Poor GPS performance.

[Flip]

I had the same problem n called sprint n they sent some kind of update to the phone n the issue was fixed

 

[Aero]

I had the same problem n called sprint n they sent some kind of update to the phone n the issue was fixed

Nope, they just reset the cache, it sticks again right same as always. I spoke with Advanced Device support for Epic (the highest internal tech support for a device) this morning there is no fix for the cache sticking.

 

[boomerbubba]

However, if you turn on Airplane Mode and reboot the phone, the cache is cleared, with the data/cell radio off the network almanac/ephimeris data is not received and when you turn the GPS on it must then get this info directly from the satelite... that is what a standalone GPS such as those on most HTC phones and all dedicated GPS navigation units do. It takes longer to acquire that data and then fix your position, that is why with the network on it can lock very quickly as the data is downloaded very rapidly from the network. The problem with my Epic is that it does not seem able to download the almanac info directly from the satelite, therefore it can never get a fix where my Evo and even my Treo Pro (HTC built Palm branded handset) can do so and get a fix in about 30-45 seconds under similar conditions.

My experience is different. My Epic GPS seems to work fine without a data connection in Standalone mode, when set up for that purpose.

Yesterday, I happened to be driving across Texas. There are places in West Texas, ranch country miles off the interstates, where there is zero data coverage. So I took such a detour to test my Epic off the grid.

After I had driven to a waypoint known to have no coverage and confirming that empirically, I followed these steps:

1) Ran the GPS Status utility's function to clear the GPS cache. Did not attempt to reload the ephemeris data from the net, and could not have done so anyway because there was no network available.

2) Ran the GpsSetup utility on the Epic by entering *#1472365# from the dialer. Then selected Setup -> Position Mode -> Operating Mode -> Standalone.

3) Rebooted.

4) Started GPS Test. A lock was achieved, using 4 satellites, in 25 seconds. Over time more satellites were tuned in and used, so the GPS settled into its routine showing multiple satellites on GPS, comparable to the G1 benchmark unit sitting next to it on the passenger seat of my car. On both of them, the number of satellites used varied as I drove -- I noticed as few as 5, sometimes as many as 11, mostly about 7 or 8. But because I was driving alone, I couldn't watch fulltime.

5) The actual tracks of the two phones, captured using My Tracks, were comparable. Generally, the G1 tended to show 3.0 meter estimated accuracy, while because of the known bug the Epic always showed 30.0-meter estimated accuracy. (This is consistent with earlier testing in the default MS Based mode.)

All this off-the-grid testing was done inside the car, with skies 100 percent overcast and rainy. I repeated the test a few miles later with similar results.

 

[AndroidOne]

I am going to set the parameters as you did on my phone... as matter of fact I just did. I will be running the phone in airplane mode (to simulate no carrier data state) and I will try to get a fix as soon as I get out of the house and see what happens.

Have you ever tried getting the GPS to work while the phone is in airplane mode? Someone else suggested the GPS radio may be turned off by this mode, although I do not believe this to be true it sure feels like it on my phone so far. Can you please try to run your GPS in airplane mode and report back your experience (I feel is my unit that is defective and not the Epic line in this regards).

 

[boomerbubba]

Have you ever tried getting the GPS to work while the phone is in airplane mode? Someone else suggested the GPS radio may be turned off by this mode, although I do not believe this to be true it sure feels like it on my phone so far.

Just tried my phone in airplane mode, which seemed to disable the GPS functionality (as seen though actual performance in GPS Test, which could not even view any satellites and zeroed all the data). However, this had no effect on the GPS on/off indicator that controls the "GPS Status" indicator on the GPS Test screen, or on the Epic's pulldown notification bar.

When I turned off airplane mode, the GPS worked again.

 

[AndroidOne]

Just tried my phone in airplane mode, which seemed to disable the GPS functionality (as seen though actual performance in GPS Test, which could not even view any satellites and zeroed all the data). However, this had no effect on the GPS on/off indicator that controls the "GPS Status" indicator on the GPS Test screen, or on the Epic's pulldown notification bar.

When I turned off airplane mode, the GPS worked again.

That is exactly what I was trying to prove... the Epic does not seem to have real standalone GPS capability like the Evo and others whose GPS locks/works perfectly from airplane mode without any time reference from the network like the Epic seems to require. Thanks.

Once the GPS locks, turning airplane mode does not "break" the GPS so the chip is certainly working. But getting any kind of initial fix from airplane mode (cold start) so far has proven impossible.

 

[boomerbubba]

That is exactly what I was trying to prove... the Epic does not seem to have real standalone GPS capability like the Evo and others whose GPS locks/works perfectly from airplane mode without any time reference from the network like the Epic seems to require. Thanks.

No. The Epic does have a Standalone mode, which is what I tested in a no-network area and you did not.

But in any case, my simple test of toggling airplane mode is not so definitive after all. I just repeated the test, and now the GPS works even when airplane mode is on!

I can't figure out what, if anything, I did differently. The only clue is that the first time, I had just experienced the known bug where the Epic GPS fails to lock after a period of inactivity, and must be rebooted. Maybe that affected this test somehow.

 

[AndroidOne]

No. The Epic does have a Standalone mode, which is what I tested in a no-network area and you did not.

But in any case, my simple test of toggling airplane mode is not so definitive after all. I just repeated the test, and now the GPS works even when airplane mode is on!

I can't figure out what, if anything, I did differently. The only clue is that the first time, I had just experienced the known bug where the Epic GPS fails to lock after a period of inactivity, and must be rebooted. Maybe that affected this test somehow.

Try this:

1. Turn airplane mode on, this will eliminate any assistance from the network as the cell/data radios are off.

2. While on airplane mode, reboot the phone, this should clear the GPS data cache and reboot the phone still on airplane mode so no new GPS cache data can be obtained.

3. Now while still on airplane mode, use the GPS Test (or other utility) to get a fix and let me know if you can.

If you are able to get a fix like this, the GPS most be getting the almanac/ephimeris and time reference data directly from the satelites and therefore working on full standalone mode. The Evo (and for that matter, my PPC6800 and Treo Pro as well) is able to get a lock in less than 30s under these circumstances, so far I have not been able to get a GPS fix on the Epic like this.

 

[boomerbubba]

Try this:

1. Turn airplane mode on, this will eliminate any assistance from the network as the cell/data radios are off.

2. While on airplane mode, reboot the phone, this should clear the GPS data cache and reboot the phone still on airplane mode so no new GPS cache data can be obtained.

3. Now while still on airplane mode, use the GPS Test (or other utility) to get a fix and let me know if you can.

If you are able to get a fix like this, the GPS most be getting the almanac/ephimeris and time reference data directly from the satelites and therefore working on full standalone mode. The Evo is able to get a lock in less than 30s under these circumstances, so far I have not been able to get a GPS fix on the Epic like this.

Why would I need to do that, trying to emulate being off the network? I already did such a test yesterday when I was actually off the network in the West Texas boondocks. The Epic worked fine, both in its initial lock and in its tracking performance, with no network connectivity whatsoever.

 

[AndroidOne]

Why would I need to do that, trying to emulate being off the network? I already did such a test yesterday when I was actually off the network in the West Texas boondocks. The Epic worked fine, both in its initial lock and in its tracking performance, with no network connectivity whatsoever.

Because what I want to know is if MY Epic's GPS is malfunctioning or other Epics behave in a similar fashion. There are plenty of reports that airplane mode "breaks" the GPS but no one is being specific about it and no one is willing to do this simple test to find out if indeed this is true or not. Could it be possible that the Airplane mode actually turns-off the GPS chip as well despite the symbol appering on the top bar, who knows, that is what I want to find out.

If is the first, then I can exchange my unit before the 30-day period, if the latter, then we need to see if this is a fixable issue.

 

[boomerbubba]

Because what I want to know is if MY Epic's GPS is malfunctioning or other Epics behave in a similar fashion. There are plenty of reports that airplane mode "breaks" the GPS but no one is being specific about it and no one is willing to do this simple test to find out if indeed this is true or not. Could it be possible that the Airplane mode actually turns-off the GPS chip as well despite the symbol appering on the top bar, who knows, that is what I want to find out.

Well, that may be worth exploring. It did appear, from my first attempt at a test, the the GPS radio had actually been turned off despite the indicator that it was on. But obviously, that did not hold up in my second test (and subsequent attempts). The GPS did work with airplane mode set. So there may be some specific problem specifically associated with airplane mode settings, or not. The results are inconclusive, at best because the behavior with airplane mode was not consistent.

[EDIT: I just did your test. Sure enough, if I turn on airplane mode, then reboot, the GPS radio does not seem to function, even though the GPS-enabled indicator is on. So this shows some anomaly related to airplane mode. I am not sure what the defined behavior of airplane mode is supposed to do in Android: Is the GPS supposed to be on or off? Or is this not defined and thus left to the implementor?]

But in any case, the hypothesis "that the Airplane mode actually turns-off the GPS chip" is a very different hypothesis from your earlier, over-generalized and speculative hypothesis that the Epic GPS just cannot work without a network connection. I disproved that hypothesis. It works fine.

So some other, narrower hypothesis is needed to explain the airplane-mode behavior.

 

[AndroidOne]

Well, that may be worth exploring. It did appear, from my first attempt at a test, the the GPS radio had actually been turned off despite the indicator that it was on. But obviously, that did not hold up in my second test (and subsequent attempts). The GPS did work with airplane mode set. So there may be some specific problem specifically associated with airplane mode settings, or not. The results are inconclusive, at best because the behavior with airplane mode was not consistent.

[EDIT: I just did your test. Sure enough, if I turn on airplane mode, then reboot, the GPS radio does not seem to function, even though the GPS-enabled indicator is on. So this shows some anomaly related to airplane mode. I am not sure what the defined behavior of airplane mode is supposed to do in Android: Is the GPS supposed to be on or off? Or is this not defined and thus left to the implementor?]

But in any case, the hypothesis "that the Airplane mode actually turns-off the GPS chip" is a very different hypothesis from your earlier, over-generalized and speculative hypothesis that the Epic GPS just cannot work without a network connection. I disproved that hypothesis. It works fine.

So some other, narrower hypothesis is needed to explain the airplane-mode behavior.

I have a question for you but before that let me layout some results of recent trials:

1. I went out and turned on the GPS with the network radios on and it got a fix without a problem. While still navigating I put the phone on airplane mode and the GPS continued to work for the rest of the trip ~ about an hour. This clearly indicates to me that the GPS chip IS indeed working regardless of the cell/data radio status, otherwise the GPS reception would have been lost.

2. I then rebooted the phone while still on airplane mode ~ I hope you agree this should have the effect of erasing any data contained on a cache, including that from the GPS. When I attempt to get a GPS fix after this (cell/data radio off), I can't, no satilites are even visible. This suggests to me the GPS radio is working (proven from #1) when the other data/cell radio is off, but seems unable to get the needed information to get an initial fix by itself.

3. While the GPS is still trying (unsuccessfully) to get an initial fix, data/cell radio are turned on and within a few seconds the GPS starts seeing satelites and gets a fix a few seconds later. This suggests to me some information was passed to the GPS after activating the radios that allowed the chip to finally get the initial fix.

Based on the above I am going to conclude the following:

1. The GPS on the Epic after an initial fix, is capable of maintaining a fix with or without network assistance (once working, it continues to work even when data/cell radios are turned off).

2. The Epic GPS uses network assistance to obtain an initial fix.

3. The Epic GPS seems unable to obtain an initial fix in the absence of network assistance eventhough it does not seem to need such assistance for continued operation once the initial fix is locked.

4. I define standalone GPS as a device that being a self contained unit is able to download almanac/ephimeris information and time reference directly from the satelite signal in order to establish an initial fix with no outside references needed.

So given the above, wouldn't you conclude the Epic's GPS is not able to work in standalone mode? If not, then how do you explain the fact that the Epic is unable to get an initial fix with radios off and empty cache?

BTW: The Evo, can get an initial fix with data/cell radios off after a reboot (like the tests conditions on #2 above) usually in less than 30-45 sec depending on obstructions.

 

[wrxdrunkie]

Ya, this thing takes muuuch longer to get a gps lock compared to my evo, much much longer.

 

[AndroidOne]

Ya, this thing takes muuuch longer to get a gps lock compared to my evo, much much longer.

Agreed, but can you actually get a fix using the test conditions described on #2 above?

 

[boomerbubba]

I have a question for you but before that let me layout some results of recent trials:

1. I went out and turned on the GPS with the network radios on and it got a fix without a problem. While still navigating I put the phone on airplane mode and the GPS continued to work for the rest of the trip ~ about an hour. This clearly indicates to me that the GPS chip IS indeed working regardless of the cell/data radio status, otherwise the GPS reception would have been lost.

2. I then rebooted the phone while still on airplane mode ~ I hope you agree this should have the effect of erasing any data contained on a cache, including that from the GPS. When I attempt to get a GPS fix after this (cell/data radio off), I can't, no satilites are even visible. This suggests to me the GPS radio is working (proven from #1) when the other data/cell radio is off, but seems unable to get the needed information to get an initial fix by itself.

3. While the GPS is still trying (unsuccessfully) to get an initial fix, data/cell radio are turned on and within a few seconds the GPS starts seeing satelites and gets a fix a few seconds later. This suggests to me some information was passed to the GPS after activating the radios that allowed the chip to finally get the initial fix.

Based on the above I am going to conclude the following:

1. The GPS on the Epic after an initial fix, is capable of maintaining a fix with or without network assistance (once working, it continues to work even when data/cell radios are turned off).

2. The Epic GPS uses network assistance to obtain an initial fix.

3. The Epic GPS seems unable to obtain an initial fix in the absence of network assistance eventhough it does not seem to need such assistance for continued operation once the initial fix is locked.

4. I define standalone GPS as a device that being a self contained unit is able to download almanac/ephimeris information and time reference directly from the satelite signal in order to establish an initial fix with no outside references needed.

So given the above, wouldn't you conclude the Epic's GPS is not able to work in standalone mode? If not, then how do you explain the fact that the Epic is unable to get an initial fix with radios off and empty cache?

BTW: The Evo, can get an initial fix with data/cell radios off after a reboot (like the tests conditions on #2 above) usually in less than 30-45 sec depending on obstructions.

No, I don't necessarily agree with the multiple assumptions you have made. I don't think you have any more certainty than I have about what happens internally when you reboot in airplane mode. You are guessing.

As for your conclusion that "the Epic's GPS is not able to work in standalone mode," I emphatically disagree. Because I have actually set mine to Standalone mode using the onboard utility with that expressly named setting, cleared the GPS cache, rebooted and SUCCESSFULLY OPERATED THE GPS WITHOUT ANY NETWORK AVAILABLE, BOTH IN GETTING AN INITIAL FIX AND IN SUBSEQUENT TRACKING. (Sorry for the all-caps, but you stubbornly refuse to give up your theory to accommodate this rather salient empirical fact.)

As for your rhetorical question: "how do you explain the fact that the Epic is unable to get an initial fix with radios off and empty cache?" The answer is simple: I don't. That is because I don't know, and neither do you. In scientific testing, sometimes "I don't know" is the most valid conclusion to make until further facts are established and someone says Eureka. But just because I don't have an alternative hypothesis to offer yet, that does not mean that yours is true. I think the facts of my empirical test in Standalone mode show your hypothesis to be false. My GPS worked in Standalone mode without a network. Period.

So we both await some other hypothesis that explains all the facts and is not disproven. I suspect maybe something is going on in your test that is specific to airplane mode. Maybe some programmer at Samsung just thought, correctly or not, that GPS is supposed to be disabled in airplane mode. Or the phenomenon may even be related to the known bug that causes the GPS locking to fail after some period of disuse, which also seems to be mysteriously related to the cache.

But these are just speculative suggestions. Unlike you, I do not pretend to have the answer.

 

[AndroidOne]

No, I don't necessarily agree with the multiple assumptions you have made. I don't think you have any more certainty than I have about what happens internally when you reboot in airplane mode. You are guessing.

I agree with your disagreement, I am just making an assumption based on my empirical observations and I do accept that my assumptions may be incorrect.

As for your conclusion that "the Epic's GPS is not able to work in standalone mode," I emphatically disagree. Because I have actually set mine to Standalone mode using the onboard utility with that expressly named setting, cleared the GPS cache, rebooted and SUCCESSFULLY OPERATED THE GPS WITHOUT ANY NETWORK AVAILABLE, BOTH IN GETTING AN INITIAL FIX AND IN SUBSEQUENT TRACKING. (Sorry for the all-caps, but you stubbornly refuse to give up your theory to accommodate this rather salient empirical fact.)

Not stubborn, I am looking for an explanation and thought having an exchange of ideas with you and perhaps additional troubleshooting might get us closer to uncover the causes.

As for your rhetorical question: "how do you explain the fact that the Epic is unable to get an initial fix with radios off and empty cache?" The answer is simple: I don't. That is because I don't know, and neither do you. In scientific testing, sometimes "I don't know" is the most valid conclusion to make until further facts are established and someone says Eureka. But just because I don't have an alternative hypothesis to offer yet, that does not mean that yours is true. I think the facts of my empirical test in Standalone mode show your hypothesis to be false. My GPS worked in Standalone mode without a network. Period.

I do certainly accept "I don't know"; in fact, I don't know myself... that is the reason I seek additional testing.

Unlike you, I do not pretend to have the answer.

I don't have any answer, that is why I am here... sorry if me asking offended you in such grave maner; I won't ask no more.

 

[boomerbubba]

Regarding airplane mode simply turning off GPS, the Occam's Razor answer might just be that the phone is working that way at least partly by design.

According to this post at a Google support site for the Nexus One, which is supposed to be the official dev reference model for the current generation of Androids, that is what occurs on the N1. So whether or not one agrees that is what airplane mode ought to do -- and there seem to be many opinions out there about how airplane mode "ought" to work, and maybe Evo programmers had a different idea -- Samsung programmers might be forgiven for defining airplane mode the way Google did: that airplane mode disables GPS along with other radios.

What is odd is that the GPS still seems to work if airplane mode is selected without rebooting. So there may be a bug here one way or another.

Edit: Or, maybe this is an issue within the Android OS. See this item, and follow the links. Just what the expected behavior is seems to be ambiguous

 

[cb2367]

I got the Epic last week, and right away I had problems with it.

When I ran google maps, it wouldn't scroll to my location. When I tried to make the application locate me, it first displayed 'waiting for location' and then a few seconds later, it displayed 'your location is temporarily unavailable.' I had to have wireless networks and GPS satellites turned on, just to have a location come up. With both of those settings turned on, the location that came up, was at least a mile away from where I was located. In one instance, Google maps plotted my position and when I looked at the details, it said that the location was accurate to within 2200FT. I've tried getting Google maps to correctly locate me both inside and outside without any success.

An application that I had installed on my phone called RunKeeper used the GPS to track distance travelled. When the application is working correctly, the GPS signal strength shows a green dot. It would only show an orange dot.

I spent all weekend calling Sprint and going to the Sprint stores to get this issue resolved. They did everything from refreshing the data services on their end to doing a factory reset. They even had me go to the dialpad and enter GPSCLRX and reset the GPS. All this had no effect on fixing the problem.

In my calls to Sprint, they wouldn't do anything more than reset the phone. When I tried to explain about the RunKeeper app which uses the GPS signal, they then stated that this was a third party app and there wasn't any guarantee that the app itself was faulty. I upgraded to the Epic from the Hero and I can say that I had no problems with Google Maps or the GPS on the Hero phone.

I finally decided that since I hadn't yet exchanged the Epic for a new one, that I would go ahead and do that since I could then say that I have tried all options. I've read through some of the topics in this thread, and I am going to try My Tracks and GPS Test and see what results I get.

I did find that Google Maps would not work unless I had wireless networks and GPS satellites turned on. I noticed that the Sprint NAV app would not update my location unless I had the WiFi on and connected. The Google Maps provided a better and more accurate location with the WiFi turned on. I don't know if all the issues I am experiencing are because of a low signal (2-3 bars inside) or if there is something else going on. I did try the GPS outside with Google Maps and got the messsage about my location being temporarily unavailable.

 

[wynand32]

Okay, just bought an Epic today to replace my Palm Pre. So far, the GPS is essentially unusable. In fact, I'd have to say it's just not working. I can get a fix to within 30m according to Smartrunner (although Maps seems to be accurate) with the wifi location turned on. But, it never gets closer than that, and so SmartRunner (and Runkeeper, which I also tried) won't work. Also, sometimes when SmartRunner goes to the background and then is brought back to the foreground, there's no attempt at all to access GPS.

I haven't read through the entire thread, but has anyone returned an Epic and had a replacement work as expected? I can't imagine that this thing is anything but defective the way it's working right now...

 

[boomerbubba]

Okay, just bought an Epic today to replace my Palm Pre. So far, the GPS is essentially unusable. In fact, I'd have to say it's just not working. I can get a fix to within 30m according to Smartrunner (although Maps seems to be accurate) with the wifi location turned on. But, it never gets closer than that, and so SmartRunner (and Runkeeper, which I also tried) won't work. Also, sometimes when SmartRunner goes to the background and then is brought back to the foreground, there's no attempt at all to access GPS.

I haven't read through the entire thread, but has anyone returned an Epic and had a replacement work as expected? I can't imagine that this thing is anything but defective the way it's working right now...

Swapping the phone will do nothing. The GPS bugs are in the firmware shared by all Epics. Neither Sprint nor Samsung acknowledge them to be bugs. But users are hoping they get fixed anyway.

The 30-meter estimate of accuracy is itself a bug. Instead of reporting a dynamicly calculated estimate of accuracy, Epics are hard-coded to always report 30 meters. The actual accuracy, relative to your actual position on the ground, is actually pretty good -- when you have a satellite lock. So the 30-meter estimate should be ignored. Some apps might treat it as significant, however.

(This 30-meter bug is reported to be fixed in the next release, along with Froyo. That is, assuming that is ever released at all.)

But the worst GPS bug is in failure to get satellite locks over time, as the GPS cache becomes stale. There is a workaround for that bug. When using that Cold Start workaround, make sure you only use 3G data service when acquiring the initial satellite lock. Then you can switch to 4G.

If you can't live with these shortcomings, I advise you to return the Epic while you can and wait. Then, after the Froyo release, you can try buying another one and maybe the GPS bugs will be fixed. (Anecdotal reports, without controlled testing, indicate that they might be.) The Froyo release is supposed to be out Real Soon Now, if you believe that.