Anandtech - "Investigating the Galaxy Nexus LTE Signal Issue"

[nkd]

It is weird Brandon. As you know my tests were probably more indicative of the results you got in your apartment (another marginal 4G signal area?). Both my Bionic and, particularly my Rezound did a better job holding on to 4G in these marginal areas.

I really don't think mine is defective as each and every VZW store I went into showed very similar performance between their display Nexus and mine. It seems that given locations are very critical in your experiences...not overly surprising.

True the 4g signal seems to get worse indoor on this phone. I can walk outside my work and I will get all full bars and around -85dbm but walk inside 4g connects at around -100dbm and only 2 bars. However I have rooted the phone and with the new leaked radios.

The 3g to 4g and 4g to 3g switch is incredibly fast. The signal strength seems the same or may be slightly better, but I do admit that the signal holds much much better. I have not lost 4g at my house for 2 days I will say it went out once or twice but quickly reconnected, after flashing the new radios. With the stock radio I would lose the 4g signal here and there, about 3-4 times a day and would take forever to reconnect. Not bad but it never happened with the razr I had. So I am liking the phone holding on to 4g so far after flashing the leaked radios.

 

[Droid800]

Smartphones aren't area controlled and it sounds really funny when you say "Rezound is drastically better in my area". They all have the comparable LTE baseband, and there are so many variables and factors that are responsible for LTE airlink performance you'll be seeing on your handset.

What I meant was that, for the terrain and topography in my area, the Rezound gets better speeds and seems to cope better. The MDM, for whatever reason, seems to be able to overcome the relatively dense housing around here, where the Nexus just stalls out.

 

[Johnly]

Running 4.0.3 and -75 16asu. 3g.

 

[Droid800]

I wonder if 4.0.3 itself has some tweaks in it. I flashed just the radios and noticed no difference.

 

[karn101]

Have any of you guys used an app like wifi analyzer to compare wifi dbm signal strength of the gnex to another phone in the same space? I'd be curious of the results....

I have. My Rezound pulls between 15 and 20 dbm better. Every now and then they would get within 5dbm of each other but the Rezound was always on top.

 

[cerktman]

As multiple people have stated earlier, it doesn?t matter what the number of bars read or what the indicated signal strength in dBm is showing.

It matters how the phone works in the real world. And in that respect it does not work as well as other phones (with regards to radio performance. Android 4.x is much better than Froyo or Gingerbread and has much more potential, in my opinion). Period. In driving the same route to work each day (64 miles round trip), the phone will lose connection to the 4G network quicker. The voice quality when switching between towers suffers quicker (there are more dropouts in voice calls than other phones I?ve used recently). In real world, everyday use, it seems the radios in this phone may not be up to par with others on Verizon?s network.

I have tried various combinations of Android 4.x and different radios. My current is 4.03 OS and the 4.03 radios. I will see how these perform today (another Walmart run is in the mix).

But we shouldn?t have to try this & try that just to make the phone usable in the real world. Real world to me is defined as driving around town, going to appointments, going to retail stores, going to work, sitting on my porch, taking a walk, doing normal everyday things that me and my family do. Not constantly sitting in one spot, not moving the phone, looking at dBm readings and running speed tests. Based on what I have encountered with the GNex, normal people (not us phone geeks) could have a hard time keeping this phone and using it as their only phone.

Now granted I?m a phone geek/nut. I love messing with them. I?m not really complaining about the GNex. I love the form factor, the way it feels, the ease of unlocking the bootloader, rooting it and loading custom roms. It?s fun. It?s a hobby. But again in the real world, I need the phone I?m carrying to do its job. And that is to make phone calls and stayed connected reliably to a data network.

The phones I currently I have for Verizon are:

Thunderbolt (rooted and running custom ROM)
RAZR (stock)
GNex (rooted and running custom ROM)
Droid Incredible 2 (stock)
Trophy (stock)
Blackberry Bold 9650 (actually two of these, one with 5.0, the other with 6.0 BBOS)

Out of these, the RAZR is hands down the best everyday performer in regards to voice calls and data. It might not ultimately be the fastest (although with fairly constant 30mbps down and 15mbps up readings on LTE, it?s up there), but voice performance is outstanding with the RAZR. The others are ok, with the Blackberries, DINC 2 and the Trophy being towards the bottom of the group with regards to 3G data performance and but fairly equal with regards to voice.

This is in no way a slam against the GNex (definitely keeping it). I?m just stating what I have encountered with the device in my neck of the woods. Not really woods, just typical suburban life here in middle TN. I?m not sure what kind of real world testing Verizon, Samsung & Google did with this phone. But in typical fashion with regards to the radios that Samsung uses, they just don?t seem up to par (and this coming from someone who gets to use phones from all major carriers, except for TMobile, in the US).

 

[Ken7]

There isn't a signal strength issue. There is a signal QUALITY issue, and also a problem with how the phone is processing that signal.

Whatever it is, whatever we call it, the bottom line is that the Nexus does not give us 4G where other phones do in marginal areas. I just hope the damage done by the Anandtech article can be undone. It's important for the parties involved to understand that this whole bars/displayed signal issue is different than the problem many are seeing in these marginal area. The article should not be used as 'cover' for a real issue.

 

[Ken7]

This is a display rezound in my local store. Got about 23 down and 12 up. Side by side, my nexus got 12 down and 4 up. So we're talking about a 50% hit in speed. Same tower, same location.

I still don't know if that alone would be enough for me to want to switch to a rezound. Sense felt so ugly after getting used to ics' beauty.
Sent from my Galaxy Nexus using Tapatalk

Was this in a solid 4G area? If it is, I don't know why you'd be getting 1/2 the speeds of the Rezound. I honestly don't recall the speeds I was getting on my Rezound.

On another note, I've always wondered what the real world impact of these speed tests are. I say that because I've often seen situations where phone A has higher speed test results than phone B, yet phone B loads web pages faster. So I often wonder what the real value of these tests are. It's kind of a joke with my son since he's seen me do these tests and it rarely tracks with actual real world results.

 

[Ken7]

I hope you ran more than one test, and I really hope you didn't run them side by side, but you actually ran one at a time.

Sent from my Galaxy Nexus using Tapatalk

What detracts from the validity when they're run side by side?

 

[Ken7]

Now that I am thinking about it, if the GN drops to 3G quicker than other LTE phones on the market, wouldn't that mean that its cutoff threshold was set too high not too low? Meaning that the GN will switch to 3G at a higher 4G signal strength than the other LTE phones?

That's what I would think. I suspect that's what he meant.

 

[Ken7]

Multiple tests. The rezounds were all around what that picture shows.

The nexus' were all around what I posted.

Sent from my Galaxy Nexus using Tapatalk

So how do we explain this? Is this difference an anomaly or more common than we think. I haven't seen people talking about this much. I may run over to a VZW store and test mine against the Rezound. Either way I'd still try to access web pages to see if it translates to real world differences.

 

[Ken7]

Rezound has an absolutely wonderful Qualcomm's MDM9600, really fine tuned, I've had it myself, but I'm seeing comparable or better throughput and latency on my galaxy Nexus under the exact same conditions.
I wouldn't go back to 2.3.4 + sense even if you paid me to do it...

OK, that's good to hear.

 

[Ken7]

What are the advertised LTE speeds we should be seeing? I'm only getting around 6-9mbps down and around 3-10 up, I thought it was maybe just my phone but I went up to VZW today and put my GNEX right next to the display one. Both had full bars and got the same low speed, around 7mbps down and 8 up. When I'm at home I only get 2 bars but get around that same speed. How is that possible? I wasn't able to compare it to the Rezound(VZW rep couldnt figure how to turn 4G on properly) but it seems like most of the 4G devices there where getting the same 4G speeds.

When doing a speedtest are you suppose to pick the server closest to you? The farther away I go, the better speeds I get. I'll click on one server close to me(Pittsburgh) and get 8/5mbps then ill click on a server in Aalabama and get 25/10mbps. How can it jump like that.

I like the GNEX and I also like the Rezound but I don't want to return my GNEX unless it's definitely a problem with the phone.

When you say all the LTE phones had about the same speeds, does that include the Nexus? If so, I wouldn't worry.

 

[milan03]

When you say all the LTE phones had about the same speeds, does that include the Nexus? If so, I wouldn't worry.

Has anyone ever seen higher speeds than what I'm getting on my nexus? I haven't, please post them.

 

[Ken7]

True the 4g signal seems to get worse indoor on this phone. I can walk outside my work and I will get all full bars and around -85dbm but walk inside 4g connects at around -100dbm and only 2 bars. However I have rooted the phone and with the new leaked radios.

The 3g to 4g and 4g to 3g switch is incredibly fast. The signal strength seems the same or may be slightly better, but I do admit that the signal holds much much better. I have not lost 4g at my house for 2 days I will say it went out once or twice but quickly reconnected, after flashing the new radios. With the stock radio I would lose the 4g signal here and there, about 3-4 times a day and would take forever to reconnect. Not bad but it never happened with the razr I had. So I am liking the phone holding on to 4g so far after flashing the leaked radios.

Sounds encouraging. What I find interesting is that your results seen to contradict the 'authenticating' theory. You said you're getting incredibly fast 3G-4G switches which I thought doesn't happen with this authenticating issue...unless I'm misunderstanding.

 

[Ken7]

As multiple people have stated earlier, it doesn’t matter what the number of bars read or what the indicated signal strength in dBm is showing.

It matters how the phone works in the real world. And in that respect it does not work as well as other phones (with regards to radio performance. Android 4.x is much better than Froyo or Gingerbread and has much more potential, in my opinion). Period. In driving the same route to work each day (64 miles round trip), the phone will lose connection to the 4G network quicker. The voice quality when switching between towers suffers quicker (there are more dropouts in voice calls than other phones I’ve used recently). In real world, everyday use, it seems the radios in this phone may not be up to par with others on Verizon’s network.

I have tried various combinations of Android 4.x and different radios. My current is 4.03 OS and the 4.03 radios. I will see how these perform today (another Walmart run is in the mix).

But we shouldn’t have to try this & try that just to make the phone usable in the real world. Real world to me is defined as driving around town, going to appointments, going to retail stores, going to work, sitting on my porch, taking a walk, doing normal everyday things that me and my family do. Not constantly sitting in one spot, not moving the phone, looking at dBm readings and running speed tests. Based on what I have encountered with the GNex, normal people (not us phone geeks) could have a hard time keeping this phone and using it as their only phone.

Now granted I’m a phone geek/nut. I love messing with them. I’m not really complaining about the GNex. I love the form factor, the way it feels, the ease of unlocking the bootloader, rooting it and loading custom roms. It’s fun. It’s a hobby. But again in the real world, I need the phone I’m carrying to do its job. And that is to make phone calls and stayed connected reliably to a data network.

The phones I currently I have for Verizon are:

Thunderbolt (rooted and running custom ROM)
RAZR (stock)
GNex (rooted and running custom ROM)
Droid Incredible 2 (stock)
Trophy (stock)
Blackberry Bold 9650 (actually two of these, one with 5.0, the other with 6.0 BBOS)

Out of these, the RAZR is hands down the best everyday performer in regards to voice calls and data. It might not ultimately be the fastest (although with fairly constant 30mbps down and 15mbps up readings on LTE, it’s up there), but voice performance is outstanding with the RAZR. The others are ok, with the Blackberries, DINC 2 and the Trophy being towards the bottom of the group with regards to 3G data performance and but fairly equal with regards to voice.

This is in no way a slam against the GNex (definitely keeping it). I’m just stating what I have encountered with the device in my neck of the woods. Not really woods, just typical suburban life here in middle TN. I’m not sure what kind of real world testing Verizon, Samsung & Google did with this phone. But in typical fashion with regards to the radios that Samsung uses, they just don’t seem up to par (and this coming from someone who gets to use phones from all major carriers, except for TMobile, in the US).

Great comparison and pretty much what I've found. But I'm surprised you're having issues with calls, I've yet to drop one yet. The sound quality is OK on calls, but I know the Motorolas will usually be better.

OTOH, the best down speed I've gotten was last night, and that was only about 18.

So my question to you is, if you had to keep one phone from your arsenal, which would it be?

 

[Ken7]

Has anyone ever seen higher speeds than what I'm getting on my nexus? I haven't, please post them.
Click to view quoted image

And that's why I wonder why 18 down is about the best I've seen.

 

[CarryMe]

And that's why I wonder why 18 down is about the best I've seen.

I believe that can be explained by the relationship between signal strength and download speeds. The weaker the signal, the slower the download speed. So if you are in a marginal area with weaker 4G signals than milan, that would explain it.

As someone else posted, the way VZW upgraded to 4G LTE was to upgrade some towers in an area to 4G LTE and leave some as only 3G capable. Eventually all will be 4G LTE, but for now, you have spotty 4G LTE coverage except in the major metro areas. The closer you are to the 4G towers, the stronger the signal and the faster the download speeds.

 

[Ken7]

CarryMe, actually the highest speed I got was last night in Manhattan. I had a full set of bars and a SS of about 86dbm. So I would think that's high enough to do better than what I saw, but I'll check mine against another Nexus today to makes sure it's not my unit.

I also think the question that was asked about which server you connect to for these Speed Tests is an interesting one. If there's such a variability between the servers, it really questions how we can be benchmarking one guy's speed against another.

 

[milan03]

CarryMe, actually the highest speed I got was last night in Manhattan. I had a full set of bars and a SS of about 86dbm. So I would think that's high enough to do better than what I saw, but I'll check mine against another Nexus today to makes sure it's not my unit.

I also think the question that was asked about which server you connect to for these Speed Tests is an interesting one. If there's such a variability between the servers, it really questions how we can be benchmarking one guy's speed against another.

Plus Manhattan is a very large island. Would have to be a very specific location.

Sent from my Galaxy Nexus using Tapatalk