Sponsored Links

Is there a way to mask my desktop computer so it appears to be a mobile phone when connecting to a network device?

I honestly think you have a chipset incompatibility between the Marvell nic and the 5g router.

What about my Intel(R) Wi-Fi AX201? How likely is it that both of these NICs have the same problem you're alluding to?

Also, I have now tested with Ubuntu (running on an Optiplex 3020 with a Gig ethernet port) and an Apple MacBook Pro 2017, again, same ~100Mb/s speeds.

Ubuntu:

Ubuntu speedtest.webp


MAC:

MAC speedtest.webp
 
Three do a variety of traffic shaping on their network along with deep packet inspection. If you call them, they say they don’t, but they absolutely do. Arguing with them about it isn’t going to change how their network works either. If the shaping is affecting you, try switching to Vodafone.
 
Good idea, alas, tried that already a few days ago. Manually set it to 2.5g, and then 5g, all to no avail.

View attachment 4892



I have a few linux distros I can test with actually. I really don't think that'll make any difference at all, but I'll give it a quick go and report back.
Sorry missed you’d already done that.

Looks like there is an issue on the 5G router. Tell them it’s faulty and needs replacing. It’s their kit and their responsibility to ensure it works as advertised.

Clearly with two unrelated machines having the same behaviour it’s the common denominator.

I think you’ve done enough exhaustive testing now. 👍
 
Hello,
Have you remembered to check that the eND ID is the same on your Iphone and your 5G router?
And what are the RF signal levels: RSRP, RSSI on the devices?
They are likely to be different.
 
Perhaps a utility to spoof the MAC address of the NIC on my PC, or WMIC/WMI? I want to test a theory that my mobile broadband provider is shaping, or rate-limiting the down-/upload speeds depending on what type of device is connecting, even though they proudly and explicitly exclaim "No restrictions, no limits" with regards to their broadband packages.
The MAC address is not visible inside the mobile network. so mucking around with that won't make any difference. 4G carries layer 3 upwards (5G can in theory carry layer 2 as well but I am not aware that it is deployed like that anywhere).

The MNO can use deep packet inspection to guess what you are doing and looking at TTL could be a way to determine whether the client runs on a phone or on a desktop behind a router. They can obviously also look at the browser's User-Agent string.
I'm using Three's 5G router, and get 450Mb/s from my phone (iPhone SE 2020) connected via WiFi but only around 100Mb's from my PC, via a CAT6 cable, using a 5Gb ethernet NIC called Marvell AQC111C.
If I understand this correctly, you are always using a 5G router but different devices behind the router. I cannot see why the MNO would treat client devices behind the same router differently (as opposed to making a difference between a smartphone and a router).
 
Sponsored Links
Interesting! However, like I mentioned a few posts earlier, I used to have Hyperoptic, and got ~800Mb/s consistently and never had any troubles. If the above applied to my system, I would not have had such fast speeds eith Hyperoptic.
I think everything being known-good with Hyperoptic is probably pretty decisive. I agree there's a lot of weird behaviour being observed and evidenced.

The big picture view is that if 100 Mbps is good enough for daily usage, it might be fine to live with until the root cause reveals itself. Typically such root causes can be obvious in hindsight. It's pretty likely that if one device can achieve 450 Mbps, others should be able to get close to that, especially if it's been seen before with another ISP and their hardware.

I actually saw some weird speedtest-cli behaviour which previously I was ignoring but might look into further because of this thread. I currently get speedtest-cli v2.1.3 download measurements of sub 1 Mbps on wired 1 GbE which is sufficiently laughable to not be worth debugging (fastcom shows 84.64 Mbps with all tests using a Python venv and fresh pip installs).

(bwvenv) [x@x bwtest]$ time speedtest-cli --server 23344 --share; date
Retrieving speedtest.net configuration...
Testing from EE (109.249.187.x)...
Retrieving speedtest.net server list...
Retrieving information for the selected server...
Hosted by 31173 Services AB (London) [35.30 km]: 1200039.52 ms
Testing download speed................................................................................
Download: 0.69 Mbit/s
Testing upload speed......................................................................................................
Upload: 41.28 Mbit/s
Share results: http://www.speedtest.net/result/14224550928.png

real 0m12.237s
user 0m0.470s
sys 0m0.231s
Wed 18 Jan 09:09:07 GMT 2023
(bwvenv) [x@x bwtest]$ time fastcom --quiet
Max speed: 84.64 Mbps
Mean speed: 63.31 Mbps

real 1m2.256s
user 0m2.150s
sys 0m1.047s
(bwvenv) [x@x bwtest]$ date
Wed 18 Jan 09:11:26 GMT 2023
(bwvenv) [x@x bwtest]$ speedtest-cli --version; date
speedtest-cli 2.1.3
Python 3.6.8 (default, Sep 10 2021, 09:13:53) [GCC 8.5.0 20210514 (Red Hat 8.5.0-3)]
Wed 18 Jan 10:22:52 GMT 2023
 
Thanks all for your replies and input. Indeed, I have no issues with ~100Mb/s, it goes up to 150 at times if I hold my breath. It's almost double my previous speeds using BT's wired fibre.

At the end of the day, I'm just a little peeved at not getting the same speeds on all my devices, regardless if these speeds are more than adequate for my needs. If Three shaped mobile devices to the same ~100Mb/s, I would not bat an eyelid and be none the wiser. Having 450Mb/s on mobile phones is kind of pointless anyways in my opinion.
 
I am with Three too, and just did the same test. 227Mb/s PC, 200 Mb/s Mobile.

"Also, I have now tested with Ubuntu (running on an Optiplex 3020 with a Gig ethernet port) and an Apple MacBook Pro 2017, again, same ~100Mb/s speeds."

My thought would be its the cable from between your high speed switch and the 5g router that is connecting at 100 Mb/s. Did you try all the above directly plugged into the router?

Did you try sharing out your phone internet as a hot spot?

Have you thought about testing the network speed itself? I did a quick Google and found Tutosoft LAN Speed Test v4.4. I have never used this, but from the look of it it will allow you to benchmark the speed of the network between Windows and Macbook.

Don't give up. I'm sure you'll find the answer!
 
Hey @fnanfne its me that replied to you on the Three Community forums :)

What Motherboard model is in the PC? I assume ASUS as searches for the AQC111C seem to reference ASUS and ASUS ROG a lot?

Does the Motherboard have 2 ethernet ports or just a single 5gbit port?

Are you running the latest Motherboard BIOS version?

Do you have the latest Motherboard chipset drivers installed?

Also it seems there's a firmware update tool for the AQC111C NIC (lots of various different issues reported online regarding this NIC btw) so might be worth updating both Motherboard BIOS and also the NIC firmware (I assume the NIC firmware update tool will be on the Motherboard vendors download section)...
 
WIndows 10. PC. Can I ask are you testing in a web browser with Speedtest or have you tried with the app you can download and install. I became aware of the app, after finding issues with speed testing through Firefox and Chrome. Through the app seems best way with WIN 10 PC (to me)
 
Sponsored Links
Also, I have now tested with Ubuntu (running on an Optiplex 3020 with a Gig ethernet port) and an Apple MacBook Pro 2017, again, same ~100Mb/s speeds.
Why not run a speedtest using iperf3 between the MacBook and the Optiplex which should confirm whether it's an internal nic/cable or router issue or something with Three?

Then repeat using the Netgear switch if it's internal in order to see if it's the router.
 
If this is a
Sorry missed you’d already done that.

Looks like there is an issue on the 5G router. Tell them it’s faulty and needs replacing. It’s their kit and their responsibility to ensure it works as advertised.

Clearly with two unrelated machines having the same behaviour it’s the common denominator.

I think you’ve done enough exhaustive testing now. 👍
If the Three supplied box is an NR5103E then I recall several folk on other threads having throughput related issues. WiFi was fine but the GbE and 2.5GbE ports were problematic.

They ended up getting a replacement router from Three. You just need to frame your complaint in the right way.

I’ve very fortunately never had an issue with mine. Though it’s not my main connection - and when initially bench tested with wired conn. the net throughout always was superior to WiFi and was well in excess of 100 Mbps throughout as tested with a Lenovo Yoga laptop with a Belkin GbE dongle and a MacBook Pro with a 10 GbE connection via an OWC dock.
 
I know it's a long shot but have you tried a different ethernet cable? I had a device that the RJ-45 cable had got slightly damaged and only on certain devices wouldn't let the gold contacts engage properly. The result was that the NIC downgraded the entire cable to 10M despite the NIC being capable of a lot more!

The fix was to file away the slight nick in the RJ-45 that was preventing the proper connection of that one pin/line!
 
Thanks all for your replies and input. Indeed, I have no issues with ~100Mb/s, it goes up to 150 at times if I hold my breath. It's almost double my previous speeds using BT's wired fibre.

At the end of the day, I'm just a little peeved at not getting the same speeds on all my devices, regardless if these speeds are more than adequate for my needs. If Three shaped mobile devices to the same ~100Mb/s, I would not bat an eyelid and be none the wiser. Having 450Mb/s on mobile phones is kind of pointless anyways in my opinion.
Three definitely are not shaping it.

I just tried on my NR5103E via WiFi:
- new-ish iPhone 14 with AX2 connection, 722 Mbps down
- 4-year old Lenovo ThinkPad with AC2 connection, 580 Mbps down

I'd try with a wired connection, but I've gone an lost my Belkin dongle....

Edit:
So plugged in old ThinkPad into dock with GbE connection. Wired into LAN1 of NR5103E....

2023-01-21 (1).webp
 
I am with Three too, and just did the same test. 227Mb/s PC, 200 Mb/s Mobile.

Thanks for confirming your speeds!

Did you try sharing out your phone internet as a hot spot?

Have you thought about testing the network speed itself? I did a quick Google and found Tutosoft LAN Speed Test v4.4. I have never used this, but from the look of it it will allow you to benchmark the speed of the network between Windows and Macbook.

Before getting Three's 5G router, I got a new 5G mobile contract I intended to use as a hotspot. I experienced the same slowish speeds and Three told me that this is normal speeds while "hotspotting". I then got the 5G router in the hopes of getting true speeds. Yes, I will do those test, well worth it!

My thought would be its the cable from between your high speed switch and the 5g router that is connecting at 100 Mb/s. Did you try all the above directly plugged into the router?

Indeed, only one CAT 6 cable plugged into my PC and the router.

Hey @fnanfne its me that replied to you on the Three Community forums :)

What Motherboard model is in the PC? I assume ASUS as searches for the AQC111C seem to reference ASUS and ASUS ROG a lot?

Does the Motherboard have 2 ethernet ports or just a single 5gbit port?

Are you running the latest Motherboard BIOS version?

Do you have the latest Motherboard chipset drivers installed?

Also it seems there's a firmware update tool for the AQC111C NIC (lots of various different issues reported online regarding this NIC btw) so might be worth updating both Motherboard BIOS and also the NIC firmware (I assume the NIC firmware update tool will be on the Motherboard vendors download section)...

Hey @crypt0ninja! Yup exactly, I have the ASUStek ROG Maximus XII Hero (Wi-Fi) motherboard. The mobo does have two NIC's, and I have tried both, but mainly trying the 5Gb/s one. I actually have not checked drivers for any other system components, it's defo worth a go thanks.

WIndows 10. PC. Can I ask are you testing in a web browser with Speedtest or have you tried with the app you can download and install. I became aware of the app, after finding issues with speed testing through Firefox and Chrome. Through the app seems best way with WIN 10 PC (to me)

I have been using speedtest.net so often, I started seeing an ad to try their Windows Store App, so I am using that at the moment. I was thinking of upgrading my Windows 10 to Windows 11 (since most of the issues have now been addressed) and wondering if this might make any difference? Just thinking about what @mikeliuk said before about Windows itself limiting network speeds.

Why not run a speedtest using iperf3 between the MacBook and the Optiplex which should confirm whether it's an internal nic/cable or router issue or something with Three?

Then repeat using the Netgear switch if it's internal in order to see if it's the router.

Good idea, I'll give that a go to get some baseline!

I know it's a long shot but have you tried a different ethernet cable? I had a device that the RJ-45 cable had got slightly damaged and only on certain devices wouldn't let the gold contacts engage properly. The result was that the NIC downgraded the entire cable to 10M despite the NIC being capable of a lot more!

The fix was to file away the slight nick in the RJ-45 that was preventing the proper connection of that one pin/line!

I have. At first I used an unlabelled ethernet cable, and it was suggested I try a new cable, so I dug up a brand new (still wrapped) CAT 6 cable, same slowish speeds. I have a CAT 7 cable, but I'm sure that's going to be overkill.

Three definitely are not shaping it.

I just tried on my NR5103E via WiFi:
- new-ish iPhone 14 with AX2 connection, 722 Mbps down
- 4-year old Lenovo ThinkPad with AC2 connection, 580 Mbps down

Awesome thanks for confirming your speeds!
 
Sponsored Links
@fnanfne - have you tried another wired connection to another computer in lieu of your Asus motherboard with the NR5103E router?

If you have and getting the same results the common denominator is now the 5G router (given you've swapped the ethernet cable).

Think we've all demonstrated that otherwise the Three network is well capable of full speeds using computers wired up to the 5103E
 
@fnanfne - have you tried another wired connection to another computer in lieu of your Asus motherboard with the NR5103E router?

If you have and getting the same results the common denominator is now the 5G router (given you've swapped the ethernet cable).

Think we've all demonstrated that otherwise the Three network is well capable of full speeds using computers wired up to the 5103E

No not yet, I will defo do some more fundamental ground work like this now and what others have suggested. Thanks to yours and @johnf 's feedback on your own speeds, I am now not suspicious of Three being up to any shenanigans anymore!

In addition to what has been suggested already, I have a few NAS which I can use to test the speed of my PC NIC's directly. I'll also use some other NIC's I can find, I know I have a USB one somewhere.

Will post back some results when I get home next week
 
No not yet, I will defo do some more fundamental ground work like this now and what others have suggested. Thanks to yours and @johnf 's feedback on your own speeds, I am now not suspicious of Three being up to any shenanigans anymore!

In addition to what has been suggested already, I have a few NAS which I can use to test the speed of my PC NIC's directly. I'll also use some other NIC's I can find, I know I have a USB one somewhere.

Will post back some results when I get home next week
It’s definitely not the network.

Either you have some very odd weirdness going on between the onboard NIC on your Asis mobo and the router wired ports or the router is simply duff.

My money is on the latter, given others experience with this model, but will await your testing with another wired device to conclude.
 
The issue impacts so many devices that I would personally try swapping out the router (USB tether to laptop might be equivalent if not bottlenecked too low). ethtool can confirm Ubuntu links at 1000 Mbps instead of 100 Mbps.

I happily got my Raspberry Pi 3 B up and running again hoping to test my Devolo Magic-2 network and found a 100 Mbps bottleneck with iperf3 because I was a dope for not realizing that Pi is USB bottlenecked internally. I did at least confirm the server (above speedtest and fastcom results) got ~940 Mbps to OPNsense.
 
Top
Cheap BIG ISPs for 100Mbps+
Community Fibre UK ISP Logo
150Mbps
Gift: None
Virgin Media UK ISP Logo
Virgin Media £22.99
132Mbps
Gift: None
Vodafone UK ISP Logo
Vodafone £24.00 - 26.00
150Mbps
Gift: None
NOW UK ISP Logo
NOW £24.00
100Mbps
Gift: None
Plusnet UK ISP Logo
Plusnet £25.99
145Mbps
Gift: £50 Reward Card
Large Availability | View All
Cheapest ISPs for 100Mbps+
Gigaclear UK ISP Logo
Gigaclear £17.00
200Mbps
Gift: None
Community Fibre UK ISP Logo
150Mbps
Gift: None
Virgin Media UK ISP Logo
Virgin Media £22.99
132Mbps
Gift: None
Hey! Broadband UK ISP Logo
150Mbps
Gift: None
Youfibre UK ISP Logo
Youfibre £23.99
150Mbps
Gift: None
Large Availability | View All
Sponsored Links
The Top 15 Category Tags
  1. FTTP (6028)
  2. BT (3639)
  3. Politics (2721)
  4. Business (2440)
  5. Openreach (2405)
  6. Building Digital UK (2330)
  7. Mobile Broadband (2146)
  8. FTTC (2083)
  9. Statistics (1902)
  10. 4G (1816)
  11. Virgin Media (1764)
  12. Ofcom Regulation (1582)
  13. Fibre Optic (1467)
  14. Wireless Internet (1462)
  15. 5G (1407)
Sponsored

Copyright © 1999 to Present - ISPreview.co.uk - All Rights Reserved - Terms  ,  Privacy and Cookie Policy  ,  Links  ,  Website Rules