****The Official 5G Home Broadband Thread**** (Three/EE/Vodafone/etc)

[lovelyhead]

OK so I did a factory reset and all the fun that brings what with having to reconnect all my devices and for the first few speedtests the speeds were back up to around the 150mbps mark which was great. About an half an hour later my speeds dropped back down again to around 25mbps whilst my upload was still at around 70mbps.

I did a little experiment and changed the router to only 4G mode and my download stayed at around 25mbps and my upload was also around 25mbps.

I am not sure what to make of it.

Edit now my speedtests on 4G are at around 60mbps down and 25mbps up. I can't make any sense of it, it is so variable.

 

[GregI]

I don't know much about this stuff but I was planning to get an unlimited Lebara SIM (no contract £9pm) to tide me over during the switch from FTTC to FTTP.

I looked at something basic like a TP-Link TL-MR105 but on some deal site someone said "Cat4 lte modems do not provide lte carrier aggregation .. Also its not cutting edge. The average new phone has cat ratings above 10".

If I just want a basic internet connection that can stream YouTube etc for a couple weeks, does it really matter?

 

[fray_bentos]

OK so I did a factory reset and all the fun that brings what with having to reconnect all my devices and for the first few speedtests the speeds were back up to around the 150mbps mark which was great. About an half an hour later my speeds dropped back down again to around 25mbps whilst my upload was still at around 70mbps.

I did a little experiment and changed the router to only 4G mode and my download stayed at around 25mbps and my upload was also around 25mbps.

I am not sure what to make of it.

Edit now my speedtests on 4G are at around 60mbps down and 25mbps up. I can't make any sense of it, it is so variable.

Possibly too many WiFi devices connected, or some other setting you are changing that is doing it then. Have you checked wired ethernet speeds?

 

[lovelyhead]

Today I just achieved one of my fastest speed tests, 230 down 60 up with no changes.

The problem has to lie with Three. Too many people using the one mast? Would wet trees in the line of sight to the mast cause big problems?

 

[fray_bentos]

Today I just achieved one of my fastest speed tests, 230 down 60 up with no changes.

The problem has to lie with Three. Too many people using the one mast? Would wet trees in the line of sight to the mast cause big problems?

Rain is known to affect 5G signal significantly, so possibly. However, there will be much less water on wet trees vs a longer distance of signals passing through rain drops. Do you have a loft? If so, put the router up there, high up in the rafters. My transfer speeds jumped from 50-100 Mbps to <900 Mbps (1 mile away from the mast, but with direct line of sight from the roof, but not from upstairs windows). Obviously not possible in flats/tenements...

 

[lovelyhead]

Rain is known to affect 5G signal significantly, so possibly. However, there will be much less water on wet trees vs a longer distance of signals passing through rain drops. Do you have a loft? If so, put the router up there, high up in the rafters. My transfer speeds jumped from 50-100 Mbps to <900 Mbps (1 mile away from the mast, but with direct line of sight from the roof, but not from upstairs windows). Obviously not possible in flats/tenements...

I actually tried sticking it in the loft to test it but get a worse signal up there.

I am pretty sure that it is an outside factor causing the problem as my signal is always excellent every time I have checked. It never has to drop down to 4G unless I force it to.

 

[fray_bentos]

I actually tried sticking it in the loft to test it but get a worse signal up there.

I am pretty sure that it is an outside factor causing the problem as my signal is always excellent every time I have checked. It never has to drop down to 4G unless I force it to.

If your signal is excellent, that does not fit with your speeds. If you are not using ethernet cable, could simply be the quality of your internal WiFi signal/channel congestion from other users / lots of device sharing the same channel. You can check channel congestion using inSSIDer.

 

[Pulseammo]

If your signal is excellent, that does not fit with your speeds. If you are not using ethernet cable, could simply be the quality of your internal WiFi signal/channel congestion from other users / lots of device sharing the same channel. You can check channel congestion using inSSIDer.

Honestly I have this as well in my house from my testing the other day (signal strength being disjointed from throughput). I put the router outside with clear line of sight to the mast. RSRP was lowest I've ever seen around low -70s, down from -85 usual, SINR and RSRQ also good and speeds were only 150mbit...

Moved the router to the attic again, tried a few places, one place I'm getting -93 RSRP, -11 RSRQ and 24 SINR giving 250mbit... move the modem a bit again, RSRP drops to -85, other stats similar... 200mbit.

Unless I'm reading RSRP backwards? I am under the impression numbers closer to zero are better?

 

[fray_bentos]

Honestly I have this as well in my house from my testing the other day (signal strength being disjointed from throughput). I put the router outside with clear line of sight to the mast. RSRP was lowest I've ever seen around low -70s, down from -85 usual, SINR and RSRQ also good and speeds were only 150mbit...

Moved the router to the attic again, tried a few places, one place I'm getting -93 RSRP, -11 RSRQ and 24 SINR giving 250mbit... move the modem a bit again, RSRP drops to -85, other stats similar... 200mbit.

Unless I'm reading RSRP backwards? I am under the impression numbers closer to zero are better?

Try putting the SIM in your phone. What speedtest speeds do you get outside? If faster on the phone, it is the router going slow / WiFi limited speed (e.g. device using WiFi5 ~200 mbps vs. WiFi 6 ~600 mbps when near the router). Incidentally, on WiFi6 downstairs I get 250 mbps on my phone connected to the router, but ethernet wired to gets 500 mbps (at 9pm). If the same download speed on the router and on your phone, then that's due to 5G users if you are in a populated area /mast congestion. I am on the very far outskirts of Edinburgh near the bypass/Pentland hills with a relatively low population density, which may explain my speeds.

 

[Pulseammo]

Try putting the SIM in your phone. What speedtest speeds do you get outside? If faster on the phone, it is the router going slow / WiFi limited speed (e.g. device using WiFi5 ~200 mbps vs. WiFi 6 ~600 mbps when near the router). Incidentally, on WiFi6 downstairs I get 250 mbps on my phone connected to the router, but ethernet wired to gets 500 mbps (at 9pm). If the same download speed on the router and on your phone, then that's due to 5G users if you are in a populated area /mast congestion. I am on the very far outskirts of Edinburgh near the bypass/Pentland hills with a relatively low population density, which may explain my speeds.

I'll continue to tinker, but all that testing was done with a 30metre long ethernet cable plugged into the Zyxel modem to try and remove as many other factors as possible. I've driven through the area you live in before on the way home from the airport so I know how the houses are dotted about, I'm rural as well on the border but in a village of about 4k and the village over which the mast also serves is about 4k population (so basically higher density than you). That said I don't see speed fluctuations even in the middle of the night so I don't think it's an contention issue.

Still, I'm delighted with the speeds I'm getting, I just found it curious how disjointed the speeds were from signal strength when I was faffing about with it the other day. Also it doesn't matter because I'm definitely limited by the speed of my router at the moment, it can only route around 200mbps even with hardware offload, so any further testing will need to wait until I could solve that.

Just thought I'd pitch in since @lovelyhead is having throughput issues that seem to be not linked to signal strength...

 

[lovelyhead]

I was reading through some forums about low 5G speeds and someone suggested changing the IPv4/IPv6 Mode from "IPv4 Only" to "IPv4 IPv6 DualStack". The speedtest I did before this gave me 25mbps download and the speedtest 5 minutes after changing the settings gave me 80mbps download. I just did another speedtest and it is at 106mbps download. Time will tell if that has been the problem.

I connect to the mast at Culross, Fife which is quite a tourist destination due to outlander and the palace so I am wondering if the mast is being overloaded. Well if changing the IP Mode settings fixes it long-term I will have my answer.

Edit: scratch the idea of enabling IPv4 IPv6 DualStack. If you enable this Steam and PC Game pass stop being able to connect. I put it back on to IPv4 and the speed stayed high. Currently downloading from Steam at 100mbps. I am convinced it is a Three problem rather than a me problem!

 

[Pulseammo]

Edit: scratch the idea of enabling IPv4 IPv6 DualStack. If you enable this Steam and PC Game pass stop being able to connect. I put it back on to IPv4 and the speed stayed high. Currently downloading from Steam at 100mbps. I am convinced it is a Three problem rather than a me problem!

Are you changing the stack mode in the APN settings or in the broadband cellular wan settings or both?

I'm swiftly coming to the conclusion that speedtest.net results are entirely unreliable, three is definitely prioritising their traffic. Today was the first day that I've realised contention, around the middle of the day speeds dropped to around 40mbps downstream.

At other points at the weekend speedtest will happily report 200mbps+, but actually downloading files barely achieves 100mbps, so I've been using waveform's speedtest at the same time to compare results since it is closer to my actual speed in downloads.

So I'd suggest people not bother with speedtest.net on three since they are rigging it (imo).

Test Batch 1:



Test Batch 2 later in the day:



My signal strengths:

 

[lovelyhead]

I just tried the waveform speedtest. I did a Ookla speedtest then the Waveform speedtest then started a large game download on steam, all within about a minute. The Steam download speed was fairly close to the Ookla speedtest result. The Waveform result was about half of the Ookla speedtest and the Steam download.

 

[Pulseammo]

I just tried the waveform speedtest. I did a Ookla speedtest then the Waveform speedtest then started a large game download on steam, all within about a minute. The Steam download speed was fairly close to the Ookla speedtest result. The Waveform result was about half of the Ookla speedtest and the Steam download.

I've found steam is generally good actually also... but for instance using something like a file download from here https://proof.ovh.net/files/ or testing against scaleway's iperf3 servers I've found them to match waveform's results.

 

[lovelyhead]

I just tried that site and was getting 1.5MB per second download then downloaded a game from Steam and was getting around 8MB per second.