Qualcomm GNSS chips with L1 L2 L5

[lodrog...@gmail.com]

we have the first android smartphone with qualcomm GNSS chips that have triple band and 4 frequencies. This is the the Realme GT5

https://www.realme.com/realme-gt-5/specs

https://i.postimg.cc/XqpmMhNg/Clipboard01.jpg

Snapdragon 8+ Gen 2

https://i.postimg.cc/qMsFKP8q/Clipboard02.jpg

other chips

https://i.postimg.cc/QNnBDvD6/Clipboard03.jpg

https://i.postimg.cc/1XT4KVMx/Clipboard02.jpg

https://i.postimg.cc/7hQJQWHW/Clipboard01.jpg

[lodrog...@gmail.com]

there is no Snapdragon 8+ Gen 2

with a plus (+) on the end. this is type error when i was writing

[Robert Enger]

I note the Realme spec calls for up to 13.7 amps to be delivered across the USB type-c connector (and cable).  (normally USB-C current rate is limited to 5amps; although I found an advert from a connector company that says they can supply a 9amp jack on special order).

The Realme spec says it supports 240w charge rate, but does not support the (new) standard 48 volt operating mode.

It seems none of the standards compliant USB PD chargers will charge this phone at full rate (perhaps not even at 100w rate?).

The enhanced GNSS system is very welcome (especially is Realme have done a good job on the antennas and RF system, and if the FCC doesn't require it to be downgraded or disabled).

But Realme seems to be quite divergent on USB power delivery.  

--
You received this message because you are subscribed to the Google Groups "GPSTest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gpstest_andro...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/5987370e-d580-4522-a0af-cc5e74eae9bbn%40googlegroups.com.

[Sean Barbeau]

Interesting, thanks for sharing!

--

[ndo...@gmail.com]

Thanks for sharing.  Glad it also has SBAS support.  Wonder if it's locked to only those service providers (via provider ID in MT17 broadcast by SBAS), or if new systems will work as well as they come online (e.g. Australia's SPAN, etc.).

The VEPP support looks interesting.  Looks like you can connect in external sensors as well, perhaps via bluetooth, USB hub, other?  Inertial sensor, Vehicle sensors, and vision for object recognition.  Competing with the likes of SwiftNav, Hexagon, Trimble, and others.  The market for precision navigation for automated vehicles is growing by leaps and bounds.

From the catalog, denoting the asterisks shown in the image linked above :

*Satellite Based Augmentation Systems – Provides DGPS Corrections & Satellite Integrity monitoring 

**Vision Enhanced Precise Positioning supported on select external Snapdragon Processors using Qualcomm® Multi Frequency GNSS, Vehicle & IMU Sensor information, and access to either a front- or rear-facing camera feed  

[binghy17]

Hi all!
That is interesting! Does it support SBAS too hence?
As far as you know, is there any other smartphone available on the market with L2C and (eventually) SBAS support? I've tried searching on the GPSTest Database but I found nothing.

[lodrog...@gmail.com]

about the question for SBAS support to this new qualcomm GNSS engine with internal name "iZat" the anwer is YES. From year 2010 with qualcomm GNSS engine gen8 with first support of dual systems of GPS + Glonass, there was SBAS also. Unfortunately qualcomm company doesn't give detail information to end users.

there is also one more problem. Over 80% of smartphones manufacturer has disabled a lot of iZat capabilities for some reasons. For battery economy, for lower hardware cost of antenna (dual band), etc

don't let those advertisements for triple band, multi constellation systems and SBAS to fool you for cm accuracy. The qualcomm company has limit down at 3 meters the gnss iZat engine, with US government command for civilian low cost devices. Only professional high cost devices allowed to have down to 30-50 cm static accuracy.
The only profit with this new technology for the smartphones is to have "driving lane accuracy 3-5 meters" at urban canyon, compared with gnss receivers of only L1 band that have 20-30 meters when driving with a car

[Jim bell]

I am interested in its possible support for Galileo HAS (High Accuracy Service) corrections, including its ability to receive the E6 signal that contains the corrections.  Or, alternatively or in addition, the IDD (Internet Data Distribution,) source of HAS corrections. 

--

[Jim bell]

Is there any more recent word about this Realme GT5 and its triple band GPS, including L2?

[Jim bell]

I found this:

https://a.co/d/aYM4v3y

[Guillaume]

That's the RealMe GT 5G not the GT5 unfortunately.

--

You received this message because you are subscribed to the Google Groups "GPSTest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gpstest_andro...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/07fda736-ab1e-4f1c-a6f6-922f356f0467n%40googlegroups.com.

[gpsfan]

There are quite a few YT reviews of the GT5 https://www.youtube.com/results?search_query=realme+gt5

Haven't looked at them yet but maybe one of them does look into GPS aspects ?

Has it made its way into the database ?

[Invalid Data]

The newer Realme GT5 Pro with Qualcomm 8 Gen 3 lacks support for L2 E5b B2b.

Apparently L2 support requires extra RF design cost and they decided not to do so in newer models.

But E5b and B2b support is suspicious. Qualcomm never officially pointed out their chips support the  "b signal". Then  Realme said that with "Qualcomm enhanced SDK", GT5 is granted with E5b B2b support.

If so, the "Enhance SDK" might be something have to pay for.

Or something so hacky like software patches, that it was difficult to implement.

Which explains why not a singel SD 8 Gen 3 phones has E5b support.

At the meantime, Vivo and Xiaomi's mediatek model get full E5b/B2b support. Even the relatively low end K60 Ultra series has it. 

I know that mediatek is well known for packing the whole software/hardware solution to device manufacturer. That's probably why every 9200/9300 phone has proper GNSS support 

Device like high-end Xiaomi 14 series also has only dual frequency L1+L5. So probably dual band is sufficient for many daily life scenes now. 

The triple/quad band GNSS is more like a placebo pills. 

gpsfan <web...@gmail.com> 于 2023年12月22日周五 下午9:17写道：

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/28094eeb-47a6-4477-8552-dff53ac96262n%40googlegroups.com.

[Invalid Data]

Based on the current smartphone industry, there's only 2 phone support L2. They're Nubia Z50 Ultra and Realme GT5. 

The dedicated L2 antenna must cost more. Until the technology evolves to some degree. But L2 is very robust signal, it can endure city valley, electromagnetic interference and signal interruption, well that's in theory at least. 

And then E5b and B2b come. These signals share antenna with L5, which means the cost added is zero.

If the chipset has it, and the antenna design is dual band based, then the phone will be L1+L5+E5b triple band naturally. 

And then when the chipset can parsing so many signal in 100ms, up to 120 signal at the same time (20 GPS/8 GLO/30 GAL/60 BDS/6 QZSS/4 Navic), also combining with IMU sensor navigation and carrier phase tracking. 

I would say its position calculation ability is very outstanding if not the best. 

So both GT5 and any D9200/9300 phone will be the best accuracy phone for obsession people like me. I use GPS altitude to log my trips.

For average use, any dual band phone is great actually 

Invalid Data <weiti...@gmail.com> 于 2023年12月24日周日 上午1:22写道：

[Invalid Data]

Another information: Realme GT5 user reported 3 weeks ago the test shows l2 is currently not available in this software version. Also the E5b is absent too.

That explains why Qualcomm never officially specify the GNSS support list. Because they don't. 

Invalid Data <weiti...@gmail.com> 于 2023年12月24日周日 上午11:31写道：

[gpsfan]

Interesting, where was that discussed ? Could it be a limitation of the GPSTest software maybe that's not "ready" to show L2 or E5b signals ?

[gpsfan]

I think I read somewhere that receiving E6 signals would require a dedicated antenna ? And quality antennas is always going to be a problem on smartphones that are packed with all kinds of RF interfering technologies...

[Jim bell]

Here is the Galileo signal plan:

Signal	Carrier Frequency (MHz)	Receiver Reference Bandwidth (MHz)
E1	1575.420	24.552
E6	1278.750	40.920
E5	1191.795	51.150
E5a	1176.450	20.460
E5b	1207.140	20.460

https://gssc.esa.int/navipedia/index.php/Galileo_Signal_Plan

You received this message because you are subscribed to a topic in the Google Groups "GPSTest" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/gpstest_android/IcyIxQoCGwc/unsubscribe.
To unsubscribe from this group and all its topics, send an email to gpstest_andro...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/7336bdb5-a83c-4074-8c9e-56f84d882035n%40googlegroups.com.

[Invalid Data]

I think GPStest can show L2 (1227.60 MHz).

Realme said it needs a software update to activate L2 support and right now it's not available. 

So Qualcomm probably never intended to improve smartphone accuracy in favor of selling their own Auto chipset.

If you want good accuracy, it's wiser to buy a Mediatek chip phone. It has carrier phase tracking, E5b band, and PPP positioning. 

Jim bell <jimd...@gmail.com> 于 2023年12月27日周三 上午1:51写道：

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/CAMvzYpie0JE-%2BnGu%2BjWM3EWSRkYj4LPA%3Drug%2Brt_Q7iK94%2BdhA%40mail.gmail.com.

[gpsfan]

Thanks for the info, hopefully that OTA update will happen at some point !

Which current Mediatek smartphones have E5b and PPP implemented ? Is PPP done via an app that's provided by Mediatek ? The only Android app I'm aware of that used to provide PPP was 3GPO and that doesn't seem to work anymore, at least on my P6 and the app looks discontinued since the Google Maps don't show anymore.

[binghy17]

Hi,

are there still no entries on the GPSTest database about Realme GT5 and Nubia Z50 Ultra ( GPSTest Database - Fogli Google)?

I've seen some entries about Realme GT 5G, but I guess that one is another model, since the date reported is around 2021/2022.

[binghy17]

Agree totally. But it depends also on the smartphone manufacturer side, and how all these things will be assembled together to minimize interference.

I also add that, over the classical GNSS algorithms to be applied on GNSS and IMU raw measurements (to be coupled for), more powerful chipset like Dimensity 9200/9300 will aid the possibility to introduce AI algorithms too. A lot of work and research has to be done.

Il giorno domenica 24 dicembre 2023 alle 04:31:30 UTC+1 weiti...@gmail.com ha scritto:

[binghy17]

It seems there will be several smartphones, as announced on 11/23/2023:  List of Smartphones Confirmed to Feature MediaTek Dimensity 9300 SoC (mysmartprice.com)

1) Vivo X100 Series 5G seems to be launched with this chipset, as announced on 01/03/2024:  vivo X100 Series 5G, powered by MediaTek Dimensity 9300 | MediaTek (en)

2) Oppo Find X7 seems to be launched with this chipset, as announced on 01/08/2024:  Oppo Find X7 comes with Dimensity 9300, only one telephoto camera - GSMArena.com news

As found on Mediatek D9300 page, it seems there will be (https://www.mediatek.com/products/smartphones-2/mediatek-dimensity-9300):

- New L1C/B1C frequency support
- Full global frequency coverage: L1+L5+B2B+E5B with PPP for high precision positioning

So hence no L2 support, as claimed by Qualcomm with Snapdragon 8 Gen 3.

Just a question aside, in general. Since it seems the battle shall be between MediaTek Dimensity 9300 and Qualcomm Snapdragon 8 Gen 3, looking for smartphones with the latter, it seems there is no real support for L2, as instead claimed on the official Qualcomm page:

Qualcomm Snapdragon 8 Gen 3: https://www.qualcomm.com/products/mobile/snapdragon/smartphones/snapdragon-8-series-mobile-platforms/snapdragon-8-gen-3-mobile-platform#Overview

Smartphones:

1) OnePlus 12

- Released: 12/11/2023

- Chipset: Qualcomm SM8650-AB Snapdragon 8 Gen 3 (4 nm)
- Positioning: GPS (L1+L5), GLONASS (G1), BDS (B1I+B1c+B2a), GALILEO (E1+E5a), QZSS (L1+L5)

- Info: https://www.gsmarena.com/oneplus_12-12725.php

2) Realme GT5 Pro

- Released: 12/14/2023
- Chipset: Qualcomm SM8650-AB Snapdragon 8 Gen 3 (4 nm)
- Positioning: GPS (L1+L5), GLONASS, BDS (B1I+B1c+B2a), GALILEO (E1+E5a), QZSS (L1+L5), NavIC

- Info: https://www.gsmarena.com/realme_gt5_pro-12646.php

Which is the reality? L2 support on Snapdragon 8 Gen 3: yes/no?

I've checked on the GPSTest Database and there are still no entries obviously, since the smartphones have been sold just recently. Anyone here with one of those two smartphones?

[Invalid Data]

Just for you information, all phones with 8 Gen 2/3 that whether specified L2 support or not, haven't report L2 in real GPStest screenshot.

And if my knowledge is correct, L2 receiver is merely accurate because it used to be the only dual band combination to calculate ionosphere correction, before L5 operated in 2014. After 2014, it's been replaced by superior L5/E5a/B2a. 

The reason is, according to GNSS theory, L2 is optimize for signal stability, it has very long code length and high power gain, which means very high robustness against almost any interference. But also it has very low 0.5115MHz chip rate (or code frequency) which means by standalone, L2 is poor in accuracy. And I have read a paper about the standalone accuracy for L1 L2 L5, the results match the theory (L5>L1>L2)

In comparison, L1 chip rate is 1.023MHz. L5/E5a/B2a chip rate is 10.23MHz. Which means the pseudo-range information is 20 times denser than L2. That's the reason modern GNSS signal is more accurate. 

Also for ionosphere correction, since L5 frequency is more distantly away from L1, also in theory, the correction results will be better ^ ^

Given the above evidence, it's not that important to have L2 when it has to trade off other features like carrier phase, PPP positioning, IMU sensor algorithm 

binghy17 <m.gian...@gmail.com> 于 2024年1月9日周二 下午6:08写道：

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/3fa197af-2391-4049-82ce-50c03a1354c9n%40googlegroups.com.

[Jim bell]

I forgot to send this to everybody.

That sounds very interesting.  Apparent the old survey-grade GPS units just use L1 and L2, as does the SparkFun ZED-F9P I just bought a few months ago.   (And which I have been very satisfied with.)

But if I understand your material, survey-grade units would be better if they use L1 and L5.  So, is that beginning to happen?  

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/CAFPPBd9%2BY6DMdVBSND%3Dd3xPBq6ZOKwCW%2BcwkCMr7e48-rVg%3Dmg%40mail.gmail.com.

[Invalid Data]

Hello buddy it's glad to hear you find a satisfying device!

The biggest user error in GNSS is ionosphere delay, so both types of dual band device already eliminates the same  error.

While L1+L5 has lower pseudo range noise, so it's more suitable for high accuracy measurement.  

L1+L2 is suitable for good accuracy. Its lower bandwidth and chip rate consume less battery, so very helpful in outdoor survey. 

Although L5 is very accurate, it isn't the bottleneck of achieving actual accuracy. 

Alongside the ionosphere effect, there's smaller but important error like: satellite clock error, orbit ephemeris inaccuracy, signal group delay, antenna geometry, which requires other technique in order to be resolved, like PPP/RTK/SBAS 

And some method like carrier phase measurement will improve tracking accuracy too. 

So in general both dual frequency combination will be great as 99.9% of ionosphere effect is eliminated in the calculation. 

Jim bell <jimd...@gmail.com> 于 2024年1月16日周二 下午1:00写道：

[Jim bell]

Quite fascinating.  

And, I'd sure like to see a smartphone installed with Galileo HAS IDD, the IDD substituting for E6 SIS service.

I think Galileo made a mistake to define a different frequency to deliver SBAS corrections.

Yes, I believe I understand that for a smartphone to actually receive E6, an antenna approximately resonant at that frequency is necessary.  Who would take the trouble to design and install that?!?!  At least, not in a smartphone form-factor.  Maybe in a survey-grade receiver?

But what I don't see is any excuse for the apparent failure of the Galileo people to provide the data necessary, to smartphone designers, to install IDD.   EXCEPT...

I notice from the smartphone gnss database that virtually no gnss-chip generations are after '2020'.  I'd like to see an explanation.  Did the GNSS chip designers just die due to COVID?

https://docs.google.com/spreadsheets/d/1jXtRCoEnnFNWj6_oFlVWflsf-b0jkfZpyhN-BXsv7uo/edit?pli=1#gid=0

I think this amounts to a built-up demand.  When will the next wave of GNSS appear, like a dam breaking. (I hope!)

[Invalid Data]

No these designer didn't die. 

Jim bell <jimd...@gmail.com> 于 2024年1月19日周五 上午5:40写道：

[Invalid Data]

See the difference from these two different chip. Soc MT6985 also branding as Dimensity 9200, is using a new dedicated GNSS chip MT6866 that is separately operating from Bluetooth WiFi.

While the older and mid-class MT6873 soc (Dimensity 800) use a single integrated RF chip MT6635 to provide GNSS, Bluetooth, WiFi features.

Although I haven't found the schematic for Dimensity 9300, but they do add GPS L1C update to the GNSS chip. Which is the latest technology they shared from  AG3335 ( a Mediatek GNSS chip too)

(It's integrated BTW, so it can't be inspected it by disassembling the phone.)

Also Qualcomm add some "AI" enhanced GNSS improvement to there location stack , I'm not sure if it's related to chip design changes. But anyway things are gradually improving. 

The Android hardware year is reported through a API, so it's software defined. It won't reflect the actual chip design 

Invalid Data <weiti...@gmail.com> 于 2024年2月1日周四 上午8:34写道：

[Invalid Data]

Typo: the dedicated GNSS chip in D9200 is MT6686, not MT6866.

its specification is never published. All we have is its name, so never mind. 

Invalid Data <weiti...@gmail.com> 于 2024年2月1日周四 上午8:56写道：

[Invalid Data]

Looking at Nubia Z60 Ultra, it actually reports L2 with GPS test database 

Invalid Data <weiti...@gmail.com> 于 2024年2月1日周四 上午9:05写道：