boomerbubba

Correct. However, HTC's FieldTrial.apk, which contains the various engineering screens, has not changed significantly since at least Froyo.

 

AJ

Thanks again. The reason I asked about Ice Cream Sandwich is that I have read various places that at the operating system and Android API level, which probably affects most applications, ICS devices report LTE dBm signals in a different manner than Gingerbread devices do, and the reported dBm might differ by about 10 dB. (This does not apply to EVDO or 1X levels.)

Do the dBm levels that the HTC FieldTrial app report for LTE signal match those reported under the Settings -> About Phone -> Etc..... screen?

I use HTC's own LTE Engineering screen. I am not a fan of third party apps for this purpose. Or, at least, I have never found any app particularly to my liking.

 

AJ

Thanks. Just to confirm, your handset is running Ice Cream Sandwich?

@WiWavelength:

What handset, operating system and application are you using to report the LTE signal strength?

No, sorry, none of those are Sprint sites. That app (or the sites themselves) is not giving you accurate locations.

 

AJ

Sounds like this market includes examples of towers that squawk offset coordinates for each sector, rather than simply the coordinates of the tower. Dkoellerwx's screen was obviouly from Netmonitor, which reports the coordinates the phone picks up from the base station broadcasts.

The problem with the NetMonitor app accuracy is not the app's fault. It's Sprint's fault. For whatever reason, some Sprint market sites do not provide accurate info to NetMonitor. However, in other markets its fine. Like here in New Mexico, NetMonitor is wonderfully accurate.

 

Sprint has no interest in fixing the issue, because they don't want you to know which site you're connected to. Some carriers even block the info. Have you checked out our site maps?

 

Robert

Just to amplify on Robert's response, Netmonitor and several other Android apps, such as CDMA Field Test, are getting the coordinates from the underlying Android API, which is just capturing the coordinates broadcast by Sprint's base station transmitters. Each tower typically has multiple directional sectors (usually three). Each sector has its own base station ID, and under the CDMA standard it squawks coordinates that are programmed in by the carrier.

There seem to be two distinct cases:

• All sector base stations on a given tower broadcast the same coordinates, which are the actual location of the tower.
  
• Each sector base station on a given tower broadcasts different coordinates for just that sector, forming a triangle apparently proportional to the shape of that cell's antenna pattern. Those coordinates may even be miles away from the tower site.
  

We don't know why this. (There is a fuller thread discussing this phenomenon in the Sponsors' forum, but this is a decent summary.)

Correct. The Galaxy S3 uses either its Qualcomm MSM8960 SoC or an outboard Broadcom solution for GNSS. Regardless, both Qualcomm and Broadcom support GLONASS, as does the S3.

Do we know for sure from teardown reports, etc., which GPS chip the Sprint GS3 employs? And do we know which GPS chiip the other carriers (particularly the GSM carriers) use?

The reason I ask is that the first generation Galaxy S units had different GPS hardware and associated software across carriers, which led to much misiniformation and confusion in user forums because people ignorantly thought "the GPS" was the same on all carriers. But the only thing they all had in common was that Samsung and the carriers managed to screw up the GPS on all the devices, just in different ways.

With all the positive anecdotal reports out there about the different GS3 implementations this time, the common news is good, so few users will care to do much testing. But it is worth noting that details for one carrier's GPS may not apply to another's.

It's the same site. The site number is 61. The sector number connected to at that site is 8 or 9.

In my experience, the sector-based BSID numbers on a tower are sequential, but it is not as simple as trimming the last digit and calling it the site number. For example, one hypothetical tower could have BSIDs 198, 199 and 200. I have seen multiple examples of that. I think the BSID is just a sector number, and Netmonitor does not report a site number.

How far away are you from your closest cell tower, you can find out this information with Netmonitor.

.. unless it is one of those base stations that squawk coordinates that are not the actual location, but rather are programmed to be offset out some distance -- perhaps miles away -- for each sector. Netmonitor and some other utilities, such as CDMA Field Test, read these values from the Android API, which gets them from the base stations' broadcast. We don't know why this is the way it is, but Sprint's inconsistent programming of these coordinates on different base stations can confuse such utilities.

There is a Beltway 8 and Loop 610 in Houston.

And in Austin, there is Loop 1, which isn't a loop at all, but just another name for the MoPac Expressway that basically runs north-south.

From what I've heard Samsung relied on Google's coding for GPS and it had issues.

I doubt that very much, having been deeply involved in troubleshooting the GPS on the first generation Galaxies. Google does not write the low-level drivers and firmware that interfaced to the GPS chips and radios, the manufacturers did, and the carriers were responsible for their own acceptance testing. And Samsung screwed up the GPS firmware on multiple Galaxy platforms. (The original Galaxy S generation actually had different GPS hardware and low-level system firmware on different carrier-branded phones, and Samsung botched them all in different ways at different times.)

I am encouraged by all the positive anecdotal reports about the GPS performance on the GS3 models. I still have not seen any controlled testing that includes not just locking times and imputed accuracy from GPS Test, but also actual accuracy over the ground using My Tracks. When I get my hands on a Sprint GS3, I will do some testing to satisfy myself. But at least we don't see the forums full of posts saying "the GPS sucks," etc. Quite the opposite.

If they did merge, maybe this would be the new logo? Or not.

 

I don't care about the logo, but can we get the pretty girl in the pink dress?

No need to though. Write the 3g roaming prl and be done with it. Flip between native and roaming with roam control or unlock the roam only option on the network settings. No need to flash prls anymore.

I see from your BatonRouge site that you have a lot of experience with PRLs. I have two elementary questions:

1) How the heck does one read or edit a PRL file? They look like binary files to me.

2) Would the PRLs have to be updated to include new eHRPD sites? If so, would these have to be different on LTE-capable phones so they will connect to the eHRPD signal, but non-LTE phones would need a different PRL?

I would wait until the launch to see how much more your 3G speeds improved and make a decision at the end of July or early August.

I would not get my expectations up about improved 3G performance. My general understanding is that the effect on 3G should be positive, but marginal. The huge benefit of the NV rollout is 4G LTE, which does require using a new LTE-capable phone. So my advice would be to wait until then and try out a new phone such as the Galaxy S III during the usual 14-day return period.

Did folks register with Sprint asking for such notice? I recall doing so in a short form on the Sprint website.

Another naive theory: Could it simply be that in cases where the new NV network is installed side-by-side with the legacy network, blasting away but being blocked, the NV signal (or at least its pilot channel) is interference that clobbers the Ec/Io of the legacy tower?

It is hard for me to map the Netmonitor towers to sites on the Sponor maps since I only see one NV site in this area on the Sponsor map (BS54XC906, Broad St. and MIlk St., which I assume is BID 608). If this is the case, then tower with BID 864 does not even appear on the NV maps.

I have found that the lat/lon coordinates reported in Netmonitor can be very flaky. In my area, Netmonitor reports two towers at least a couple of miles apart with close BID numbers. And Netmonitor's log screen color-coded these sites Blue, which is supposed to mean that the towers themselves are squawking their coordinates. But I went looking for them yesterday (in a residential neighborhood) and could find no towers whatsoever near those coordinates. There is a big tower site roughly halfway between them on a north-south line, but offset a bit to the west, which does appear on the S4GRU maps.

I don't yet have an LTE-capable phone because I am keeping my powder dry for the moment. But I have been watching several forums lately, and notice an unexplained pattern:

Some end users of the EVO 4G LTE and Galaxy S III -- and I emphasize some -- report anecdotally what is typically described as "poor signal" when connecting to the legacy Sprint network They generally say that a certain location they used to get good connections and displayed "bars" on their previous handsets, but the LTE-capable models show weak signal and might even kick into roaming mode.

Of course, this leads to some angry comments that "this phone radio sucks," etc. But I am not at all convinced that the antennas or radio sensitivity on these new phones are really the root cause, but rather suspect some interaction between the handsets and the network -- which may be running side-by-side with the NV network under construction. I am wondering what might account for this. For example, in certain areas could the phones not be playing well with a developmental NV tower whose signal is being broadcast but blocked, so the phone can't lock onto the legacy tower base station that it should be using? Or even being fooled by eHRPD upgrades as Robert suggests in this case.

Sprint surely is aware of these issues, but publicly is not saying much.

Note that this is a different case from the EVO 4G LTE anomaly reported here involving faulty connections to a live LTE test site. The case I am talking about is that of LTE-capable phones with bad connections to a legacy site.

Ideas? What should such users do to isolate and troubleshoot the problem?

BTW, I was just reviewing the FCC specs of this phone, which Robert published a few weeks ago.

Do I understand correctly that the FCC certification only covers LTE in the 1900 MHz band, so when Sprint starts converting the 800 Mhz band to LTE a year or so from now, these phones can't support that functionality?

(And ditto for the HTC EVO 4G LTE ?)

One day of testing is what was reported and I wasnt connected to eHRPD. I dont know if thats the difference or not. Im going to assume it is.

There doesn't seem to be any significant difference with this eHPRD report and your earlier tests of the GS3 performance. My point is that your recollections of your old Epic 4G's performance at the same site have been all over the lot, which led to confusing comparisons.

.

EDIT: I don't mean this personally. None of us can really trust our subjective memories on stuff like this. That's why scientists do double-blind studies, etc. I know I have no reliable memory of what my phone's historical performance has been.

Home Depot.....

 

My OG EPIC was about the same.....maybe one bar better.

Okay, now you have confused me. We started this conversation with you reporting,:

I get half the bars indoors that my OG EPIC would get..

Then you did some more testing of your GS3, and said:

1bar -101dbm

2bar -97dbm

 

This is what I have been averaging all day inside work.

You did not correct my extrapolation of that:

And your old Epic 4G was typically 3 bars or better in the same location? (That would correspond to -91 dBm or better on my reference scale.)

And now you say the Epic 4G's performance "was about the same" after all? Sorry, but this seems like a wild goose chase. It goes to show that anecdotal reports of performance, especially those that rely on somebody's subjective memory, are not very reliable. There has to be some controlled testing.

Do you work in a lead lined building? That's a weak signal for being relatively close.

 

Robert via Kindle Fire using Forum Runner

See comments in this prior thread in the Samsung forum here about testing aerxx's signal performance with a GS3 in his workplace. What we don't have are any controlled comparison tests of other devices at this location.

Slightly off-topic: Will there ever be a 4G LTE Airave?

I have no personal experience to report. But I spotted this posting from the Sprint forum for the EVO 4G LTE a couple of days ago. This guy lives near Waco TX, where we know the field test has been live:

Q: What's the difference between a used-car salesman and a smartphone salesman?

A: The car salesman knows when he is lying to you.

Sorry for the slightly off topic and probably previously answered question but, what is(are) the advantage(s) of of the dual Rx antennas we're seeing in the new Sprint phones?

I frankly don't know. The serious RF engineering stuff is over my head. But it is not over the head of some veterans on this forum.