sure why this is happening. Overall, very happy with the dual xgps 160. I like it a lot with the predictive lap timing feature and the 10hz rate.
  • spaparizos, you should not replace the 818XT battery with Nokia one. It works, but it is little bit loose, so it disconnects with vibration.
  • how did you do the firmware update? Wonder if I should have updated those new XGPS160s. I couldn't find anything on the Dual site...
  • had track day with 818 and got disappointed again.. Overal lap timing looked pretty accurate, but sector times were confusing and when I checked logged trajectory it appeared to be unusable for any further analysis...
    Kind of pissed because it was in fact training session so I was expecting to get some info. Will search how to lock it in 5 hz mode instead of 10 hz...
  • edited March 2015
    Just wanted to update about the 818XT. I tried out the replaced/fixed unit for Qstarz and have to say I am satisfied. they fixed the erratic frequency and my position data is back to what it used to be. I still occasionally use the battery trick, but I have rarely gotten awful data from it. I have taken it out to a few track events now on top of some street testing and its been consistently good. This is about 6 months after I received it, so I am half expecting it to start failing again. Will keep posting about it. I was considering picking up a 160 again and trying the new firmware, but honestly I don't feel a need at this time.
  • edited May 2015
    I am using a Nexus 5 Android 5.1. Had an issue the last two track events where the 818XT would intermittently (approx. every 13 seconds) drop to 0Hz then come back after a second or two. Turns out this is an android issue and not necessarily related to the 818XT. I tested this with a friend's Nexus 5:

    -Connected the 818XT to my phone
    -Saw the intermittent drop
    -turned my phone's BT off.
    -Connected the 818XT to the other phone (without powering down any device)
    -Did not see the issue
    -Reconnected to my phone
    -Experienced the issue again.


    Ultimately, I fixed it by restarting my phone. Its happened twice now (on a vanilla android 5.1), so keep a look out if you are using a similar phone/version. Obviously the culprit here is the android BT stack (possibly RC, but that is doubtful)...which has had issues before.
  • Maybe its an issue related to 5.1? Kept my Nexus 5 on 4.4.4 and not had an issue using a Garmin Glo across multiple days now.
  • edited May 2015
    I did some testing on Android 5.0 (Samsung S4) and Android 5.1 (Google Nexus 5):

    1) Dual XGPS 160:
    - Rock steady 10 Hz for both devices. Still the best device in my opinion.

    2) Garmin GLO:
    - Update rate is about 7 Hz on both phones. It's little bit worse than on Android 4, I remember getting around 8.5 Hz, so this is little bit worse. Also saw one 2 second break in the data on my 5 minute test session. Garmin GLO has some flow control problem with Android that causes this. The data is still pretty good as the precision of this device is great. Second best GPS after Dual XGPS in my opinion. Would be equal if this flow control issue was fixed.

    3) Qstarz 818XT
    - Receiving steady 10 Hz right now on both devices, BUT it's quite common to have some flow control issues with this device. And the kind where you get the 1-2s pause every 5-10 seconds is probably worst one you can have with it. I've seen this on all Android version from 4.1 to 4.4. I don't know what is causing this, other than some incompatibility in flow control. This device lacks GLONASS support so would not recommend buying it anymore, as the lock to satellites is not nearly as good compared to two devices above.
  • edited May 2015
    yea, I definitely don't know where the issue is exactly. I definitely point my finger at the Nexus 5 though since restarting it seems to fix the issue (although its possible there is an issue with the 818XT and how it communicates with BT on android...but then why did it work on friend's phone fine without a restart or anything?).

    Planning on retesting the XGPS160 in July, so I will post my experience then.
  • Hi I don't know what gps receiver i have in my phone but it is an LG G2 and the thing just doesn't want to work the way it should. If there is anybody with some time to spare please do it , yesterday I was about to take the phone and through it on the ground after my 1st session in the track I don't know why it is not working.

    Before the start of the 1st session I told my friend I hope it works today we have clear sky with nothing but blue sky over us.......it was a balls up again.

    Please help me with this , I thank you!!!
  • I bought the XGPS 160 after years of using the 818XT.
    Having the XGPS160 mounted in the front windscreen gave me very very bad position accuracy.
    I had to mount it outside the car to get better position accuracy.

    So similar issues with the 818XT.
    I am a bit disappointed.
  • Helo, my father have this gps activepilot. Is it ok for the first time to use?
    Sry link is in german. I hope you can help me :)

    http://www.handytechnik.de/assets/files/bedienungsanleitungen/Jentro activepilot GPS Maus/89150_Jentro-bt-gps-8_UM.pdf
Sign In or Register to comment.