Marcelo Claure, Town Hall Meetings, New Family Share Pack Plan, Unlimited Individual Plan, Discussion Thread

[burnout8488]

B26 has had the same range/penetration as B25 in my area. They've had B26 broadcasting for around 8 months now but they still haven't optimized it. I know they're short on technicians, but can't they just temporarily hire another carriers or something?

 

Sent from my M8

It took about 9 months in my area, FWIW. Hopefully your market is right around the corner.

[BlueAngel]

That experience suggests a device issue.  Dropping LTE at -105 dBm RSRP is not common.  And optimization should not factor in to the equation.  Before optimization, the signal may be at lower power.  But that just means the cell radius potentially will be smaller, not that LTE will drop at a higher signal level.

 

Lastly, if you are running an alternative ROM, all bets are off.  Many of those developers do not know what they are doing when it comes to RF engineering.  They can muck things up with the baseband firmware.

 

AJ

Happened on my Note 4 as well it's not a device issue.

[WiWavelength]

The baseband firmware is completely proprietary and cannot be modified by a ROM.

 

I am not suggesting that developers are modifying baseband firmware.  But they sure can alter, add, or remove parts of stock firmware that could then unintentionally affect RF performance.

 

I stay away from alternative ROMs like the plague.  They are completely unnecessary.  And maybe my recollection is faulty, but I seem to recall reports of alternative ROMs sometimes causing connectivity issues on certain devices on certain networks.

 

If you have experience that no alternative ROM has ever affected any RF performance parameter, please offer it.  That is not asking you to prove a negative, just to say honestly that you have not encountered that nor heard of it yourself.

 

AJ

[dnwk]

I might as well be, 800 voice is solid but LTE on 800 is a different story. So many places I drop to 3G when I could be on LTE 800 but it drops to 3G at only 105dBm. They need to optimize this soon it's only been almost a year now I don't know what is taking so long.

 

Sent from my Nexus 5

 

It appears to me that Sprint deploy B41 faster than B26. Is it because there are many more wireless carriers use B41 than B26?

[EvanA]

I am not suggesting that developers are modifying baseband firmware.  But they sure can alter, add, or remove parts of stock firmware that could then unintentionally affect RF performance.

 

I stay away from alternative ROMs like the plague.  They are completely unnecessary.  And maybe my recollection is faulty, but I seem to recall reports of alternative ROMs sometimes causing connectivity issues on certain devices on certain networks.

 

If you have experience that no alternative ROM has ever affected any RF performance parameter, please offer it.  That is not asking you to prove a negative, just to say honestly that you have not encountered that nor heard of it yourself.

 

AJ

Speaking as a developer for Paranoid Android (a major custom ROM distribution), that is entirely impossible.  All RF considerations are dealt with in the modem firmware.  There is no way to modify RF performance by modifying AOSP.  

[WiWavelength]

Happened on my Note 4 as well it's not a device issue.

 

It still could be a device issue.  You could have gotten two subpar devices in a row.  Or you could have installed two problematic ROMs in a row.  Either way, several others with stock devices need to corroborate your experiences.  As factually stated, dropping LTE at -105 dBm RSRP is not common.  If that happens for others, then the network in that area is somehow fakakta.

 

AJ

[WiWavelength]

Speaking as a developer for Paranoid Android (a major custom ROM distribution), that is entirely impossible.  All RF considerations are dealt with in the modem firmware.  There is no way to modify RF performance by modifying AOSP.  

 

Does a third party ROM package then typically or even necessarily include a version of the stock baseband firmware?

 

The reason I ask is this:  Nexus 5.  It is a quintessential example, having gone through numerous baseband firmware revisions from Qualcomm.  Some revisions have performed better on Sprint than have others.  And the shoe may be on the other foot on other operator's networks.  If a stock baseband firmware revision is included in a third party ROM, who is to decide which one?

 

Maybe I am off base here, but could that not affect RF performance?

 

AJ

[EvanA]

 

 

Does a third party ROM package then typically or even necessarily include a version of the stock baseband firmware?

 

The reason I ask is this: Nexus 5. It is a quintessential example, having gone through numerous baseband firmware revisions from Qualcomm. Some revisions have performed better on Sprint than have others. And the shoe may be on the other foot on other operator's networks. If a stock baseband firmware revision is included in a third party ROM, who is to decide which one?

 

Maybe I am off base here, but could that not affect RF performance?

 

AJ

Custom ROMs typically do not include baseband firmware. It's up to the user to manually flash that themselves. Performance differences from baseband to baseband are typically minor, with the exception of the Nexus 5 when it went from non-Spark to Spark enabled. Since that update the version jumps have been minor, suggesting only minor differences between the basebands.

[BlueAngel]

It still could be a device issue.  You could have gotten two subpar devices in a row.  Or you could have installed two problematic ROMs in a row.  Either way, several others with stock devices need to corroborate your experiences.  As factually stated, dropping LTE at -105 dBm RSRP is not common.  If that happens for others, then the network in that area is somehow fakakta.

 

AJ

While it's possible I highly doubt it. It's true for my friend as well we both connect to it and then drop it in favor of 3G. I only use the stock rom sometimes rooted and other times not.

[Flompholph]

I have been kicked off lte with a signal as low as -95. It was because the tower was overloaded when they turned on band 41 everywhere around that area. It now drops lte around -125.

 

To get the highest throughput for the sector the tower should keep the UEs with the highest snr and highest cqi. High snr and high cqi means higher bps the tower can push to a device. That is why you get slower top speeds as you go away from the tower.

 

If they didn't do that the towers max throughput would be cut by a fourth.

[Ascertion]

It took about 9 months in my area, FWIW. Hopefully your market is right around the corner.

How much of a difference did it make once it was optimized?  Is it comparable to EvDo signal (ie: -80 dBm Evdo like -94 dBm LTE B26?)

[S4GRU]

How much of a difference did it make once it was optimized?  Is it comparable to EvDo signal (ie: -80 dBm Evdo like -94 dBm LTE B26?)

 

Post optimization typically yields around a 10dBm improvement on B26 from the same site, according to what most people around the forums report.  But it can be quite a bit more.  And sometimes it doesn't change at all.

 

For instance, the downtilt between the 800 and 1900 can be as much as 10 degrees different.  So if the 1900 is set 10 degrees lower than the 800 and you are at edge of service for 1900, it wouldn't be too far fetched that you could have -120dBm on B25 and -88dBm on B26.  Whereas, the difference could be only slight if you are close to the tower as the B26 would be going more over your head, and the B25 pointed more down toward you.

 

There is going to be huge variability, from sector to sector and site to site between bands.  B26 can be designed to reach completely different set of customers than B25 on the same sector.  So, between difference in downtilit and power broadcasted, the difference can be huge.  Or, it may be very little.

 

In general, it is more likely to be much higher of a difference in rural areas where they are trying to get the signal farther, and more likely to be less in urban areas where they already have signal overlap and trying to focus more on indoor signal penetration.

[burnout8488]

To go along with what Robert said, I experience an 8-10 dBm difference indoors, typically, compared to whatever B25 I am able to get in the same spot.

[Ascertion]

Post optimization typically yields around a 10dBm improvement on B26 from the same site, according to what most people around the forums report.  But it can be quite a bit more.  And sometimes it doesn't change at all.

 

For instance, the downtilt between the 800 and 1900 can be as much as 10 degrees different.  So if the 1900 is set 10 degrees lower than the 800 and you are at edge of service for 1900, it wouldn't be too far fetched that you could have -120dBm on B25 and -88dBm on B26.  Whereas, the difference could be only slight closer to the tower as the B26 would be going more over your head.

 

There is going to be huge variability, from sector to sector and band to band.  B26 can be designed to reach completely different set of customers than B25 on the same sector.  So, between difference in downtilit and power broadcasted, the difference can be huge.  Or, it may be very little.

 

In general, it is more likely to be much higher of a difference in rural areas where they are trying to get the signal farther, and more likely to be less in urban areas where they already have signal overlap and trying to focus more on indoor signal penetration.

This is actually interesting.  Thanks for the info.  I'll report my experience when my area gets optimized.

I'm just hoping, for the sake of using internet radio in my vehicle on  my commutes, that 4G does not drop.  Sprint already covers ~90% of my commute in 4G LTE, but once it drops down to 3G, it won't scan for 4G since it'll have an active data session ongoing (internet radio.)

[jreuschl]

It took about 9 months in my area, FWIW. Hopefully your market is right around the corner.

9 months from what.. Spark launch?

[Ascertion]

9 months from what.. Spark launch?

To optimize B26 in his area.  He was giving me the timeline it took Sprint for his area.  It'll likely be comparable to mine, since my area has been un-optimized for ~8 months on B26.

[burnout8488]

9 months from what.. Spark launch?

Since the day the Band 26 carrier card was installed and turned on. :-)

[gusherb]

It's been over a year since band 26 started getting turned on around here. Sometimes the network acts like a turd, and sometimes it surprises me and acts like its optimized. I can't stress enough how much I wish it were optimized already. Today I was stuck on B41 that was anywhere from 1 Mbps to page timeouts and nothing working at all. (Had to cycle airplane mode so it would attach to B26)

 

I would hope after optimization a situation like I just mentioned above wouldn't happen.

[JosefTor]

It's been over a year since band 26 started getting turned on around here. Sometimes the network acts like a turd, and sometimes it surprises me and acts like its optimized. I can't stress enough how much I wish it were optimized already. Today I was stuck on B41 that was anywhere from 1 Mbps to page timeouts and nothing working at all. (Had to cycle airplane mode so it would attach to B26)

 

I would hope after optimization a situation like I just mentioned above wouldn't happen.

I thought optimizing was primarily about optimizing range and less so speed. If you are getting such a low throughput, I would say it has too much range or not enough brandwith.

[lilotimz]

It's been over a year since band 26 started getting turned on around here. Sometimes the network acts like a turd, and sometimes it surprises me and acts like its optimized. I can't stress enough how much I wish it were optimized already. Today I was stuck on B41 that was anywhere from 1 Mbps to page timeouts and nothing working at all. (Had to cycle airplane mode so it would attach to B26)

 

I would hope after optimization a situation like I just mentioned above wouldn't happen.

 

I thought optimizing was primarily about optimizing range and less so speed. If you are getting such a low throughput, I would say it has too much range or not enough brandwith.

 

Optimization is coverage and consistency of the signal.

 

You cannot optimize congestion. 

[gusherb]

Optimization is coverage and consistency of the signal.

 

You cannot optimize congestion.

 

I thought optimization also included getting the load balancing working right. Either way the network is pushing everything it possibly can to B41 when it could be spreading it out more evenly.

I could've sworn everything i read about optimization also included load balancing...

[JossMan]

To optimize B26 in his area.  He was giving me the timeline it took Sprint for his area.  It'll likely be comparable to mine, since my area has been un-optimized for ~8 months on B26.

Every situation will be different but usually when LTE 800 is initially turned on it yields the same signal strength as LTE 1900, maybe a couple dBm better.  Post optimization it will vary, the site nearest my home use to have a -105dBm connection indoors on 800 now it hovers around -90dBm indoors, again this was my experience and about 90% of the sites in this area are running dual-band 800/1900 LTE now.  

[greenbastard]

Optimization is coverage and consistency of the signal.

 

You cannot optimize congestion.

Well, you could if you live near a tower like mine in which every spark phone is seeded to Band 26 automatically for some odd reason. During the midday, download speeds can top at 0.15 Mbps on band 26 while band 25 sits at 8+ Mbps. There's nothing I can do about this since its a network bug. I usually turn off B26 on my phone or force 3G (1+ Mbps) in order to make my data experience tolerable.

[lilotimz]

I thought optimization also included getting the load balancing working right. Either way the network is pushing everything it possibly can to B41 when it could be spreading it out more evenly.

I could've sworn everything i read about optimization also included load balancing...

 

Optimizing the network ensures that all three bands cover what they're meant to do with one band handing off to each other as coverage on one band degrades or improves. 

 

Also that is how the network is meant to be configured. B41 is now / or going to be set as priority band in all cases and all devices will hand off to band 41 where avaiailable. 

 

Existing load balancing parameters as part of the old / current software sets are quite inconsistent. The next software update to the RAN equipment that is/was being pushed out for all the vendors further expands that load balancing schemes to handle congestion better but as some have experienced in several initial markets that they pushed it out to it's not perfect still even after months of FIT testing.   

[nexgencpu]

Optimizing the network ensures that all three bands cover what they're meant to do with one band handing off to each other as coverage on one band degrades or improves. 

 

Also that is how the network is meant to be configured. B41 is now / or going to be set as priority band in all cases and all devices will hand off to band 41 where avaiailable. 

 

Existing load balancing parameters as part of the old / current software sets are quite inconsistent. The next software update to the RAN equipment that is/was being pushed out for all the vendors further expands that load balancing schemes to handle congestion better but as some have experienced in several initial markets that they pushed it out to it's not perfect still even after months of FIT testing.   

The only reason I can think that Sprint is removing triband users from B25 is allow single band users to still have a decent experience since B26/B41 should be sufficient in most cases. 

 

But in reality its cramming a lot of users onto B26 and slowing it down considerably, while B25 cruises along. Must be a very high percentage of tri band users already out there.