• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

All things GPS

Is this a poll?

  • Yes, it looks like a poll.

    Votes: 42 26.3%

  • No, there is no way this is a poll.

    Votes: 31 19.4%

  • Why in the world is there a poll here?

    Votes: 87 54.4%
  • Total voters
    160
This is a universal thread to discuss the GPS on the Samsung Captivate. If you are having trouble, need help or simply want to discuss the GPS this is the thread for it.

Please keep in mind the rules as well as the zero tolerance policy when posting.

All new GPS threads will be merged into this one.
 
Many GPS fixes are proposed without describing the underlined GPS issues in Galaxy S. This new thread will discuss the underlined GPS issues first and then evaluate the fixes.

Almanac can be viewed as a complete GPS satellite schedule (satellite ID, time, and position) broadcasting over 6 orbits by about 27 satellites currently. Since the almanac is so huge, a GPS receiver cannot store it completely. But intelligent collection of some relevant almanac for a GPS receiver is crucial to compute the first fix without blindly searching the sky.

Based different computing strategies needed, we can categorize the functions, related to computing GPS receiver’s position, into 4 types of tasks. Samsung Captivate and Vibrant are completely flunked in Task 0 and Task 1, shown below

Task 0: Automatic collection of your GPS receiver’s interested almanac (self-initialization)

Your GPS receiver should automatically collect some almanac information relevant to current position and store it locally for future reference. The best timing to do it is at time of turning on GPS receiver before running other GPS applications.

Task 1. First fix without sufficient almanac.

The GPS receiver should automatically find the missing almanac on-the-fly to compute the position (need to identify at least 3 or 4 satellites signals for 2D or 3D fix, and update the almanac information to the GPS receiver storage, and choose the best one GPS satellite to lock

Task 2. First fix with sufficient almanac

Need to compute current GPS receiver position efficiently and choose the best one GPS satellite to lock

Task 3. Tracking current receiver’s position continuously, by using the locked satellite, previous position, assisted information such as WiFi, cell towers, etc.

All posted GPS fix methods are considered “cheating” methods. Only one reliable cheating method was posted: to run other “intelligent” GPS applications first such as LabTestMode, GPS Status for Task 1. The main side-effect (purpose) is to collect the missing almanac for the GPS receiver. Thus, other GPS-based applications such as Samsung geo-tagging camera, Google maps, etc only need to successfully perform Task 2 and Task 3.

Once a dumb GPS receiver is given with sufficient almanac, for instance, too many testings by using GPS Status, or LabTestMode in an area, it can “miraculously” get a quick GPS satellite fix for Samsung geo-tagging camera, Google maps, etc. without need of running those intelligent GPS applications first.

Note my favorite cheating settings are: GPS plus off, WIFI/ wireless,Skyhook off to make sure my dumb GPS receiver to store the almanac truthfully from GPS Status or others. The settings themselves would not help too much in Task 0 and Task 1. How the GPS algorithm/strategy inside the phone for collecting almanac is important.

Without capabilities of Task 0 and Task 1, this proposed fix will not work if your GPS receiver is transported to new regions far from the stored almanac. Your GPS receiver needs to re-train again from the intelligent GPS applications to feed local relevant almanac. Currently, the GPS receiver in Galaxy S phone cannot perform an unsupervised learning for automatic collecting the relavant almanac for first fix.
 
am i crazy or does my GPS work totally fine? ive used it 10 or so times in different locations and never really had a problem with it
 
Can't say I understand your analysis. But do you think a firmware fix is possible for the problem? Or would a completely satisfactory fix require a hardware mod?

BTW, like the other poster the GPS on my second Captivate seems to work. Have had it for 4 weeks now. If this is not a placebo effect, does it suggest that the problem is in the hardware (as no one has suggested that the firmware's been updated?)

DW
 
Let me clarify the terms first: GPS driver and firmware.

The firmware is inside GPS chip provided from the GPS vendor, Broadcom BCM4571 in captivate/vibrate phone BCM4571 is supposed to be a better version of GPS chip BCM4570 used in IPone 4.

The GPS vendor also provides SDK (Software Development Kit) to develop periperal device driver on Android OS side. The dirver development is Samsung's responisbility. Samsung need to provide a good GPS device driver to support higher level functional calls to other GPS applications and support the initialization

The Samsung GPS driver has a problem. Samsung does not implement the capabilities of task 0 and task 1 properly. It needs a reliable unsupervised learning method to collect almanac automatically. Currently the almanac can only be feeded manually by other applications such as GPS Status, GPSTestMode, others.
 
I think for many, Task 3 is also a bit of a failure. We can see that from tracks that jump all over. I can stand completely still, and my GPS will sometimes show me jumping around a 200m square area. I've been wondering lately if this is not a "GPS Fix" issue, but maybe an issue with lag within the phone? Would the phone's internal lag cause information to be fed in a bad timing sequence, resulting in a "jumping around" pattern? I don't know that much about the way GPS interprets data, but when I see the pattern of the jumping around on my screen, it seems to be similar to the lag sequences I see in other motion-based activities.
 
Yes. the task 3 is not perfect in the Samsung GPS driver.
Just summerize the factors in computing GPS receiver position from Satelitte signals only for your reference.

Uncorrected positions from GPS satellite signals The accuracies are in the range of 50 to 100 meters.

When using correction techniques and below GPS error budget, the good GPS driver should get positions accurate to within 5 meters or less

GPS Error Budget
Source 50-100 meters (uncorrected error range)
Ionosphere 0-30 meters
Troposphere 0-30 meters
Measurement Noise 0-10 meters
Ephemeris Data 1-5 meters
Clock Drift 0-1.5 meters
Multipath 0-1 meter
Selective Availability 0-70 meters

see details Introduction to GPS
 
I think I do see some positive news in this: If I understand you correctly, this is a software problem.

" The dirver development is Samsung's responisbility. Samsung need to provide a good GPS device driver to support higher level functional calls to other GPS applications and support the initialization"

So, it is fixable. If it were a hardware problem, I presume it is not fixable.

So theoretically Samsung is working on this, and ultimately we should have a fix.
 
Yes. I think it is a GPS software driver issue. If it is from GPS hardware side, then these phones should be recalled in the future since it is a manufacture defect. If it is from GPS firmware, then it can be updated in the future. Thus, we do not need to worry too much in either case.

But there are not good signs from Samsung for dealing with GPS issues technically so far.

1. Samsung provides a note to AT&T for turning-on wireless for
this GPS issues
2. The level 3 Samusung support told me that Samsung already worked
out the solutions and doing side-by-side (both Captivate and
Vibrate phone) testing, and to be released in September


Too soon to be true! I am not sure that Samsung understands the issues correctly. To overhall the GPS driver is not an easy task, particularly there is no track record for any GPS solutions from Samsung.
 
ewingr said:
I think I do see some positive news in this: If I understand you correctly, this is a software problem.

" The dirver development is Samsung's responisbility. Samsung need to provide a good GPS device driver to support higher level functional calls to other GPS applications and support the initialization"

So, it is fixable. If it were a hardware problem, I presume it is not fixable.

So theoretically Samsung is working on this, and ultimately we should have a fix.
Click to expand...

This issue was supposedly fixed in the Epic before it was launched. If the fix required more than a simple software change then I don't think that would be possible. I doubt they used a different GPS chip between the captivate, vibrant and epic.
 
What Dr. GPS says makes perfect sense. I have posted before that Runkeeper (for example) provides a good kick-start (i.e., Tasks 0 and 1) for the Captivate GPS. Once started, other GPS apps appear to work ok. Clearly a software (or firmware) issue!
 
How well does Runkeeper work for you, over time? I've been trying to use it, but it freezes up constantly. it'll go 45 seconds without any updates on screen, then occasionally lock up to the point that the phone decides to terminate it. Sorry, a bit off-topic, just curious if you're able to use it for more than a jump-start.
 
Wow, thanks to someone for finally opening an inteligent discussion on the issue rather than just the complaining that seems to be rampant.

I am no expert in GPS, but my hypothesis on this issue is that; Since some peoples seem to work propperly, and mine worked perfectly for a few weeks before crapping out. Added to that many people have found various tinkerings to "fix" or improve it's performance is that;
Whatever the problem is it seems to be related to something screwed up in the software, including the drivers and or firmware. So, if that being the case, it is possible that Samsung, or even someone else could figure out what the error is and fix it. Its seems the chip is fine, but the software controlling it is not. Does the Epic and Fascinate have the same GPS chip as the Capt. and Vib.? Because as I understand, the GPS work on those ones.

If it where a hardware problem, wouldnt the GPS not work at all, ever? In my case, since mine worked at one point, I dont think my chip broke itself over time, i think something is wrong with the software.
 
To those saying yours works fine, have you used the turn-by-turn nav or tried to get a lock while moving? Mine gets always gets a lock while I'm stationary, but the location is almost always a couple blocks away from where I actually am. On a rare occasion, it has been exact. Once I start moving, it goes haywire.
 
Segfault said:
This issue was supposedly fixed in the Epic before it was launched. .
Click to expand...

Yeah, I've seen that stated several times for the Epic and for Verizon's version of the SGS. I wonder who says that. Only now is the Epic actually available, and the Verizon one isn't yet available. I'm skeptical that anyone has had enough use of them to tell. I've read too many times that someone says theirs is fixed, only to reply back days later that, well, no it isn't.

I'm also skeptical that those who say theirs has always worked fine use them that much or in the right places that shows problems. The one I had worked perfect most of the time, but occasionally was completely nuts.

And if they are fixed, then why can't the get the fix out for the Captivate now?

I'm still hopeful that September will bring the fix.
 
Mr. GPS said:
When using correction techniques and below GPS error budget, the good GPS driver should get positions accurate to within 5 meters or less
Click to expand...
Great thread. And I agree with the above statement.

Just because a phone locks on a few sats and gives a location of 10 meters or more does not mean it's working. It's good enough for some GPS duties but not others that require a more precise location fix.
 
I pretended I was buying a phone and started asking questions, I asked the sales lady at best buy if her GPS was working and she said it worked great and she used it all the time. She had her phone rooted and everything. She offered to prove to me that the GPS does not have a problem. We walked out the front of the store and she pulled up google maps and after 30 seconds had a lock in the general vicinity of the parking lot. She said look, it works perfect. I had her download GPS test and watch it. It had an accuracy of 300 meters and was dropping and reaquiring sats constantly. She doesn't think her GPS is perfect anymore.
 
I explain the reason why I think it is a GPS software driver issue and the BCM4571/firmware provides sufficient and fair signals for GPS drivers to use or functionl call.

Under the same condition, some GPS drivers can do better jobs than others based on the algorithm and strategy used inside the drivers. I love to share my favorite comparison GPS test in a AT&T store.

This AT&T store (brick building, glass door with WIFI and wireless phone network) provides the demo phones
(A) iPhone 4,
(B) One Captivate near glass door
(C) one Captivate near iPhone 4 (both more away from glass door)
(D) HTC Aria.
Also I carries my own trained Captivate phone (E) by using GPS Status for storing almanac information covering the store area.

As I expected, initially, phones A, D, and E are OK on GoogleMaps. it is a pity that two demo Captivate phones B and C performed very badly on GoogleMaps and could not lock at all.

But I can train phone B to make it work shortly by using LabTestMode, not phone C(weak signals and far away from glass door). I could not download GPS Status to these two demo phones inside the store to tain phone C. However, when I moved my own phone E into the same position of B or C for testing, there is no problem at all on GoogleMaps.

It is a GPS driver problem!! Also, I believe the third-party GPS Status uses its own GPS driver to obtain the signals from the same BCM4571 instead of Samsung GPS driver. Broadcom claims BCM4571 chip is more sensitive to weak signals than previous versions such as BCM4570 in iPhone 4.

Sometimes, I feel that Samung changes the name from A(ssisted)-GPS to A(bused)-GPS.
 
I am glad to hear that someone has finally posed a rational hypothesis of the GPS problem and possibility of a fix in the future. For my needs it works fine, Google Maps is another story, I know my neighborhood very well and my location is within 20 meters or so of where I am but the map gives inncorrect directions. It says that a location is on the wrong side of the street. This makes me nuts, the GPS functions and Google Maps are wrong. I don't use GPS on the phone, if I want that I will buy a GPS.

Is there a better app for directions than Google's? I can sideload apps too so I don't care about AT&T's market ristriction.
 
Thank you for this excellent thread.

So I learned a new trick from this and other GPS threads.

1. Start AT&T Navigator.
2. Let it sit on the Terms of Service page for a few seconds.
3. Click Exit. (Don't agree, don't sign up, don't pay anything.)
4. Use GPS Test, Maps or other apps that need GPS

I get an excellent lock on all of a sudden. It's accurate and stays accurate. Typically 15-30 meters accuracy indoors, sometimes 5-10 meters outdoors or in car.

So that AT&T bloatware is good for something. More proof it's a software fix.
 
You must log in or register to reply here.
Back
Top Bottom