5G Home Internet vs. phone data speeds

  • 29 December 2022
  • 15 replies
  • 5473 views

Badge

So I’m a Magenta Max customer and mostly happy with my service. We live in a town where the 5G can be very strong depending on time of day/which part of town you’re in. 
 

So I started considering switching my home internet to T-Mobile’s 5G Home Internet, and did some research. I turned off my phone’s wifi to run speed tests on my phone’s 5G data from inside my home. It varies a TON depending on the room and the time of day. I’ve seen download speeds as high as 300+ and upload speeds of about 6-8 (at the times you’d expect… late night or early morning) but as low as 6Mbps for downloads and 1 for uploads (during peak times). That lower number concerned me as I work from home during the day, participating in a lot of mandatory video calls. I can’t have my internet just disappear during peak work hours. A great price for home internet won’t matter if I lose my job because I can’t stay online on a Thursday afternoon.

 

So I went to a T-Mobile outlet to ask an employee about it. The kid told me the phone data speeds are worse than what I’d get with Home Internet because the wifi customers are “on a different tower.” I’m no expert on this stuff, but this sounded made-up to me, like he just wanted me to feel better so I’d buy something. But maybe someone who knows more can confirm if the T-Mobile employee was correct: Are Home Internet speeds better than what I get from my T-Mobile 5G device in the same building, and if so, is it because they deliberately use different towers for that service, as this guy told me? 
 

I should add that this kid was very obviously trying to get me to sign up on the spot, showing me 1000+ speed tests on his device (irrelevant to me because I don’t live at the T-Mobile store so my speed won’t be the same). So I protected myself by leaving without signing up for the trial service, although I am still considering it. 
 

So anyone with better knowledge than me - and with no stake in trying to sell me something on commission - what’s the real story here? Can I reasonably expect better speeds (and most importantly, no frequent drops in service while I’m working) then I get on a 5G mobile device from within the same residence? Thanks for any help!


15 replies

Badge

Mine is  just the opposite. My S22 Ultra  will get 400 or better  during  off peak time standing  next to the modem while the home internet  .odem will  peak at a little  over 300

Userlevel 5
Badge +7

Denny, they definitely use the same tower. The “kid” in the T-Mobile store was wrong.

As far as speeds, iPhones and Androids are throttled because of limited battery and CPU power. If you place a T-Mobile gateway in one of the areas of your home where you have great reception of 5G, you'll be pleasantly surprised at the speed you'll get OTA.

Also, there is a 2-week, no questions asked trail period. Give it a shot while you still have your old ISP and see if you can achieve speeds that will satisfy your home’s needs. Otherwise, return it.

For me, I had Xfinity 800 Mbps, but I hated having to negotiate the price every year, and I really didn't like the data caps imposed on users. I am not a gamer, but we do stream a lot of TV and have multiple computers, iPhones, smart switches and gadgets. In my environment I get approximately 50-70 Mbps down and 10-11 mbps up, but I can only connect via 4G. The 5G signal isn't strong enough or can't make it through all the trees surrounding our home or through the concrete block construction of our home in S. Florida. 

The bottom line is that the bandwidth I receive is more that enough for our activities. We also use T-Mobile as the provider for our iPhones, so that qualifies us for $25 per month rate, all-you-can-eat for life, I'm a very happy camper. From what I can tell on this forum, there are some problems, especially for gamers, but for most people who can get good access to 5G, people really like the service.

Userlevel 7
Badge +11

Mostly… T-Mobile has deprioritized Home Internet over cell.

I can get as good, if not faster on my Home Internet .. just not as often,

S21 ≈ 150-200Mbps indoors - 5G disabled (becomes worse with 5G enabled, as it is weak from +1 mile away), LTE ≈ 1000’ away.  During the day, ≈ 150Mbps.  Night time ~200Mbps.

Home Internet ≈ 50-200Mbps, typically 50-150Mbps, but some times as low are  10-30Mbps.  During the day ≈ 75Mbps.  Night time ~=150 upwards to 200Mbps or better.

Upload is typically 30Mbps on both.

I should notice better performance once T-Mobile upgrades the local tower.

Userlevel 7
Badge +8

I agree 100% with “formercanuck” the phones do have priority over the fixed broadband gateways due to the nature of cell phone calls and roaming from tower to tower. Testing with a cell phone vs. the gateway can be rather deceptive. From one manufacturer’s cell phone model to another there are often huge differences in the radios in use. If you buy a premium cell phone you are getting better tech than the average cell phone so it is not a comparison that will lead to conclusive results. 

The service in one area vs another can vary greatly and it seems to come down to higher density urban areas with heavier user loads requiring the use of bandwidth throttling due to load. T-Mobile may be expanding their network but it seems to me the cells are still often loaded heavier than is optimal for service to be what users are/were expecting. The ONLY way you will know how the T-Mobile solution will work for you in your location is to try it out. Exercise the solution as much as you can during the trial period to profile proper expectations. You can read threads on Reddit, here and other forums and get a better idea as to if it is what you will find satisfying or not.

T-Mobile runs a IPv6 network and a 464XLAT solution so don’t expect NAT and port forwarding controls you might have become reliant upon without a VPN solution. Games that rely upon UPnP are not going to be functional on the T-Mobile solution. Go read the FAQs on the T-Mobile pages. Don’t be blind sided by not researching the details. I know the T-Mobile home gateway solution can be pretty darn great in some markets. Here we are in a rural area with n41 for the data load and it is very good. We started out with n71 delivery and then they upgraded the cell to n41 and the speeds just about doubled. Still the speeds vary depending upon the time of day/night and load. Weather can also be a factor. Well, there are a number of factors but if your location is close to a cell source and good clear delivery it can be pretty good. In higher density markets where user load is heavier that equation may not work for you.

If you are interested in the way the tech is used: (interesting read)

https://prateek-mishra.medium.com/en-dc-overview-62d35521edc4

Badge

Thanks for the feedback, everyone. I decided to do the trial run. I’ve had the gateway up and running for about an hour on a Friday afternoon, which I assume is a pretty decent time to test signal strength at peak (or close to peak) usage time.

 

Im getting speeds ranging from 45-220 down depending on the device and the room I’m standing in, and from 7-12 up. The lowest of those numbers is still better than I’m used to from AT&T in my neighborhood. So far, so good.

One curious thing though - the video streaming is nice and fast on my work laptop. However, when I sign into my work’s VPN and attempt to load one of the web pages that requires the VPN, they no longer load. The VPNs are signed in and connected, but the work-related pages that need the VPN are not loading anymore on the T-Mobile wifi. Am I just missing a setting or something?

Badge

Welp. At about 4:30pm on Friday the download speed plummeted to 11. By 5pm it was 6Mbps. On top of that, my cellular service dropped from 3 bars to 1 (even briefly went into SOS mode). 
 

I expected a drop at peak hours. But I’m not sure total paralysis of the service during work hours is something I can deal with long-term. Those are peak hours for a reason. People need internet / cell service most at that time of day. 

Userlevel 7
Badge +8

That is exactly why I strongly recommend anyone who is interested to do the trial run and not dive in after termination of the current service. It could be T-MOBILE is still working on the towers in the area. They tend to start loading before they are done I think. Some instances it seems as if users come on and speeds are great only to go south after a month or two. 
With a VPN keep in mind your packets will be encapsulated in the tunnel. You might need to lower the MTU from 1500 bytes to I think 1440-1460 bytes. Large max size packets are where the problem can be encountered. 
You might check downdetector.com to see if there are reports of outages for T-MOBILE in your area. 

Userlevel 7
Badge +8

The safe size of packets for an IPsec VPN is 1328 bytes. 
TechTarget.com (search for Managing VPN bandwidth)

There is also a MTU article in the Dig Deeper on WAN tech. 
This assumes you want to geek out. 😎

Badge

This seems like a similar topic to what I am looking at.  My 5G home internet is typically about 60-90 MBS down.  But I am curious why my phone on 5G in the same location (iPhone 12) is getting about 3X the speed down?  If my phone and the home internet gateway are in the same location, why would the gateway be that much slower?

Userlevel 7
Badge +8

The difference between the two is somewhat explainable. On the T-Mobile 5G cellular network phones on the same cells/frequencies as the gateway devices will have priority due to being phones. Often people overlook the cellular antennas and chips in the devices. They are not equivalent so the difference in the two cellular hardware components has to be considered. If you are running testing from the gateway is is probably tested over IPv4 whereas your phone could well be IPv6 not IPv4 so the routing for the two can different. There would also be a bit more overhead with the gateway over IPv4 due to the 464XLAT and the translation from IPv4 to IPv6 and back to IPv4. Plus you also have to run a speed test against the same server in many cases. The “optimal” server the test each uses also can be different and render results that are not even close. 

As these speeds go down when they say it is high congestion, is T-Mobile just pawning home internet customers to different isp servers for their phone customers to use their towers only? I wondered because I tested 2 phones with Ookla for internet speed over 100 times within an hour (10-11PM). On every test with both phones when WiFi was turned off and it was on 5G, all tests download speeds were between 180 to 251 and t-mobile phones connected to t-mobile Pittsburgh server (same server all tests show the Home Internet gateway connects to early morning around 350 until it slows down around 4pm around 100). But when tested on the phones with t-mobile home internet, same amount of tests for the “recommended best signal switching between 5G & 2.4G”, -tests on 5G and tests on 2.4G during this same time as the cellular 180-251 5G tests, all scores ranged from 29 to 50 with almost all between 29-33. But the thing that I have a question about is because all tests were connected to different isp servers (Armstrong Cable in Zelienople, PA, Magna5 in Boyers PA or wVNET in Morgantown WV). From between 25 & 30 tests for each network, 100% of the T-mobile Home Internet WiFi tests connected to a non-t-mobile isp server far away from where I am located  and 100% of the 5G cellular network tests connected to the t-mobile Pittsburgh  network close to my home (that my TMHI is normally connected to except for evenings). Are they throttling TMHI customers or just pawning us off to other companies at certain times of the day?

Update to previous post. A knowledgeable T-Mobile source stated T-Mobile does pawn off their home internet customers to 3rd party isp companies at low speeds during times of expected congestion (evenings/late night) and when towers are congested so high speed goes to cell phone customers.

Userlevel 3
Badge +4

@Otis824, "...to 3rd party isp companies..."

Were you given any more details as to who these ISP might be? 

 

 

Denny, calling 611 will not be anymore helpful than the store. For a year after the tower, gateway came out no one could explain how it would get better reception than my phone. On accident I came across a video and it explained all. It's kinda backwards. I just got off the phone and got nowhere but I think I have found the place to get my answers. Or videos. T mobile had a load of videos. 

So I thought I would jump in and add my speeds...first, I changed from ATT wired Internet back in December. Due to where I live, I had no other option. ATT was incredibly slow peaking at an average of only 10-14 down and maybe 7-9 up! Plus, we could only have a couple of things connected at a time! It was horrible! So with this 5g , average is 70-135 down and 30-40 up and as many devices as we need to use at any given time! I'm home all the time so I run tests randomly and the speed has been as slow as maybe 50 down but then only for a few minutes or so. Since we have had it for 3 months or so now, we have only had an issue twice that I can think of! I've been pleasantly surprised with this service and with the price! ATT kept increasing until it was almost 70 a month with those extremely slow speeds! 

Reply