Cellular

I talked earlier about the Note 2 being the first Samsung handset I know of to include MDM9x15. In the case of the T-Mobile Note 2 I was sampled, the device includes MDM9215 which is capable of category 3 LTE FDD and TDD, Release 9 DC-HSPA+, GSM/EDGE, and TD-SCDMA along with onboard gpsOneGen 8A GNSS. I’ve confirmed that MDM9215 is present without having to disassemble or otherwise tamper with the T-Mobile Galaxy Note 2.

One of the most interesting things about the Galaxy Note 2 on T-Mobile is that it literally is the same hardware as the AT&T Note 2 with LTE. Sure, the model number is different, but the T-Mobile Note 2 includes support for LTE bands 17 (which AT&T uses) and 4 (AWS, which AT&T has specced devices out for and T-Mobile has confirmed it will deploy LTE on). This is to my knowledge the first T-Mobile handset with overt LTE support, and thus a solid future-proof purchase if you’re determined to have a T-Mobile handset that will work with the carrier’s upcoming LTE on AWS plans.

In addition, if you unlock the handset there’s no reason it shouldn’t work on AT&T’s LTE network that I can see. I managed to unlock the T-Mobile Galaxy Note 2 but not before leaving the AT&T LTE market in Dallas. It does however work as expected on WCDMA 1900 in my own market on AT&T after unlocking.

 
ServiceMode showing DC_HSPA+=1 (left), One of my fastest T-Mobile Tests (right)

As usual Samsung’s awesome ServiceMode is on the device and confirms that T-Mobile DC-HSPA+ is working. If you haven’t read discussion of DC-HSPA+ before, this is WCDMA carrier aggregation that combines two 5 MHz WCDMA downlink carriers, statistically multiplexes across them, and effectively doubles throughput on the downlink. That gets you from the theoretical maximum of 21.1 Mbps on 64QAM WCDMA up to 42 Mbps on dual carrier at the moment, which T-Mobile does run in its “4G” markets at present. I’m actually a huge fan of T-Mobile’s commitment to continually rolling out the latest physical layer upgrades for WCDMA and DC-HSPA+ keeps things feeling very speedy when you’re in good coverage and still helps at cell edge. Upstream is still limited to a single WCDMA uplink carrier, but most mobile traffic is so asymmetric to begin with it isn’t a huge deal.

Galaxy Note 2 T-Mobile - Network Support
GSM/EDGE Support 850 / 900 / 1800 / 1900 MHz
UMTS/WCDMA Support 850 / 1700 / 1900 / 2100 MHz
HSDPA/HSUPA Speeds 42.2 Mbps (DC-HSPA+ Rel9) / 5.76 Mbps
LTE Support Band 17 (700 Lower B+C), 4 (AWS), UE Cat 3 FDD-LTE (Up to 100 Mbps DL)
Baseband Hardware Qualcomm MDM9215 (confirmed) + RTR8600 (?)

Interestingly enough ServiceMode has some hints about there being possibly even more bands, but these are probably for other Galaxy Note 2 variants based on MDM9215 that will pop up or have popped up for other locales.

Running speedtests and outputting the results on the Galaxy Note 2 is more of just a sanity check than something very interesting since we’ve seen dual carrier HSPA+ before. I had to test partially in Dallas, Texas during the Big Android BBQ and partially at home. For whatever reason the conference venue definitely had some T-Mobile propagation issues or loading from all the attendees, but the averages are still decent. I’ve seen speeds around 25 Mbps down on T-Mobile dual carrier HSPA+ which is pretty impressive honestly.

 

Stats Download Throughput (Mbps)
Avg: 8.65, Max: 26.53, Min: 0.10, StDev: 5.60
 
Stats Upload Throughput (Mbps)
Avg: 1.49, Max: 3.56, Min: 0.01, StDev: 0.71
 
Stats Latency (ms)
Avg: 416.30, Max: 2394.00, Min: 57.00, StDev: 563.20

We see a weird double distribution of latency as well since there’s some additional setup and negotiation. I suspect getting out of PCH and into DCH results in some of this T-Mobile behavior when using speedtest.net, because tests without letting the radio go into IDLE have very low follow-up latency.

WiFi

The Note 2 uses BCM4334 for WiFi and is enabled for both 2.4 and 5 GHz. Just like the Galaxy S 3 that means 40 MHz channels on 5 GHz for a maximum physical layer speed of 150 Mbps. I did have some weird issues here, no matter what I did I couldn’t get the Note 2 to go on the 5 GHz AP, or get up to 150 Mbps.

Curiously enough there isn’t any 2.4 or 5 GHz priority toggle under the advanced tab in WiFi settings, yet the Galaxy S 3 models have this toggle and the same WiFi hardware stack. I suspect there’s some software or configuration issue here.

WiFi Performance - iPerf

As a result we see iPerf performance out of the Galaxy Note 2 along the lines of what you’d expect for just 2.4 GHz operation. I’m puzzled as to why this is the case quite honestly.

GNSS

There’s not really too much to say about GNSS on the Galaxy Note 2 other than that like other MDM9x15 based handsets, it leverages Qualcomm’s on-baseband GpsOneGen 8A for both GPS and Russian GLONASS based positioning. Gen 8A includes better LTE coexistence and lower power consumption compared to 8. I tested GNSS and found that locks are speedy and accurate, as expected. I suspect that the days of Samsung phones shipping with flakey GPS are now well behind us.

Display - A new Subpixel Geometry Speakerphone and Noise Suppression
POST A COMMENT

131 Comments

View All Comments

  • The0ne - Wednesday, October 24, 2012 - link

    Don't worry, I'm 40 years old myself and screen specs are very important for me due to my aging eyes. I've since replaced all my LCDs with 30" IPS ones, e-readers and tablets have at least x800 and now this phone if I decide to buy one (if pricing is right).

    I was glad to read that statement in the review as well. It definitely put a smile onto my face to actually read it and have someone share the perspective. Mind you my eyes aren't too bad being .75 off but it does make a huge difference having a good screen to look at.
    Reply
  • jjj - Wednesday, October 24, 2012 - link

    Not a fan of the S3 but for some reason i kinda like this one.
    The weight seems rather high,after all most of the time the phone sits in a pocket,hope they get rid of some layers of glass in future models.Maybe by then we also see Corning's Willow Glass and the flexible Atmel touch sensor (not controller) for a thinner bezel.
    It does feel a bit outdated already with quad Krait devices announced and dual core A15 arriving soon hopefully (Gigaom had some numbers for the A15 based Chromebook and they look impressive)
    Reply
  • enezneb - Wednesday, October 24, 2012 - link

    That 11 million contrast ratio is just amazing; a true testament to the potential of AMOLED.

    I could forsee Samsung improving their color calibration standards for the next generation of flagships seeing how they're under considerable pressure from the likes of SLCD2 and Apple's retina display. Paired with this new pentile matrix ultra-high ppi displays in the range of 400 may be possible as well (a la SLCD3)

    Looks like next year will be another exciting year for mobile display technology once again.
    Reply
  • schmitty338 - Wednesday, October 24, 2012 - link

    You can change the colour calibration to be more 'natural' in the TouchWiz software ont eh Note II.

    Also, personally I don't see the need for 'more accurate' colours on a phone. Maybe if you are a pro photog who reviews pics on their phone, but otherwise, I love the colourful pop of AMOLED displays. Even the low res pentile AMOLED of my old galaxy S (original...getting the Note 2 soon) looks great for media. Text, not so much, due to the low res and pentile.
    Reply
  • slysly - Wednesday, October 24, 2012 - link

    Why would colour accuracy ever not be important? I thought the point of a big phone is to make it easier to consume all sorts of media, from websites to photos to movies. Wouldn't more accurate colours be better for all of these activities? To me, it's a bit like saying, I don't see the need to eat delicious food during lunch, or I don't need to be with a beautiful woman on the weekdays. Why settle for something markedly inferior? Reply
  • Calista - Thursday, October 25, 2012 - link

    Is your screen properly colour calibrated? Are your walls painted in a neutral colour to avoid a colour cast? Are your lamps casting a specified type of light and shielded to avoid glare? Do you use high-quality blinds to prevent sunlight?

    This is only a few of the things to consider when dealing with calibration. And a cellphone is unable to deal with any of them unless it stays in the lab.

    So for a cellphone the criteria is:
    Is it bright enough?
    Does it look pleasing to the eye, overly saturated or not?
    Reply
  • PeteH - Thursday, October 25, 2012 - link

    Eh, depends what you do with it. I can understand wanting an accurate representation of a photo you're taking.

    And given the option between an accurate display and a less accurate display (all other things being equal) I think most people would opt for the more accurate option. Note that I'm not saying they would choose it as better visually (people seem to be suckers for over saturated displays).
    Reply
  • Zink - Wednesday, October 24, 2012 - link

    Great review. Just the right amount of detail and I really like your perspective on day to day use. Reply
  • Mbonus - Wednesday, October 24, 2012 - link

    Battery Life Question: I have noticed that since I have started using audio streaming apps my battery has taken quite a hit. I wonder if that could be added to your battery life benchmarks?

    It might not matter for devices like this where you have a large storage upgrade ability, but some other devices are forcing us to the cloud where streaming matters.

    Great and thorough review as always!
    Reply
  • Brian Klug - Wednesday, October 24, 2012 - link

    So with streaming apps and such, even though the bitrate is low, if they're not very bursty (eg download and fill a big buffer, then wait 30 seconds or minutes, then repeat) they can hold the phone in CELL_DCH on UMTS or the appropriate equivalent on other air interface types, and that's what really burns power. It's time spent in that connected state that really destroys things.

    This is actually why I do the tethering test as well (which has a streaming audio component), but I'm beholden to whether or not the review unit that I'm sampled is provisioned for tethering or not, which is the real problem.

    -Brian
    Reply

Log in

Don't have an account? Sign up now