Official Tmobile-Sprint merger discussion thread

[iansltx]

Last night I forced B25 downtown and got a T-Mobile carrier using only the G block, rather than the typical Sprint 10x10 carrier that showed up when I went a few blocks north (this was downtown Austin, 1st St bridge). Center frequency was offset 100 KHz vs. the usual G-only Sprint B25 installs. Got like 10 Mbps down, 1 Mbps up in the middle of a fireworks show when AT&T was basically dead, VZW was iffy, and deprio'd T-Mobile was hit or miss (I did too many speed tests and hit the 50 GB soft cap last night).

[Grabber5.0]

Ugh, my S21 is suddenly parked on B12 sitting in the front room of my house. I bet they turned off the Sprint B41 at my closest tower to start moving it over. Will have to play with the band selector to see if it's really offline.

[Grabber5.0]

I think I was right (or I know less about TNX band preference than I thought). When I disabled everything except B41, I connected to a weak T-Mobile B41 signal from (I assume) a different tower. With all 5G enabled and just LTE B41, I am sitting on N71, which still shows the 5GUC icon for some reason, but at least it's faster than B12. I assume they're going to move that B41 to N41? Hopefully sooner than later.

[iansltx]

1 hour ago, Grabber5.0 said:

I think I was right (or I know less about TNX band preference than I thought). When I disabled everything except B41, I connected to a weak T-Mobile B41 signal from (I assume) a different tower. With all 5G enabled and just LTE B41, I am sitting on N71, which still shows the 5GUC icon for some reason, but at least it's faster than B12. I assume they're going to move that B41 to N41? Hopefully sooner than later.

B41 goes further than n41, so my bet is that the site is broadcasting n41, just not loudly enough for your phone to hear it.

[chamb]

1 hour ago, iansltx said:

B41 goes further than n41, so my bet is that the site is broadcasting n41, just not loudly enough for your phone to hear it.

I THINK the above statement may be true, but somebody please verify it.  And if you can verify it, please tell me why?

[dkyeager]

1 hour ago, iansltx said:

B41 goes further than n41, so my bet is that the site is broadcasting n41, just not loudly enough for your phone to hear it.

I would think this would be determined by the amount of bandwidth.  My hypothesis would be in a rural area with 20+20 of b41 would go further than 10Mhz of n41, while in a more endowed urban area 100Mhz of n41 would go further than 20+20 of b41. Larger bandwidth can more easily overcome interference.

 

Another hypothesis would be the age and capabilities of the site equipment

[iansltx]

In terms of X goes further than Y, I've seen this IRL on sites around here. Same equipment broadcasting both n41 and B41, with B41 usually at a higher frequency than n41. LTE is a less complex signal than NR, and chipsets are more mature, so it just goes further before dropping out. When the closest n41 site to me was ~1 mi away I could pull B41 from that site consistently, but to get anything near reliable n41 I'd have to get within 0.7 mi or so. It's less obvious at this point as n41 coverage is closer to contiguous, but I know for a fact that I was looking at B41 and n41 from the same site, same panels (though IIRC LTE gets broadcast from different M-MIMO antenna elements), and these are the perf characteristics I've seen.

Note that the interference robustness issue shouldn't ever come into play as we're talking about licensed spectrum. Attenuation, sure, but fancier modulations require higher SNR to make work.

Likewise, VZW n2 doesn't reach nearly as far as B2 here, though that may be a function of VZW not trying very hard to keep phones on DSS rather than plain LTE.

[dkyeager]

15 hours ago, iansltx said:

In terms of X goes further than Y, I've seen this IRL on sites around here. Same equipment broadcasting both n41 and B41, with B41 usually at a higher frequency than n41. LTE is a less complex signal than NR, and chipsets are more mature, so it just goes further before dropping out. When the closest n41 site to me was ~1 mi away I could pull B41 from that site consistently, but to get anything near reliable n41 I'd have to get within 0.7 mi or so. It's less obvious at this point as n41 coverage is closer to contiguous, but I know for a fact that I was looking at B41 and n41 from the same site, same panels (though IIRC LTE gets broadcast from different M-MIMO antenna elements), and these are the perf characteristics I've seen.

Note that the interference robustness issue shouldn't ever come into play as we're talking about licensed spectrum. Attenuation, sure, but fancier modulations require higher SNR to make work.

Likewise, VZW n2 doesn't reach nearly as far as B2 here, though that may be a function of VZW not trying very hard to keep phones on DSS rather than plain LTE.

Agree that we must discount Verizon. 

What you're observing could also be a design issue of 5g: the complexity of being able to serve so many more devices where signal is strong leads to edge deterioration. I agree that this might be able to be addressed in the future.

I would like you to discount the bandwidth size theory if you can. What is the bandwidth difference in places where you have observed n41 falling short of b41?

[dkyeager]

T-Mobile continues to buyout ED (2.5GHz) spectrum in small amounts.  Here is California:

https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1671153433&attachmentKey=21487162&attachmentInd=applAttach

 

https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1861609718&attachmentKey=21487164&attachmentInd=applAttach

 

And here is a new lease from the Ewiiaapaayp Band of Kumeyaay Indian, which was recently acquired from the FCC:

https://wireless2.fcc.gov/UlsApp/UlsSearch/licenseMap.jsp?licKey=4345805

 

As you can see from above, this is a significant amount of spectrum that will hopefully allow this Band to receive much better internet.

 

Here is Washington state (none in Oregon):

https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=430721178&attachmentKey=21504414&attachmentInd=applAttach

 

Likely many others elsewhere in small amounts.

 

[iansltx]

5 hours ago, dkyeager said:

Agree that we must discount Verizon. 

What your observing could also be a design issue of 5g: the complexity of being able to serve so many more devices where signal is strong leads to edge deterioration. I agree that this might be able to be addressed in the future.

I would like you to discount the bandwidth size theory if you can. What is the bandwidth difference in places where you have observed n41 falling short of b41?

I believe I've seen the difference even when n41 was at 40 MHz, though it's been awhile since I checked that. My market is 100 and other places are 80 right now, vs. 2CA B41 (so 40 MHz).

[cooltech6597]

Saw 40+20 n41 CA at Findlay over the weekend. Was interesting to see...

 

[mdob07]

It looks like even the unconverted keep sites are turned off now also here in Louisville. I tried to connect to 2 known keep sites today and both times would get Verizon on B26 and AT&T on B25. On B41 it would just pick up T-Mobile.  

[floorguy]

The purge in Northern Utah is going... I thought as much as I have seen quite a few towers being "worked" on... But it was confirmed today driving by a sprint only tower... GONE  Pole is there but no antenna's I HOPE they put some Tmo back up, because it is hot garbage at and around that tower.. perfect fill in spot.

200 N angel, Kaysville utah... You can see the tower.. 

[dkyeager]

More 2.5 purchased to fulfill merger 5g obligations for Watertown SD area for 8 SD counties and 5 MN counties: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=138600801&attachmentKey=21481609&attachmentInd=applAttach

 

Change in ownership to T-Mobile from Northern Arizona University (already leased).  https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1026040304&attachmentKey=21432847&attachmentInd=applAttach

"The assignment includes EBS license areas covering parts of 42 Cellular Market Areas (“CMAs”) in markets across the United States, including Albany, New York; Knoxville, Tennessee; San Antonio, Texas; Nashville, Tennessee; Orlando, Florida; Rochester, New York; Evansville, Indiana; Rutland, Vermont and Amarillo, Texas and portions of 107 counties in these areas."

 

This one covers the greater Boston area and appears to be a new lease: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=111517181&attachmentKey=21340773&attachmentInd=applAttach

"The Lease Agreement cover portions of eight local markets in portions of seven counties in Massachusetts, four counties in New Hampshire and one county each in Connecticut and Rhode Island and there is no need for any additional competitive review."

 

A weird one from a Burlington Vermont Bank https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=453703197&attachmentKey=21526085&attachmentInd=applAttach  Can you say due diligence? IMO will likely be sold.

 

States recently covered: California, Oregon, Washington, Minnesota, Wisconsin, Michigan, Ohio,  Pennsylvania, New York, Vermont, New Hampshire, Maine, Rhode Island, Connecticut, New Jersey, Delaware, West Virginia.  Only looked at pending.

[dkyeager]

T-Mobile manages to win the state of Hawaii (sharing with Verizon), Nationally a distant third. Worst states appear to be Wyoming, Vermont, Alaska.

https://rootmetrics.com/en-US/rootscore/map/national/united-states/2021/2H

https://rootmetrics.com/en-US/rootscore/map/state/hawaii/2022/1H

https://rootmetrics.com/en-US/rootscore/map/state/wyoming/2022/1H

https://rootmetrics.com/en-US/rootscore/map/state/vermont/2022/1H

https://rootmetrics.com/en-US/rootscore/map/state/alaska/2022/1H

Edited July 13, 2022 by dkyeager
Note: National ranking not done yet for 2022

[Trip]

I wonder how far along they are in converting Sprint sites in Vermont.  When I was there in 2015, large parts of the state had Sprint (3G) but not T-Mobile at all, and I don't think that's changed significantly.

- Trip

 

[Paynefanbro]

Weird how they've already done all of the metro and state rankings for 1H 2022 but the national ranking is only on 2H 2021.

[dkyeager]

57 minutes ago, Paynefanbro said:

Weird how they've already done all of the metro and state rankings for 1H 2022 but the national ranking is only on 2H 2021.

Good catch

[dkyeager]

Maryland, DC, Virginia results:

https://wireless2.fcc.gov/UlsApp/ApplicationSearch/applLeases.jsp?applID=10128331

New lease for the Winchester VA area, about 40Mhz of spectrum.

 

https://wireless2.fcc.gov/UlsApp/ApplicationSearch/applAdmin.jsp?applID=10128331#

New lease for Roanoke, VA

 

North Carolina, South Carolina, Georgia, Florida

New purchase for Ashville, NC: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1234792620&attachmentKey=21494016&attachmentInd=applAttach

[dkyeager]

The non-keep sites in Columbus starting to go dark.

[Dkoellerwx]

Keep sites still kicking on all bands in Omaha, NE. Not sure about non-keep sites. Without band selector, hard to force my phone over to those anymore. Still getting automatically pushed to the keep sites when in range.

[mikejeep]

Quote

• Since our previous round of testing, AT&T's median download speed increased from 47.8 Mbps to 60.5 Mbps.
• T-Mobile's median download speed increased from 58.5 Mbps to 92.5 Mbps.
• Verizon's median download speed increased from 45.6 Mbps to 56.5 Mbps.

That is from the 1H22 Boston market test -- that is a statistically significant jump and massive lead by T-Mobile.. yet they finished #2 in the market for network speed. Kind of funky.

[dkyeager]

Covering ED license frequency adds in Kentucky, Tennessee, Alabama, Mississippi

License purchase in Jacksonville, MS: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1735621489&attachmentKey=21153691&attachmentInd=applAttach

 

North Dakota, South Dakota, Montana, Wyoming, Idaho:

Central South Dakota: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1789253661&attachmentKey=21511950&attachmentInd=applAttach

 

Williston, ND area purchase: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=275169369&attachmentKey=21528230&attachmentInd=applAttach

 

[dkyeager]

Louisiana, Texas, New Mexico, Arizona

Mesquite area, TX: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=715324474&attachmentKey=21366241&attachmentInd=applAttach primarily a fill-in new lease.

Houston and Galveston areas purchase: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=404866740&attachmentKey=21457038&attachmentInd=applAttach

Houston lease to buy: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1023756578&attachmentKey=21465258&attachmentInd=applAttach

and https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=1416588475&attachmentKey=21465258&attachmentInd=applAttach

Brenham Texas new lease: https://wireless2.fcc.gov/UlsApp/ApplicationSearch/applAdmin.jsp?applID=13488125#

Rice Texas New lease: https://wireless2.fcc.gov/UlsApp/ApplicationSearch/applMain.jsp?applID=13421387

Magnolia, AK purchase: https://wireless2.fcc.gov/UlsEntry/attachments/attachmentViewRD.jsp?applType=search&fileKey=330142865&attachmentKey=21456838&attachmentInd=applAttach  and https://wireless2.fcc.gov/UlsApp/ApplicationSearch/applMain.jsp?applID=13327426

[jreuschl]

My home of Milwaukee has made incredible gains:

• Since our previous round of testing, T-Mobile's median download speed increased from 49.3 Mbps to 294.3 Mbps.
• T-Mobile's median upload speed increased from 14.1 Mbps to 22.7 Mbps.

Sprint and T-Mobile used to fight for the distant bottom here.