It’s really strange with out Internet service as of late. Using Ookla I’ve been finding our downloads have been 2 MB at the best while our uploads 50 MB or better. Normally it’s been 7 or so MB down and 2 MB or so on the up. I’ve done the reboot on the trashcan and a full shutdown restart with no changes. I called TMO’s customer service and the tech probed the gateway and then had me do a full shutdown/restart with no changes. This is the only thing going where we live besides satellite or dialup, we had DSL but the speed was 1.5 down at the best and now the port we had has been given to another customer The neighbors are interested in TMO but I keep telling them to hold off until the convert an old Sprint tower near us but with these slow speeds it’s going to be a long wait.
It sounds like your stuck on the 4G LTE. Look at the Primary and Secondary Cellular signal information on the “Status” page. What bands are those on? Also look at the cellular metrics on the Overview page and record those. If you have the Nokia GW and it does have a 5G signal that should be on the secondary signal. If you have no secondary signal it is just using the 4G LTE signal.
The cellular metrics actually look pretty good. The LTE signal is a little weaker but no worse than mine. With similar 4G cell metrics I get 40-60 Mbs. I’m surprised your downloads are not much higher given the 5G n71 metrics. You should be getting 130-170 Mbs downloads with metrics like that. Are the WI-FI statistics clean without errors or packet drops? If the metrics are that good performance should be much better. I tend to believe there is something wrong at the tower. If the local LAN statistics are clean and transferring files on the WI-FI fast then it seems the problem is upstream.
If there is suspicion about wireless vs. wired connections from the client you use run speed testing with an Ethernet cable vs. via the wireless network. If all clients are connecting via wireless and the results of speed testing are consistent to a reference server then a comparison with a wired client would be good to know.
If all tests still reflect poor download speeds I would still be suspicious of the tower operation.
With respect to your neighbors interest. What I have learned here is that especially on the n71 towers they have only so many available connections before they stop taking customers in a given location. If they delay and do not request a subscription they might be left in the queue. If they go ahead and get the gateway and try it they do have the time to send it back OR if T-Mobile tells them it will be a month or two before they have the tower stable then they can have their home internet account suspended (at no cost) and then jump back on when the tower is right. Here the tower loaded up really fast and I have heard from multiple people in our area that are in the queue. Unless someone drops out OR a new cell becomes available they would have to wait or go for another solution. Just food for thought.
IF I was stuck with a slow expensive solution I would jump in right away. Delay means they will probably not get a spot on the tower. If it does not work out they can cancel at any time or suspend and wait till it does work.
Another thought on that is if your neighbor gets a gateway/subscription and it works well for them then the datapoint may be very helpful. It could suggest that the problem is more with the gateway you have vs a tower issue. You could request another gateway and have them send one out. Just a thought.
Thanks for getting back with me. Here’s what I’ve got:
How would I check the WiFi stats? We’re just using the TMO trash can for in-house WiFi and we don’t do any internal transfers. We just use email and web searches and watch videos, we’d like to do some streaming but we’ve never had the download speed to do it reliably.
Thanks for you help.
Log onto your router (from a browser) and check the statistics under WLAN. See if there are a lot of reported discards, errors, etc.
I would also try to plug directly into it with an Ethernet cable if you have one available.
//
Here’s the 2.4
Sent Bytes | 1068832079 |
Received Bytes | 166023607 |
Sent Packets | 1081189 |
Received Packets | 671984 |
Discarded Sent Packets | 0 |
Discarded Received Packets | 0 |
Sent Errors | 217 |
Here’s the 5
Sent Bytes | 5362468602 | 448922047 |
Received Bytes | 878641614 | 46062531 |
Sent Packets | 4501060 | 421593 |
Received Packets | 3655453 | 227307 |
Discarded Sent Packets | 0 | 0 |
Discarded Received Packets | 0 | 0 |
Sent Errors | 171 | 3 |
Either the gateway does not record the discarded packets or the communication is clean so if there are not signs of damaged packets then that would not account for a performance problem. If there are discards it often is due to damaged packets and leads to retransmissions which of course leads to poor performance especially when there is repeated packet damage.
I have never been able to get any answers as to the “Sent Errors” and mine reports those all the time on the WIFI radios. I can see no indication they cause any problem.
Since you have the PCI, physical cell identifier values it would be possible to locate the source for those signals. Cellmapper.net is ~80% maybe more, effective in having the location of the towers. Some of the newer 5G cells are not in the database they maintain but the 4G LTE most likely are. In this case locating the tower and evaluating another location for the gateway would not be that useful. You might be able to improve the signal reception but the cellular metrics are actually pretty good. With the 5G NR signal strength the downloads should be MUCH better than you stated in your post. There has to be something wrong on the tower or that gateway has some problem. I would lean toward the tower but it could be on either end. With such good signal quality and the low signal noise it is strange that downloads are not in the 130-180 Mbs at least.
Primary Signal
PCI 480 Band B66
Secondary Signal
PCI 873 Band n71
If you are using Speedtest.net to run speed tests be sure to get the details/all the statistics and instead of testing with which ever server is listed as the “optimal” server run several tests with select servers. I am in East TN but I use the T-Mobile Charlotte, NC server most to test against. I use it as a reference as it seems to be very consistent for results. The results are dependent upon a number of variables but using one or two locations for reference points seems to provide a better benchmark. The server in Charlotte seems to be fronted by better connectivity and resources than some. A small town ISP is probably not likely to provide results that are as good as a larger solution. It depends on traffic, router hops, distance, capability of the equipment etc...
You can use fast.com as weil which is helpful for comparison. There are a number of different speed testing solutions out there.
I would still suggest to get more information from T-Mobile on the connection you have with them and let them know how poor the download performance is. Push for answers. They may be upgrading equipment so that might be part of the problem. It takes time for them to get it nailed down it seems.
I’ll check with a network cable later today but I’m leaning toward the tower as being the problem and I always did.
It’s a good idea as it should effectively eliminate the WIFI from the equation. It may also help with support calls as one of the troubleshooting steps you have taken.
//
From the initial posting the behavior appeared to me tower equipment centric. The values from the speed test did not add up as the should have. After working on large enterprise networks resolving issues one of the most common aspects of the job was to prove where the problem was not at. Customers would often report the problem was with our gear and they knew we would work the issues or come to a site because they got more push back from the larger vendor. We had to prove our equipment was not at fault in many cases. Confirming the operation of the gateway and local gear just provides more confirmation where the issue resides. When T-Mobile engineers have the equipment right it is a very good solution for the price.
A little update. Sunday and today (Monday) up downloads are around 10 MB which is where they’ve been but the uploads are still smoking around 50 MB. With today being a holiday it could be less traffic on the site, not sure if that is so since the area I live in is more a bedroom community so I’ll keep an eye on it and make another post.
Thanks to all who posted with info.
I just have to agree the download speed is horrendous. You really need to press T-Mobile for more information. What they state and what is don’t even come close. They should step up and actually provide information about the “other factors”. The router’s cellular metrics for 5G and the actual speeds just don’t make sense. The delivery from the tower is not right.
With the 5G cellular metrics below, that you reported, you should get +10x what you get down at least. Only getting 10 Mbs download speed is pretty much blah DSL speed not high speed at all.
RSRP -82 dBm
SNR 16 dB
RSRQ -11 dB
“T-Mobile 5G Home Internet customers see typical download speeds between 33-182 Mbps, which is great speed for streaming video, surfing the web, working from home and most types of online gaming. Speeds can vary depending on location, signal strength and availability, time of day, and other factors.”
This thread may be unwatched, but for what it’s worth, I am getting download speed of 0.59Mbps and upload speed of 39.5Mbps. That is NUTS!
There must be some throttling going on. No way is 39.5 upload possible while 0.59 is all I can get down.
Is there ANY other explanation?
If the download speeds are falling to single digits download during the day there is something fundamental WRONG with the programming or the delivering cell subscriber loading. Be sure to confirm you are seeing both the 5G secondary signal and the 4G LTE primary signal when the downloads are so poor. With the n71 or n41 neither should be established to deliver such poor performance figures. It is possible that the QoS is not properly configured OR they have overloaded the cells on the tower. Someone needs to evaluate the frequency and how the cell is configured for delivery. I suppose the distance of the user’s gateways from the source may also be a factor. Be sure you have a clear idea of what tower/cell is delivering the 5G signal and how far you are from the tower. Every user’s situation is a little different so external factors have to be considered. I am not trying to make excuses for T-Mobile’s delivery in the area where you are. It is just a fact of cellular delivery that there are multiple factors to consider with cellular delivery and successful delivery over an area may be challenging due to external factors.
It may be T-Mobile intends to improve coverage with more cell coverage but T-Mobile should be up front about intentions and why the current behavior is taking place in that area. If T-Mobile cannot deliver on their expectations for download speeds between 33-182 Mbps then they are failing in that area and need to correct it.
I consistently get 5 bars of 5G all day long. I only get the n71 band, but from about 9am through midnight, I get single digit download speeds and this IS on 5G. I’ve had work tickets opened and the cell techs say there is nothing wrong with the tower and it is functioning as expected. Hard to get them to fix something that they “believe” is working correctly.
I wonder if somebody at T Mobile would be able to SEE the throttling.
The answer is YES! Mellissa in Charlotte saw the throttling started when I exceeded my 300 GB cap. She helped me to get an “uncapped” account. Problem solved! She said that when throttling is invoked, the speeds can go “crazy” hence the high upload speeds and the abysmally slow throttled download speeds.
The answer is YES! Mellissa in Charlotte saw the throttling started when I exceeded my 300 GB cap. She helped me to get an “uncapped” account. Problem solved! She said that when throttling is invoked, the speeds can go “crazy” hence the high upload speeds and the abysmally slow throttled download speeds.
Does the new uncapped account cost more? Or is it simply a change on TMO’s side?
Doing a speed test in the middle of the night should give you an idea of what the system is capable of when there are few users on. Network congestion seems to be the biggest limiting factor for me. At night I’ve seen 25 down and 45 up. During the day, it can drop to 2 down and 30 up.
Sadly I get the feeling that TMO is OK with all of this poor performance.
If the download speeds are falling to single digits download during the day there is something fundamental WRONG with the programming or the delivering cell subscriber loading. Be sure to confirm you are seeing both the 5G secondary signal and the 4G LTE primary signal when the downloads are so poor. With the n71 or n41 neither should be established to deliver such poor performance figures. It is possible that the QoS is not properly configured OR they have overloaded the cells on the tower. Someone needs to evaluate the frequency and how the cell is configured for delivery. I suppose the distance of the user’s gateways from the source may also be a factor. Be sure you have a clear idea of what tower/cell is delivering the 5G signal and how far you are from the tower. Every user’s situation is a little different so external factors have to be considered. I am not trying to make excuses for T-Mobile’s delivery in the area where you are. It is just a fact of cellular delivery that there are multiple factors to consider with cellular delivery and successful delivery over an area may be challenging due to external factors.
It may be T-Mobile intends to improve coverage with more cell coverage but T-Mobile should be up front about intentions and why the current behavior is taking place in that area. If T-Mobile cannot deliver on their expectations for download speeds between 33-182 Mbps then they are failing in that area and need to correct it.
Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.