Database Users

Weird GPS problem. Moving even while my car is motionless.

Hi i have weird problem with my GPS.
While I driving (car is moving) everything is beauty and sweet but once I stop my car my GPS is still moving for about 10 sometimes 20 sec showing 2km/h then 0km/h then 2km/h again. Sometimes when I stop car on junction my Gps will jump on the perpendicular road and from time to time my arrow on the map just turn around to opposite direction and the maps sets a new route! Once i move my car everything going back to normal.
I was using first my TomTom and i just thought it`s a map problem although i haven`t problem with that on my Rhodium either X1. So I decided to try Igo8 and it`s exactly this same. Even on the roundabout my gps can jump somewhere else on different road.
I tried all Radios available for HD2, many Roms and just have no idea what more I can do. Many people on HD2 forum in Poland have this same problem.
I also tried GPS fix cab downloaded here from xda but no luck its ever worst.
I can add that i'm getting GPS fix in few second so its good the problem is just my gps is moving around.
PS. Very, very rarely have no problem with this jumpin gps around. But is like 1/30.

Do you have access to an external Bluetooth GPS receiver you can test out? That could eliminate or isolate some possibilities.

balane said:
Do you have access to an external Bluetooth GPS receiver you can test out? That could eliminate or isolate some possibilities.
Click to expand...
Click to collapse
unfortunately i don`t. but on my Touch Pro 2 and Xperia never had this problem.

hi,
are you using the 'gpsmoddriver' software, that is floating around here in the forum?. its made for using the compass data when no gps fix is available. it causes such failures in igo and copilot on my hd2. after uninstalling it gps works much more percise to me, when im not moving.
regards mad

Happens with mine as well. I have replaced my HD2 and it seems that the 2nd unit acts the same. tried various ROMs/RADIOs (official/cooked) - nothing seemed to fix this.
It's quite funny though... most of the people rather paying attention to the fix time instead of the accuracy (which is quite bad imo) in addition to this weird phenomenon. I doubt that the AGPS is somehow related, but it will be nice to know how to disable this feature, and try to acquire the signal without it. maybe this is the cause.
Btw, check this out - http://forum.xda-developers.com/showthread.php?t=600232

Fatherboard said:
Happens with mine as well. I have replaced my HD2 and it seems that the 2nd unit acts the same. tried various ROMs/RADIOs (official/cooked) - nothing seemed to fix this.
It's quite funny though... most of the people rather paying attention to the fix time instead of the accuracy (which is quite bad imo) in addition to this weird phenomenon. I doubt that the AGPS is somehow related, but it will be nice to know how to disable this feature, and try to acquire the signal without it. maybe this is the cause.
Btw, check this out - http://forum.xda-developers.com/showthread.php?t=600232
Click to expand...
Click to collapse
Agps - this feature is available in NRGZ roms. I tried to turn on and off and it was no different.
PS. thanks for answering.
madbird said:
hi,
are you using the 'gpsmoddriver' software, that is floating around here in the forum?. its made for using the compass data when no gps fix is available. it causes such failures in igo and copilot on my hd2. after uninstalling it gps works much more percise to me, when im not moving.
regards mad
Click to expand...
Click to collapse
As i said before : i was using it. I had feelings its jumping around the places more than without it so I uninstalled it.

Tried to email HTC regarding this. they claim that their HD2's don't exhibit this behavior, even though I've tried two different units, from different batches.
I don't know what to do... this drives IGO nuts - it causes igo to switch lanes, directions and having the route recalculated without any good reason.

Fatherboard said:
Tried to email HTC regarding this. they claim that their HD2's don't exhibit this behavior, even though I've tried two different units, from different batches.
I don't know what to do... this drives IGO nuts - it causes igo to switch lanes, directions and having the route recalculated without any good reason.
Click to expand...
Click to collapse
Sometimes its just pointless to talk with those dickheads.
They ll never say true.
With pink camera was this same. They were saying at the beginning that they have no problem with pink spot at all. And what was next? We need just more people to email or ring them.
Most of them even not using HD2. They've seen for few minutes and think they'll know everything about it.
It was so many bugs with this device they just dont want to hear about another one!

In addition to that, both of my units tended to show different position and altitude readings each time i powered up igo or google maps. while exhibiting the differences I obviously didn't from my current position, not even an inch.
to try to make things clearer. that's what I did :
loading igo -> acquiring signal within seconds -> getting a certain position reading with altitude of 11 meters (for example) -> quit -> reloading igo -> acquiring signal within seconds -> slightly different position, altitude of 75 meters or a negative value (for example).
even the small blinking dot in igo (which determines the actual gps mark) doesn't fit to my physical location on the road which I'm currently driving on. it sometimes blinks few meters to the right/left, and sometimes it tracks behind the guidance arrow (as it should be).

Could be
Simply a limitation of GPS. I have seen it on other GPS's. Garmin etrex. Try walking with one and even thou you are walking in a straight line it plots a zigzag path. More satelites (10+) should reduce this issue.
Basically the GPS is only accurate to a certain range from +-50m to +- 2 meters depending on the no of sats. The location is detemined by timing from when the signal was sent to when it was received. Calcs how far away the sat is (Speed of light??). It is possible you care getting a fluctuation in your current location. Eg
Sample 1 says you are point a.
Sample 2 says you are 2 meters to the right.
Sample 3 says you are 2 meters to the left.
Still in the +- 5 meters but the gps is plotting you moving backwards and forwards.

Guys, it is a problem everyone has, gps position is not as accurate or stable on the hd2 as on most other htc devices. And with an erratic gps position automatically comes an erratic speed. Don't know yet if it is because of a different gps module that's inside the hd2 or because of the inclusion of crappy gps drivers. Will try to find out soon....
Anyway, gpsmoddriver is not the cause of this problem since the problem is also on an hd2 without having gpsmoddriver installed. Because of the functionality of gpsmoddriver to improve the gps readings, in some cases it can amplify your erratic gps position, and that is also what some of you have been experiencing. Remember you can always use gpsmoddriver to add the hardware compass functionality to your navigation software, while additionally configuring gpsmoddriver to leave gps data untouched so it won't make the erratic position and speed worse. For more info or help, please visit the gpsmoddriver thread: http://forum.xda-developers.com/showthread.php?t=571266
EDIT: Of course the HD2 has a new snapdragon chip which has built-in GPS just like many previous generations of phones. Most likely qualcomm changed the way gps works in their new line of Snapdragon processors...

The movement of vehicles around you or even the movement of the sun's rays on buildings nearby can distort GPS signals. The GPS navigator software tries to rationalize this distortion and that usually appears as a movement when you are still. On the other hand, if you drive into a tunnel using TomTom and some other software, you will often see yourself driving straight on in the tunnel well after the GPS is receiving no more satellite signals.
In top-of-the-line built-in navigation systems, the GPS data is supplemented by inertial guidance. (Someone could create a similar supplement for the HD2 The supplementary information allows the system to reconcile the GPS information with information on your car's actual movement.
But for us mere mortals, we just have to live with it.

Well... this particular gps chip is quite useless, then. if my navigation program has to recalculate the route each time I stop at traffic light, then it seems that I either have to purchase a usb-based gps to pair with the HD2, or replace this device. I counted on it to perform well.

Fatherboard said:
Well... this particular gps chip is quite useless, then. if my navigation program has to recalculate the route each time I stop at traffic light, then it seems that I either have to purchase a usb-based gps to pair with the HD2, or replace this device. I counted on it to perform well.
Click to expand...
Click to collapse
Complain to HTC, perhaps if they get enough complains they will come out with a hotfix or a rom upgrade that fixes it. Nobody said that the hardware is not that good (though it's possible), but it could also very well be a bad implementation by HTC!!

Already did.
They have obviously denied my complaints.
thanks again, anyway.

I noticed this behaviour yesterday night, at about 10pm. I was playing with NoniGPSplot, and while I was completely still (and freezing me arse outside believe me), it was tracking a movement back and fro and all around in small step, keeping track of me traveling some meters still while I was standing in the same point.
It was night and I was in my big yard, so no sunrays nor moving vehicles around nor buildings over my head.
I think this behaviour is very well hidden in tomtom since it automatically puts you on a road, and "smoothens" the gps data by making you follow that road no matter what, but a "realtime" software like nonigps tracks that all so well. Gonna notify htc about it now.

Guys, let's not go overboard on this. Non-military GPS is only claimed accurate to around 20 metres, regardless of the quality of the receiver. It's the US Defense Department messing with your HD2, not HTC.

Stephen Selby said:
Guys, let's not go overboard on this. Non-military GPS is only claimed accurate to around 20 metres, regardless of the quality of the receiver. It's the US Defense Department messing with your HD2, not HTC.
Click to expand...
Click to collapse
Yes, absolutely, I've noticed the same behaviour with all three of the BlueTooth GPS I used in the past. So, I would not be too quick to blame HTC for this.

my touch cruise's gps receiver worked with static navigation enabled to avoid this false movement. maybe this is just a usual behaviour?

Agree to that.
But regardless of the fact that the thing is not a military GPS system, It should do it's work correctly as for me the GPS is almost useless when driving slow or within a city. My 2nd phone (HTC DIAMOND) and 3 personal navigation devices are mounted in my car to compare and all work fine when the HD2 isn’t .
The GPS looses all the time the fix to the satellites and then quickly gets the fix back. when using a GPS monitor you can see that the satellites constantly moving really rapidly. (moving --> getting fix and loosing it again)
This is for 4 other reference devices which are running simultaneously not the case so it seems really an issue with the GPS device.
I tested this with different ROMs (1.44 Vodafone, 0.6 dutty WM 6.5.5, 1.66 stock and another one I cannot recall the producer anymore.)

GPS going crazy (proven with plot screenshots)

I used NoniGPSPlot to trace my movements as the GPS reported them, while I ran back and forth in my yard this evening; I took screen to show you how crazy can the GPS be in this device.
Facts:
1) these screens have been taken with the shake-to-screenshot function of BsB tweak
2) the results you see here are constant, I just decided today to take screens of what regularly happens whenever I want to test the precision of GPS
3) my movement has been a plain run back and forth on a linear path, probably 50m or so, no pauses but the ones needed to reverse the course when I reached the end of the yard
4) it is 8pm here atm, cold outside, clear sky, no buildings nearby except my own, which is still 30m away from the running path... and I do not live in a skyscraper anyway
5) I was holding the HD2 in my hand all the time while running
6) I had given enough time to the gps position to fix, I have deleted the erratic trace the gps gives as soon as it starts acquiring position
7) I can proficiently use TomTom to follow the road with my car, no erratic behaviours there, except the usual places where there's a slight turn and tomtom thinks you're still on the main course, just to realize a few seconds later that you actually are on the almost parallel street
Here are the shots:
(first one taken with GPSmodDriver disabled, nonigpsplot set to use the COM4 internal GPS)
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
(second screen, taken with GPSmodDriver enabled, profile set to walking, all default prefs, nothing changed but selecting the walking profile)
Also please notice the distance reported by nonigpsplot to have been ran, in the lower right corner. There's NOWAY I could've ran that much meters.
I would very much appreciate your thoughts, suggestions, whatever, except:
1) you should see a physiotherapist because you have a strange conception of "straight line"
2) you should check your yard for relativistic holes
3) your GPS chip is faulty, return the unit (lame excuse, were the GPS system faulty I would've gotten completely erratic readings -telling me I'm in asia for example instead of nearby rome- or no readings at all)
4) nonigpsplot is a bad software (that's, afaik, a very widely used app, can't believe it behaves erratically just for me; also, I used googlemaps connected though my wifi, and I can tell you that, even if gmaps doesn't do plots, the direction arrow representing me was pointing in the most variables place while I moved; anyway, if you have in mind other fre gps plotting softwares, I'll be happy to test them in my yard, or you can do the same at your place and report here

Well, GPS tends to be accurate if you're standing still, or if you're moving at a decent speed.
It is less accurate at slow speeds.
The accuracy per "tick" can be +/- 20m or more, but you'll usually find it's about 5m, so basically, every time it calculates the position it'll be about 5m off in any direction.
Despite that, your accuracy does seem to be all over the place, if that scale on the program is correct. How many satellites did you have a lock on? The meter on the bottom right of the screen seems to indicate very poor reception, this would also result in larger jumps in position, as would tall buildings/trees/anything around you.

I had 6 "working" satellites above me, the mean of the reception bars length was a little above 50%. My yard does have tall trees around it, not that many tho. Let's say that this Leo is not good for hicking, as for example, if I'm going to use it to find my way out of a forest, I can kiss goodbye to home The fix is still lightning fast if compared to my previous BT antenna.

theemed said:
I had 6 "working" satellites above me, the mean of the reception bars length was a little above 50%. My yard does have tall trees around it, not that many tho. Let's say that this Leo is not good for hicking, as for example, if I'm going to use it to find my way out of a forest, I can kiss goodbye to home The fix is still lightning fast if compared to my previous BT antenna.
Click to expand...
Click to collapse
You should have a lock on at least 7 satellites to get a decent reading.
Most GPSes are no good for walking, I reckon the ones in phones are good for knowing where you are on foot, and just keeping you going in generally the right direction, and they concentrate more on usability for driving.

I think I should get more info on that, had no idea 7 satellites were necessary. Following my high-school studies, afaik 3 measures are sufficient to get an evaluation of a 3d position, so I supposed 6 satellites were far more than enough
In which case, my yard is not a good point enough...

Im using my HTC HD 2 as a data collector plotting up to 80 - 200 waypoints a day with 15 - 35 metre accuracy which is quite acceptable for a recreational type GPS. Equally as good if not better then our Bluetooth Globalsat BT338 units.
We get more satellites and a lower PDOPs then the Globalsats which have the sirf 3 chip.
Ihave been really impressed with the HDT HD 2 GPS
5 or 6 satellites are good enough for an accurate position
Try marking a waypoint several times and testing the deviation in position

Well, I do not have a real "waypoint" whenever I stand still, I have a moving position, or at least that's what nonigpsplot gives me.
Out of curiosity, what's the tool you use to plot?

Were using OziExplorer for our mapping, finding our way to the waypoints etc
and a specially developed program for the data collection and plotting.
I have not used nonigpsplot if it has a trial download I will see how it goes on my HD 2

theemed said:
afaik 3 measures are sufficient to get an evaluation of a 3d position.
Click to expand...
Click to collapse
It is, but it will have awful precision due to the timing tolerances of the GPS system. The more sats you have the more precise you get, with about 10 needed to get a 5m precision. Your software does tell you the reception is low, and this normally takes more into account than just the number of sats.
One thing that plays a great role is how separated the sats you're receiving are. If you use a GPS program that shows the sat positions, you can have a look at where they are. If they're all "packed" in the same area of the sky, your precision will be low. The more separated they are the better it gets (extend on the thought that if 2 sats are at the same place, they're no better than one, try to triangulate if 2 of the 3 points are at the same place or very close).
What this can show is that the HD2's GPS could have a bit less filtering than usual, but not much more.

kilrah said:
One thing that plays a great role is how separated the sats you're receiving are. If you use a GPS program that shows the sat positions, you can have a look at where they are. If they're all "packed" in the same area of the sky, your precision will be low. The more separated they are the better it gets (extend on the thought that if 2 sats are at the same place, they're no better than one, try to triangulate if 2 of the 3 points are at the same place or very close).
Click to expand...
Click to collapse
Nonigpsplot has a nice 3d view of the sats position in the sky, and well they actually are in the medium circle between azimuth and the horizon. At least now I know how to make more out of that kind of info, I never thought about sat position to be important in that matter, even if after all it's pretty obvious
forest.ranger said:
I have not used nonigpsplot if it has a trial download I will see how it goes on my HD 2
Click to expand...
Click to collapse
If not for anything else, at least nonigpsplot has a very big advantage when compared to oziexplorer... it's free And does work with gmaps and openstreetmap

I will take a sample of mine.
However with your low number of sats I would suspect thats your problem. Poor accuracy is what I would expect from only a few satelites.

theemed said:
and well they actually are in the medium circle between azimuth and the horizon.
Click to expand...
Click to collapse
Don't forget the number of visible sats and their positions change constantly

I had a play with NoniGPSPlot today while at work. While I was standing still I was getting the same erratic jumps as in the screenshots but when moving at a walk it would show a track line.
I was getting the same weak signal indication but the satellite page of NoniGPSPlot was showing 6 satellites and OxiExplorer was showing 8.
The erratic jumps in the screenshots could be just the normal variation of each gps plot exaggerated by the program.

looks like a deflection issue to me

I took nonigpsplot with me today on a roughly 34km total long trip, back and forth.
No big problems while moving, but as soon as I stopped the car at the (frequent here) traffic jams, the position kept wandering off in various directions with some looping paths and came slowly back near the real position of the car to almost stop there. It idled there even after I moved the car, to follow me with a big jump a few second after I started moving.
It's like there's some sort of built-in "inertia" in the GPS chip....

I've been messaging with the HTC support, and they basically dany that the HD2 gps chip could ever behave like that, and that it's sure the fault of the nonigpsplot application, which may interfere with the GPS reception EVEN WHEN IT'S NOT RUNNING )
I asked them to recommend me a gps plotting utility which is compatible with the HD2 GPS chip, let's see what they're going to suggest

Have you tried Reperion?
I use it with my HD2, its not beautiful, but it works well for me, using it mainly to send tracks live to Google Earth.
http://live.reperion.com/media/

theemed said:
I've been messaging with the HTC support, and they basically dany that the HD2 gps chip could ever behave like that, and that it's sure the fault of the nonigpsplot application, which may interfere with the GPS reception EVEN WHEN IT'S NOT RUNNING )
I asked them to recommend me a gps plotting utility which is compatible with the HD2 GPS chip, let's see what they're going to suggest
Click to expand...
Click to collapse
Chances are HTC support are only trained on basic phone functions. I would not rely on them for help on anything complicated. This is what XDA is here for!

bslask said:
Have you tried Reperion?
I use it with my HD2, its not beautiful, but it works well for me, using it mainly to send tracks live to Google Earth.
http://live.reperion.com/media/
Click to expand...
Click to collapse
Downloaded and installed... alas I'm inside now so i can't really use it
Will report back... but still, my request to HTC support about an alternative plotting solution was to tease them out of the "not our fault, it's the app, uninstall it!" defense strategy...
elyl said:
Chances are HTC support are only trained on basic phone functions. I would not rely on them for help on anything complicated. This is what XDA is here for!
Click to expand...
Click to collapse
well okay, that's correct still, I can't complain with the guru's at xda for something which is probably due to HTC's shortsightedness in choosing/implementing the gps subsystem of their device. I know I should just pair this hd2 with my old BT antenna to see if I get the same weird position jump while being still that I get with the internal gps...

theemed said:
I used NoniGPSPlot to trace my movements as the GPS reported them, while I ran back and forth in my yard this evening; I took screen to show you how crazy can the GPS be in this device.
Facts:
1) these screens have been taken with the shake-to-screenshot function of BsB tweak
2) the results you see here are constant, I just decided today to take screens of what regularly happens whenever I want to test the precision of GPS
3) my movement has been a plain run back and forth on a linear path, probably 50m or so, no pauses but the ones needed to reverse the course when I reached the end of the yard
4) it is 8pm here atm, cold outside, clear sky, no buildings nearby except my own, which is still 30m away from the running path... and I do not live in a skyscraper anyway
5) I was holding the HD2 in my hand all the time while running
6) I had given enough time to the gps position to fix, I have deleted the erratic trace the gps gives as soon as it starts acquiring position
7) I can proficiently use TomTom to follow the road with my car, no erratic behaviours there, except the usual places where there's a slight turn and tomtom thinks you're still on the main course, just to realize a few seconds later that you actually are on the almost parallel street
Here are the shots:
(first one taken with GPSmodDriver disabled, nonigpsplot set to use the COM4 internal GPS)
(second screen, taken with GPSmodDriver enabled, profile set to walking, all default prefs, nothing changed but selecting the walking profile)
Also please notice the distance reported by nonigpsplot to have been ran, in the lower right corner. There's NOWAY I could've ran that much meters.
I would very much appreciate your thoughts, suggestions, whatever, except:
1) you should see a physiotherapist because you have a strange conception of "straight line"
2) you should check your yard for relativistic holes
3) your GPS chip is faulty, return the unit (lame excuse, were the GPS system faulty I would've gotten completely erratic readings -telling me I'm in asia for example instead of nearby rome- or no readings at all)
4) nonigpsplot is a bad software (that's, afaik, a very widely used app, can't believe it behaves erratically just for me; also, I used googlemaps connected though my wifi, and I can tell you that, even if gmaps doesn't do plots, the direction arrow representing me was pointing in the most variables place while I moved; anyway, if you have in mind other fre gps plotting softwares, I'll be happy to test them in my yard, or you can do the same at your place and report here
Click to expand...
Click to collapse
There's no doubt - the gps in the HD2 does not perform as expected (or rather, does not perfrom well at all - very bad accuracy, completely useless in everyday navigation - it fits only for general orientation). I can verify the same behavior. is it software or program related issue? The guys in HTC don't really (want to) help us...

RFF: Why GPS information is poor...

The bizarre 'wandering' behaviour leads me to suspect the navigation engine. When using a standalone GPS test app the Satellite lock I see is fine and the SNR (assuming it is being reported honestly) is OK.
In addition, when I'm using an app like Open GPSTracker I get convincing and accurate updates at around 1Hz (which is what you would expect) with a good solid position fix to within an accuracy I am happy with (usually 5-10m).
There are two issues I see:
First is that the speed reported by GPS apps is Bunk, or laggy at best. I haven't looked at the NMEA logs from the built-in GPS - do they contain accurate speed information?
Second is the 'wandering GPS' issue, which results in the classic continuing-straight-on-when-you-just-turned issue most of us are seeing.
This second issue, combined with the way you can observe the GPS reported position in something like Google Navigation slowly converging with the location you know you are are leads me to strongly suspect that Samsung are using some sort of IN/GPS mixing and doing it very badly.
GPS simply gives a position in its pure form. There's no interpolation or guessing at 1Hz (if you have GPS that advertises itself as a higher refresh rate then it probably is. but I assume the embedded GPS chip isn't too fancy and updates at a standard 1Hz). Usually the receiver then calculates the speed using a simple time/distance calculation between two points. Heading is simple geometry. Based in this fact I can think of no other explanation for the drifting behaviour.
What I suspect:
For some reason Samsung is mixing magnetometer and/or accelerometer data with the GPS positions (regardless of whether you select the 'use sensor data' option in the location menu), probably to avoid having to do the speed/direction computations from the raw GPS data alone or because their driver isn't providing it in the NMEA line (I assume that at some point the GPS chip provides a NMEA formatted output to the OS, this is that way every other device works!). In theory this should work really well and cut processing overhead - if you've got a good accelerometer giving acceleration and a good magnetometer giving heading why bother to calculate it from GPS position fixes, you can simply bias towards sensors and use the GPS for occasional updates.
However, while the MEM devices in the SGS while good enough for apps like Layar and Google Sky simply aren't up to the job of providing navigational quality mixing. I believe Samsung are being too clever and should re-write their positional engine using GPS position data alone and calculating speed and heading from that. It works for almost every other GPS device on the market.
At some point Samsung have tried to compensate for poor GPS performance by feeding sensors into the mix at the system level without user control. (I suppose this MIGHT be Googles fault??) This is giving a short-term bias towards interpolated positions fed from heading and accelerometer date which is provided by inaccurate sensors corrupting the perfectly good GPS information.
As I say, I don't know how good my guesses above are, it's simply an educated guess based on observation and some knowledge of IN/GPS mixing filters.
What do you guys think?

that's pretty much what i think but i dont have anyway to prove it

Well, standing still with walking directions with google Navigation I can see the zooming in and out

Pretty good assumptions there, dangrayorg. You might be spot on.

Sounds plausible. Of course the fundamental problem is probably poor GPS receiver and now Samsung is trying to work around that unfortunately with not to good results.

[GUIDE][FIX] GPS finding satellites, but not connecting

tl;dr version:
1) Switch data on. (not wifi, data. Has to be data.)
2) ? ? ?
3) PROFIT.
looong version:
I kept wondering why GPS would find many satellites, but would often not lock onto them, or would take a very very long time to lock (around 10 mins or so). Then I noticed that it works fine when data is switched on, and locks within a few seconds.
You should get a lock as soon as you're "fully connected to Google", and "go green" for people who have that enabled.
This is the a part in agps at work Assisted GPS. Click to check the wikipedia article on it.
Accuracy is still a problem, and testing with a friend's DHD, I was getting around 10-6m accuracy, and he generally got 2-4m.
And Samsung's GPS software sucks too. glgps daemon's internal smoothing algorithms are still there, and it's annoying.
~Rawat

Rawat said:
tl;dr version:
1) Switch data on. (not wifi, data. Has to be data.)
2) ? ? ?
3) PROFIT.
looong version:
I kept wondering why GPS would find many satellites, but would often not lock onto them, or would take a very very long time to lock (around 10 mins or so). Then I noticed that it works fine when data is switched on, and locks within a few seconds.
You should get a lock as soon as you're "fully connected to Google", and "go green" for people who have that enabled.
This is the a part in agps at work Assisted GPS. Click to check the wikipedia article on it.
Accuracy is still a problem, and testing with a friend's DHD, I was getting around 10-6m accuracy, and he generally got 2-4m.
And Samsung's GPS software sucks too. glgps daemon's internal smoothing algorithms are still there, and it's annoying.
~Rawat
Click to expand...
Click to collapse
If you go to the Location Settings, and Untick "Use wireless networks" and Untick "Use sensor aiding"
You will get a GPS lock even without DATA. BUT, it will take longer.

Not on my Sgs... with those unchecked and cold start I still get a lock within 10 secs. So something funky in Sgs ii his stuff.
Sent from my GT-I9000 using Tapatalk

So does the SGS2 have similar GPS issues to the original? I'm currently on the Samsung Captivate and have a preorder in for the SGS2. One of the big reasons I'm looking to spend 800 bucks to upgrade is to be done with GPS issues on my phone. If this won't be the case maybe I need to give up on Samsung and wait for HTC to release a dual core beast on AT&T bands...

TheSopranos16 said:
So does the SGS2 have similar GPS issues to the original?
Click to expand...
Click to collapse
I have no experience with the earlier Galaxy's but I can compare it to a G2 and G2X. There are three GPS settings: Use wireless networks (cell and Wi-Fi), use GPS satellites, and use sensor aiding. With "use wireless networks" checked and Wi-Fi off it just locked and loaded for me in Google maps in less than a minute from a poor location. With only "GPS satellites" checked it took closer to 2 minutes. It was accurate down to my street address both times. In order of speed from "activation" to "usable" I'd rank the phones in this order: G2, G2X, SGS2. While a bit slower to lock than the others it's fine (accuracy, holding lock) for me. Someone from the EU uses Tracks to measure his walks and was complaining about the street-level accuracy. You might want to check out that thread.

I personally never ever would use the WiFi- and Sensor functions as GPS replacement.
Why?
1) It is responsible for the collection of the location data and sending it to Google.
2) It costs battery power.
3) Why using when the GPS fixes fast enough? What's the problem with waiting for 10 seconds for a fix?
The fix can be made quicker by DL'ing the "GPS Status & Toolbox" app and using it for DL'ing the actual valid GPS sat vectors. That helps the GPS to fix quicker.
Cheers
Zap

Mine is very slow to lock, despite having plenty of sats visible with good strength.
This is compared to 4 different ZTE Blades that showed me how quick locking can be, even indoors.
There is a tool on the market called GPS Aids which was developed for the Galaxy and I'm keen to discover whether this can improve the SGS2.
However, it requires rooting and I'm reluctant to do so as I may decide to replace my phone if it can't be sorted.
Is anyone who is rooted willing to carry out some tests using that or other tools to see whether performance can be improved? I'd really appreciate it.

TheSopranos16 said:
So does the SGS2 have similar GPS issues to the original? I'm currently on the Samsung Captivate and have a preorder in for the SGS2. One of the big reasons I'm looking to spend 800 bucks to upgrade is to be done with GPS issues on my phone. If this won't be the case maybe I need to give up on Samsung and wait for HTC to release a dual core beast on AT&T bands...
Click to expand...
Click to collapse
I've used GPS a few times, and it works fine. Accuracy is a bit weak sometimes (compared to other phones) but it's fine for navigation, and even using mytracks or similar. (although when it goes off on mytracks it'll take a while to get back due to Samsung's inane smoothing)
prusling said:
Mine is very slow to lock, despite having plenty of sats visible with good strength.
This is compared to 4 different ZTE Blades that showed me how quick locking can be, even indoors.
Click to expand...
Click to collapse
Not sure why it's like this, but you have to have data enabled for it to get a quick lock on SGS II, otherwise it'll take a few mins.
I'll give GPS AIDS a test later on. Maybe.

My gps has been spot on from day one-locks in seconds and google maps always puts me in right room of house or parking space etc. The blue accuacy circle is bigger on sgs2 then sgs1 but the reall accuacy is tons better.

--deleted--

Hi everybody.. i have downloaded an app called GPS aids v 2.0 and this has helped me in getting a lock in under 15 secs after which i fire up Sygic and it hardly takes 10 secs to lock my position.
Do put in a thanks if it helps somebody

Should I give up on i9000? GPS will NEVER work! Can't we sue them or get a recall?

I actually need gps quite a bit in my line of work. I bought this phone because of it. But, it just does not work. Some claim it works fine, but at the same time they are saying they get 5-10m lock! (thats off by 30ft!) With a borrowed motorola atrix i was getting 4.9 feet accuracy (thats like accurate by 2 steps!). I think I have come to the conslusion that it will never work, and samsuck doesn't even care! Why else would they keep releasing newer phones with the exact same problems? I'm gonna jump ship...everything else about this phone is awesome. Any ideas if HTC has good gps/reception? I am thinking about the incredible S. (and i'm not a troll, I have lived with this phone's gps issues for almost 1 year now).

Get a real GPS receiver, and don't expect it will be cheap if you want a real 4.9ft accuracy.
and for i9000 is about 15ft~30ft accuracy, quite good when compare to my garmin unit.

Samsung should get sued for that piece of hardware ****... but the i9000 is still the best phone
You should try Darkys 10.1 with Darky Core and JVO Modem. GPS works okay but no perfectly !

i do like this:
android market : gps toolbox , in settings has 2 options - download agps satelites and reset gps
1st i download the satelites data
2nd i reset gps
and after closing this app , i open my gps soft and it works , i use iGO

Why do i always hear people with problems my gps works fine since day one. even on the old roms. only with these newer ones i get real fast lock.
never had any problem with navigation software and driving my car
i a using sygic mobile maps because i dont wanna waste data with google maps.

And if your unit performs worse than everyone else's SGS, then you might consider bending the contacts that connect the GPS antenna inside. Seems they don't make good contact on many units.
There is a thread about it somewhere.
Mike
Garbled meaning induced by swype when posting from XDA app on SGS I9000.

at the beginning gps was not very good but after android 2.2 and newer fw gps is totally fine for me. outside and in car it works very good. btw data network is activated for me and it helps for faster fix...

I gave up on the Internal GPS after a few Months.
Purchased an External Bluetooth GPS Receiver and now have accuracy to around 5 meters at times.
Also easier as the GPS Receiver can be turned on the get Satellite lock then connected to the Phone.

GPS on JVO ROM is all you can get. It's not enough to lead you driving on a city. GPS antenna is too small, does not update position often enough, keeps using guessing instead of real GPS position (your previous speed is used to calculate the supposed position, something very bad when you're turning around many times).
It's good enough to use it walking outside the city (main reason for me), or in highways (smooth curves and slow change of direction), but the GPS on the Galaxy S is real ****. Worst than the one in my previous two phones, and worst than the one in many current phones.
I really don't understand why having a phone that big we keep having to deal with so much unuseful plastic. 1.5 cm above an below the screen. Just make the phone the same size as the screen. The speaker can go in the top, like in many Motorola models. The battery could be bigger. If they have to make an GPS antenna that's 5 cm long, make it, running along one side.
It's like the phone hanging after a call. I would not advise this device for anyone depending of his phone to receive critical calls at any time. This **** just hangs radomly and you have to force a reboot.

Just to let people know that I've spent all afternoon trying 3 different ROMS, 2 times each, and the GPS sensivity is better on XXJVO.
I flashed the deodexed versions from Ramad, and used GPS status app to test. Just flash the ROM, install GPS status, see how many satellites it detects.
JVJV9 and XXJVP detect 1 or less while on my desk
XXJVO detects 4 and gets a fix, and sometimes it mantains the fix
Next to a window, JVJV9 and XXJVP detect 5 or 6 satellites, get a fix but lose it, most of the time 0 satellites in green.
XXJVO detects 8 satellites, 7 in green all the time
Accounting for the fact that satellite positions change between one test and another, I repeated tests twice.
I tried XXJVP on the field last weekend and was unable to get a fix while on the car.
Phone is sitting now on my desk, and has 4 satellites in green and intermitent fix. Much better than the other ROMs that see 0 or 1 satellites sitting on the same spot.

SKeijmel said:
Why do i always hear people with problems my gps works fine since day one. even on the old roms. only with these newer ones i get real fast lock.
never had any problem with navigation software and driving my car
i a using sygic mobile maps because i dont wanna waste data with google maps.
Click to expand...
Click to collapse
Gmaps 5.x and above has cache option...I'm finding very useful. Plus offline navigation should be coming this summer

Personally, I've given up on the GPS.. Samsung should honestly be sued, but it is too expensive to do so.
It has gotten better with newer ROM's, but, the sensitivity still appears to be a joke compared to my years old garmin forerunner. It may be good enough for some people who use it for navigation (because, I'd imagine there are plenty of algorithms which can probably be employed). However, the track quality produced seems semi-laughable, and the GPS appears to be nowhere near the quality of a dedicated unit.
People also need to stop checking programs like "gps status". They don't tell you if the location is correct (because it wont show you on the maps). So the accuracy value means nothing (the accuracy value means nothing without RAIM/FDE anyway regardless of unit).

GPS doesn't work at all on my SGS.
I tried that 'push the upper part of the rubberplastic thingy' trick, no result.
Whenever I try to find my location in Google maps, it can only find the WIFI center about 1km away from here. I think that's thanks to the Use Wireless Networks setting. but the GPS in my own SGS has never ever succeeded in locking any satelites.
Must be a hardware fauly I guess?
I bought it in October IIRC.
I don't mind though, never had the need to use GPS.
But it just bothers me that I got a somewhat malfunctioning device.
No big deal though!

Its been working fine on mine.. The only time i had issues was with 2.2/2.2.1
Sent from my GT-I9000 using Tapatalk

Some data points:
The signal to noise levels reported by the test tools seem reasonable.
All GPS systems use averaging to get a lock.
All GPS systems use Kalman filtering, which filters and connects location and speed, to make the system work at all. The stronger the filter, the more stable the position. But this gives bigger problems in turns.
Dedicated receivers have a much bigger antenna and thus need less averaging to get a stable position.
Receiver software has 'modes' - for example walking and driving mode. In driving mode, movements at low speeds are suppressed.
GPS is still a work in progress; there must be a reason Samsung renamed LBStestMode into AngryGPS!