T-Mobile 5G Secrets Revealed: Here’s Where It Doesn’t Work Well

T-Mobile’s new “nationwide” 5G can double your 4G speeds…or not. We explain why and give you maps of where it works well.

By Sascha Segan

For people in many major metro areas, T-Mobile’s new low-band 5G and its OnePlus 7T Pro McLaren phone will offer little advantage over 4G—unless the T-Mobile-Sprint merger goes through.

T-Mobile’s low-band 5G launched last week. I’ve been testing it with the McLaren phone—first in Maui, where it performed very well, and then in New York City, where it hasn’t performed as well. T-Mobile’s low-band 5G is most likely to benefit people in smaller cities and rural areas. In those places people could see significant speed increases. In 16 of the nation’s top metro areas, T-Mobile’s 5G will currently bring no benefit. It will bring relatively little benefit in 22 other major metros.

This will change dramatically if the T-Mobile-Sprint merger closes. If that happens, the two T-Mobile low-band phones—the McLaren and the Samsung Galaxy Note 10+ 5G—will be able to use all of Sprint’s 2.5GHz 5G airwaves as well as Dish’s 600MHz airwaves , and that promises faster speeds in many cities within a few months. That said, you should still probably wait for the Samsung Galaxy S11 5G in February. That will be the first phone to handle the full “layer cake” of T-Mobile’s 5G frequencies, including very fast but very limited-range millimeter-wave.

What Is T-Mobile Low-Band 5G?

5G isn’t a speed, a frequency, or an experience. It’s a radio encoding method. It can work on any frequency in any channel size between 5MHz and 100MHz. 4G tops out at 20MHz channels. You hear a lot about millimeter-wave because that’s where 5G is very different from 4G right now, because it’s using channels that are impossibly large for 4G, which gets you those amazing speeds.

But 5G can also work a lot like 4G in 4G bands, and that’s what T-Mobile is doing with its “nationwide” 5G. In 5MHz or 10MHz low-band channels, there’s effectively no advantage for 5G yet—I’ll get to why T-Mobile is bothering in a bit. The speeds for now are just like 4G.

T-Mobile’s low-band 5G relies on 600MHz spectrum (the old TV channels 38–51). That spectrum travels great distances, but there isn’t very much of it available. T-Mobile has either 20, 30 or 40MHz of spectrum almost everywhere in the US. Across the country, it’s using 10MHz of that for 4G-5MHz for uploads, 5MHz for downloads. It’s using the rest for 5G. It can’t combine the two approaches; a phone has to use one or the other at any given time.

So in places where T-Mobile has only 20MHz, such as Atlanta, Boston, and New York, you’re swapping 10MHz of 4G for 10MHz of 5G. They turn out to perform about the same with the same level of load.

In 30MHz zones, including Austin, Charlotte, Los Angeles, and Seattle, you swap 10MHz of 4G for 20MHz of 5G. That will make a slight difference, but it may not be noticeable.

The noticeable difference, which I saw in Maui, comes in 40MHz zones. Those would include Boise, Chattanooga, Spokane, Tucson, and Wichita. There, you’re exchanging 10MHz of 4G for a juicy 30MHz of 5G, and in Maui I saw that boost speeds by up to 100Mbps. These results are from Maui, which has 40MHz of low-band.

Here’s a set of key 40MHz cities, where performance should be good right now. (Click on the map to get a larger, more readable image.)

I had to leave some cities out here because T-Mobile’s spectrum in those cities is inhabited by TV stations that aren’t required to clear out until between January and May 2020. When that time comes around, those cities will boost up to the 40MHz level, merger or no. Some of them may have cleared already; I’m not sure. According to Spectrum Gateway, that list includes Baltimore, Baton Rouge, Buffalo, Colorado Springs, Corpus Christi, Laredo, Raleigh, Rochester, Waco, Wichita, Tallahassee, Tulsa, and Washington DC, as well as most of New Hampshire and Maine.

I don’t have a full map of all the 40MHz zones, but this map from Spectrum Omega is a decent proxy. The 600MHz band was broken up into seven blocks for sale, and to achieve 40MHz, T-Mobile needed four of them. Look at the “E” block, which makes the difference in a lot of areas. The purple regions are T-Mobile, and you can see that they steer clear of a lot of major cities.

If you’re in one of the purple places, you’re likely to see a boost similar to what I got in Maui. If you’re in one of the non-purple places, T-Mobile’s low-band 5G isn’t going to have a lot to offer until the merger goes through.

New York City Blues

Testing T-Mobile’s 5G in New York City, I got speeds that were the same as a Samsung Galaxy S10e on 4G in a best-case scenario and significantly lower than the Galaxy S10e in several locations. That’s because of what the Galaxy S10e can aggregate and what a low-band 5G phone currently can’t aggregate.

The speeds and capacity we get from phones nowadays come from combining different lanes of spectrum, something we call carrier aggregation.

I’m going to fling a lot of frequency band numbers at you now. The important one is band 71, T-Mobile’s 600MHz low-band. When it’s being used for 4G, it’s called 71. When it’s being used for 5G, it’s called n71.

On a 4G phone, T-Mobile generally aggregates up to four lanes of spectrum. These can be mid-band (bands 2 and 66), low-band (bands 12 and 71), or the Wi-Fi-like band 46, which is used for extra capacity in dense center city locations. (There is also an older band number, 4, which is a subset of 66 and which T-Mobile appears to be renaming as 66 for newer phones.)

The McLaren’s hardware can aggregate 5G and 4G bands together, combining them for better speed than 4G or 5G alone. It works well together with bands 2 and 66, T-Mobile’s most widespread bands for coverage and capacity. But it can’t aggregate 5G band n71 with T-Mobile’s other 4G bands.

(If you want to see every possible 4G carrier aggregation combination, feast your eyes.)

Bands 12 and 71 are a wash. You can’t combine 12 and 71 on a 4G phone anyway (they’re too close), so you end up exchanging either 12MHz of band 12 or 10MHz of band 71, for 10MHz of 5G n71, which is basically passing a coin from one hand to the other. In test locations where my 4G phone was combining bands 2/66/71 or 2/12/66, speeds were pretty much the same on the 5G phone.

But when I stumbled upon band 46—ouch. Band 46 is rare, but you really notice it when it’s gone. It’s used mostly in super-congested spaces—sports stadiums, downtowns, college campuses, and the like. I found it in Union Square in Manhattan, and losing it dropped my speeds from 128Mbps with a 4G phone to 25.1Mbps with a 5G phone.

Note: That’s aggregation between bands 2/66/71 or 2/12/66, because you can’t do 12/71 at the same time. But the performance is roughly the same in the two configurations.

Most important, there is no math that gets you to better speeds with the 5G phone in one of the 20MHz low-band markets. There’s no way to gain lanes. You might as well just be using a 4G phone.

Why Do This at All?

Okay, so why bother doing 5G at all in 600MHz? Other than the marketing advantage of saying you’ve got a lot of 5G, a bunch of new capabilities are coming online later in 2020 that T-Mobile wants to take advantage of. Network slicing would let it provide a reliable rural home internet business without home users getting their lanes crowded with mobile users. Low-latency communication, and the ability to address large numbers of industrial devices enable business uses like drone management and agricultural sensors. In other words, it’s about future applications that aren’t just getting internet on your phone. Those things don’t demand fast speeds, but they demand aspects of the 5G spec that haven’t been fully turned on yet.

The Sprint-T-Mobile merger would add 60MHz of mid-band spectrum to the McLaren’s accessible 5G portfolio, boosting speeds considerably and providing a real difference. Sprint’s 5G is most built out in some of the cities where T-Mobile’s low-band is weakest, such as Atlanta, Dallas, Houston, and New York.

Dish also has unused 600MHz licenses right next to the ones T-Mobile is using, often in T-Mobile’s most constricted cities. Part of the merger agreement would involve T-Mobile getting access to that low-band spectrum, which would raise it to 40MHz of low-band or more in Las Vegas, Miami, New York, Seattle, and other places where T-Mobile is now weak. Ookla has a map of that combination in this story. (Note: Ookla is owned by Ziff Davis, PCMag.com’s parent company.)

The networks are all very complementary, and if they merge, that’s a compelling reason to get a phone like the McLaren.

Fourteen state attorneys general are currently fighting the merger in court, in a case expected to last about three weeks. So we could get a resolution to this saga by the new year. If T-Mobile doesn’t merge with Sprint, it will have to rely more on its millimeter-wave holdings for in-city 5G, which means that the McLaren (which doesn’t support mmWave) would be a poor choice, and you’ll need to hold out for the Samsung Galaxy S11, which will presumably support mmWave.

I’m still working on a full review of the OnePlus 7T Pro 5G McLaren. At the moment, it’s $899, where the OnePlus 7 Pro is $549—that’s a $350 bump. The differences are pretty much just a slightly faster processor, a slightly faster-charging battery, more storage, and 5G. With the way I now understand things, I can’t justify paying that price unless the phone gets access to the Sprint and Dish spectrum as well, so keep an eye on the court case.

Originally published at https://www.pcmag.com on December 11, 2019.