Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[tubaman]

....

So plug the modem/router directly into the test socket on the master socket, turn off Wi-Fi, disconnect any network cables, turn off WiFi on your laptop and connect it via a network cable. So now you have only the laptop connected you can run speed tests, but make sure nothing on the laptop is downloading or uploading as this will affect the results.

The logs /stats we need to see are what the modems connecting at, not access point logs.

We can see from the logs provided that the Hub 6 has been up for over 5 days with a connection speed of 63/14. If the line connection was poor I think DLM would have taken action in that time and it clearly has not.
Agree that any speed tests must be run wired to ensure consistency.
If wired checks are consistently ok then you need to look at the wireless config. Although the BT Hubs claim to have 'smart wireless' - ie pick the best channels themselves - I don't believe this always works well in practice, and here are plenty of free wireless analysis apps out there that can quickly tell you which are the best channels to use.
If the wired checks are inconsistent then this is perhaps pointing at congestion on the PlusNet network, and I'm not sure what you can do about that one.
 

[Ronski]

Thanks Tubeman, I was on my tablet earlier and must have missed the logs/stats from the hub 

[guest_10]

Hi,

Thanks for the responses

So to provide a degree of clarification on the matter.

There appear to be two issues that I am experiencing.

The first and most annoying is the internet speed when a device is connected via WiFi is reading less than 10meg consistently. This is the case no matter what transmitter I am connected to. So when I'm connected to the main router less than 2 metres between the router and laptop with no obstructions the speed is less than 6meg and it is the case whether using the Plusnet hub one or the BT Smart Hub 6. The same is true if I'm connected to one of the access points which is connected to the main router via CAT 6.

The second issue, is that the internet drops out regularly and this occurs on devices, whether they are connected via WiFi or wired. I know the dropouts occur on the wired devices because the light on the access point turns orange which indicates there is no internet. Also devices that are wired lose their internet connection randomly. When speaking to Plusnet about this, the problem does not appear on their logs, An engineer has been out and tested the line and found no problems. This is why I attached the logs for the access points because I would have thought they should show the dropouts that have occurred.

The speed tests I have run are varied but all use the same device (a laptop). They are as follows:

When testing directly through the master socket via a cable the download speed I get is 68meg, this was shown to me by the BT engineer and repeated on my own device.

When the router gets plugged in and the test is done via cable with no other devices linked to the router the download speed drops to 58meg.
The next test was to connect the device via a switch and the download speed again was 58meg

I removed the switch and then linked access point 1 via CAT 6 and run a test again via cable and the speed is also 58meg, disconnected access point 1 and repeated the test for access point 2 again no other devices connected and the speed is consistent 58meg.

I then connected all the wired devices to the router and ran the tests again via cable and the speeds came back as 58meg on all connection points. Based on the wired tests there is nothing wrong with my internal network.

I repeated the tests 3 or 4 times to ensure they weren't blips and the results were consistent.

The next set of tests involved repeating the tests for the router and access points but using the WiFi and always situated less than 2 metres from the router or access points with no obstruction. Same procedure, testing with no other devices attached. Prior to testing I used the WiFi analyser app on android to check the WiFi signal and optimised the channels each time I added WiFi access points. The WiFi Analyser showed nothing unusual with the signal. The speed I got when testing via WiFi was totally different the download speed was less than 6 meg and the upload speed was over 10meg The upload speed is the same as what I was getting when wired. I check the WiFi signal strength and it always rates as excellent.

Again the tests on the WiFi were repeated and they are consistently poor download speeds.

Before switching to Plusnet the above tests were done and the download speeds were always towards the top end around 50 - 60meg on both WiFi and Wired. There is nothing affecting the WiFi connection internally because that would have appeared on my WiFi Analyser. Based on my tests the problem is with Plusnet but I don't know what they are doing to affect the speed coming out of the WiFi.

As for the internet dropping out, there is no way for me to test this.

[j0hn]

There is absolutely zero I can think of that changing to Plusnet would suddenly cause issues on WiFi.

If wired speeds are fine (they are) and the line remains in sync (it does) that shouts out internal hardware to me.

The fact that you also experience this on the BT Homehub which worked with BT leaves me at a loss.

Changing ISP and using the same modem/router can't suddenly cause WiFi issues. This is not a Plusnet issue.

The connection completely dropping is another matter.
No idea if you can set up a BQM on either of those ISP devices.

[ejs]

Have a few different wifi channels actually been tried?

There are plenty of things that can affect wifi but will not show on a wifi analyser app. Personally I think most wifi apps are largely worthless and may well be worse than simply trying out different channels. Most wifi analyser apps just show the signal strength from each AP and what channel they are on. They don't show how heavily used each network is, a channel with one busy network will be a lot busier than a channel with several mostly idle networks. They don't show the strength of the signals from any of the devices connected to the network, just the access point. They don't show non-wifi signals / interference on the frequencies. Some people might then argue that a wifi analyser app is still better than nothing, but even that is questionable. An automatic channel selection system on an access point should have more data available to it than what a wifi analyser app has.

[Ronski]

EJS beat me to it.

When testing directly through the master socket via a cable the download speed I get is 68meg, this was shown to me by the BT engineer and repeated on my own device.

When the router gets plugged in and the test is done via cable with no other devices linked to the router the download speed drops to 58meg.

I'm rather confused by the above two  statements, to test direct through the master socket surely you're using your router? The engineer would of used his own equipment, which you wouldn't have been able to plug into. So statement two surely using the same equipment as statement 1, but something must be different to get the different results.

In statement two are you still using the test socket? This means removing the front of the master socket and using the socket behind? When you remove the front of the master socket it should isolate all other sockets in the house, they should no longer work - if they do this will be affecting things. Hopefully the engineer would of checked this.

Prior to testing I used the WiFi analyser app on android to check the WiFi signal and optimised the channels each time I added WiFi access points. The WiFi Analyser showed nothing unusual with the signal. The speed I got when testing via WiFi was totally different the download speed was less than 6 meg and the upload speed was over 10meg The upload speed is the same as what I was getting when wired. I check the WiFi signal strength and it always rates as excellent.

......

 There is nothing affecting the WiFi connection internally because that would have appeared on my WiFi Analyser. Based on my tests the problem is with Plusnet but I don't know what they are doing to affect the speed coming out of the WiFi.

Unfortunately a wi-fi analyser app on Android is not going to show things that could be affecting your wi-fi, you need dedicated hardware to show that. For instance get the app on your phone, go and microwave something and see if it shows on the app. What you need is something like the Wi-Spy (very expensive) which will actually display the signal from things like microwaves, baby monitors, remote control transmitters etc. These could all be affecting you're wi-fi, especially 2.4Ghz. Are you using 2.4Ghz or 5Ghz?

 Watch this video, especially from 6:28
https://www.youtube.com/watch?v=tr0AfBO1O20

One thing I'm certain of is Plusnet are not slowing your broadband just when you use wi-fi.

As for the internet dropping out, there is no way for me to test this.

Well there is, but you'd need a modem that allows you to collect stats.

[tubaman]

The only reason I can think for the wi-fi being so slow is that it is set to the wrong protocol - you haven't accidentally managed to set it to 802.11b as that would run at about 10Mbps maximum?
 

[boozy]

One of the easiest ways to solve wifi problems is to separate the 2.4 and 5G (I suffix the network name with the frequency).  I put all the stuff I don't want cluttering 5G on 2.4 (kids things, Alexa's, cameras).  The things I want to run faster won't know about the 2.4Ghz networks and won't fail over to them when the signal is poor and visa versa for the things allocated to 2.4Ghz.  I have essentially the same as you with two access points at either end of the house, so there's always a 5Ghz and 2.4Ghz available.  Laptops and mine and the boss' phones get 5Ghz and most everything else gets 2.4Ghz.  Everything in a fixed position with an ethernet port gets a wired connection (although senior management will kill me if I put another hole in the ceiling, so I'm going to have to get creative with the infrastructure cabling).

With 4 networks being available you can deal with any internal congestion easily e.g. changing channels or promoting a device to the 5Ghz network (although I've not needed to).  The only problem is when you still have a 5Ghz connection to the other end of the house - but that's fairly rare as the internal walls are brick and flicking wifi off an on sorts this.  If you are dual banding the APs  (like most come out of the box), you have no idea what's going on without logging in to them and checking.

[snadge]

My 2 pence worth:

SMART WI-FI
I can confirm that SMART WI-FI (auto channel swap) is poor in my experience, every time I've tested such a feature across various routers on various connections (not just my own - which I've tested many routers on) it seems to always pick one that doesn't appear optimal? (i.e. switching to an active channel when there are clear channels on 1, 6 or 11 on the 2.4Ghz spectrum). I know someone has already mentioned that the router/hub may have more information when choosing a channel, but the only thing I can think of is RF Spectrum Analysis and I may be wrong but aren't these types of devices expensive? and therefore unlikely to be installed in a cheap ISP router?

PLUSNET HUB ONE
I will also add (even though its probably irrelevant at this point) that the Plusnet Hub One is absolutely DIRE in my honest opinion, I had 2 from Plusnet as the first would drop its throughput from 36Mb to 15-25Mb every 4-5 days and only a reboot would fix it, the 2nd Plusnet Hub One done the exact same.. I resolved this by using my own.

THROTTLING
Your 100% not being "throttled" by your ISP, there is no way for an ISP to manipulate what goes on in your internal network, also what gain is there for Plusnet to pump 65Mb down to the router and then to throttle it and cause you issues at your end? it just doesn't make sense and can not and does not happen.

NTE5C+Mk4 SSFP
There are loads of bad reviews about these, I myself just 3 days ago tested various SSFP's/setups including the one you have even after hearing this as I wanted to find out for myself, I had loss of speed (2MB) and errors went up 10 fold (and strange readings from DSL-stats), you would have been better leaving the NTE5A and Mk2 SSFP in to be honest, its looks nice, but is cheaply made (no circuit boards, just metal strips) in comparison to the older one

CONCLUSION
After reading your thread I have to say that it sounds like its an internal configuration issue or the device(s) your testing have issues. You mention having switches and routers etc, maybe if you could supply more information about your setup and the devices connecting to it?
What routers/switches do you have?
How are they connected?
How are they configured?
Same again for devices.
- sometimes a mockup drawing in paint or a photo can help.

[ejs]

SMART WI-FI
I can confirm that SMART WI-FI (auto channel swap) is poor in my experience, every time I've tested such a feature across various routers on various connections (not just my own - which I've tested many routers on) it seems to always pick one that doesn't appear optimal? (i.e. switching to an active channel when there are clear channels on 1, 6 or 11 on the 2.4Ghz spectrum). I know someone has already mentioned that the router/hub may have more information when choosing a channel, but the only thing I can think of is RF Spectrum Analysis and I may be wrong but aren't these types of devices expensive? and therefore unlikely to be installed in a cheap ISP router?

The automatic channel selection will probably be software provided by the wifi chip manufacturer, which will have low level access to the wifi hardware. This will be able to measure the amount of signal and the amount of noise on each channel, and measure the amount of data traffic rather than merely count the number of wireless networks. There's nothing wrong with being on the same channel as another network anyway. You claim some channel with no networks on is the optimal one, but how could you possibly know it is? Maybe all the routers are avoiding it for good reason, if they can see a lot of non-wifi noise which you won't see on a wifi analyser app.

[snadge]

The automatic channel selection will probably be software provided by the wifi chip manufacturer, which will have low level access to the wifi hardware. This will be able to measure the amount of signal and the amount of noise on each channel, and measure the amount of data traffic rather than merely count the number of wireless networks. There's nothing wrong with being on the same channel as another network anyway. You claim some channel with no networks on is the optimal one, but how could you possibly know it is? Maybe all the routers are avoiding it for good reason, if they can see a lot of non-wifi noise which you won't see on a wifi analyser app.

Actually, now I think on I remember when I set up my new Zyxel 8924 the wi-fi page had a channel scan feature, the screen below in my router seems to confirm what you say about Low-Level 'Channel Noise' detection...

[burakkucat]

Actually, now I think on I remember when I set up my new Zyxel 8924 the wi-fi page had a channel scan feature, . . .

All the ZyXEL VMGnnnn CPEs have that feature. 

[vultura]

NTE5C+Mk4 SSFP
There are loads of bad reviews about these, I myself just 3 days ago tested various SSFP's/setups including the one you have even after hearing this as I wanted to find out for myself, I had loss of speed (2MB) and errors went up 10 fold (and strange readings from DSL-stats), you would have been better leaving the NTE5A and Mk2 SSFP in to be honest, its looks nice, but is cheaply made (no circuit boards, just metal strips) in comparison to the older one

A bit off-topic, but are the NTE5C & Master Socket 5C the same thing?

[Ronski]

@Vultura, welcome to the forums. Yes, the NTE5 is the master socket, and there are multiple versions of it with the NTE5C being the latest I believe.

[vultura]

@Vultura, welcome to the forums. Yes, the NTE5 is the master socket, and there are multiple versions of it with the NTE5C being the latest I believe.

Thanks for reply and the welcome, from such a well-known name too, I feel honoured.