Deterioration in speed and stability since 31 Dec - Watton at Stone SG14

[#1]

@Prologica  I realise you are not able to run an ethernet connected speedtest at present but I just wanted to point out that the BT wholesale test you did over wi-fi is showing an impossible result. Whether it is because it was done over wi-fi or just the tester playing up I don't know. It isn't too good for me and always reports an upload speed of zero!

Looking at the result it is reporting an IP profile of 14.51 Mbps,  a throughput speed cap based on percentage of your sync speed. Your throughput speed can not exceed this value so the 19.85Mbps result is higher than any sync speed you have reported, also the test is saying you could get up to 21Mbps, again impossibly high.

Your sync speed is good but as @Townman has said it seems to be changing, is this because you are restarting the router or is the connection dropping and resyncing?

In my experience Wi-Fi is very unpredictable and speeds are often good one minute and poor the next. To get good results one or more wi-fi access points connected to the router by ethernet is the best solution. I am not a fan of powerline connections but they can be an easier answer.

[#2]

Hi @RealAleMadrid  I agree with you about the wholesale test result. I’ve just run it again three times in quick succession over Wi-fi and download speeds it showed ranged from 13-53 mpbs!

[#3]

@RealAleMadrid  @Prologica 

This might interest you  https://community.plus.net/t5/Everything-else/BTw-Speedtest-Gone-Bananas/m-p/1789174#M25389

[#4]

@Prologica 

Your service is in the right ball park.  One can expect a drop off in performance of ADSL during the hours of darkness, arsing from ADSL transmission radio interferance from EU countries.

I fear that to take the investigation of this scenario further, you need to look at finding a different router which can have the xDSL stats dynamically monitored. That is one that has not had Telnet access deliberately "obstructed" by BT in their attempt to prevent users from measuring the performance of the delivered service.  From that perspective, all of the BT supplied routers provided by Plusnet since the TG582n was replaced have been pigs for being monitored.

The Plusnet Hub One (based on the BT HH5A) does lend itself to being part monitored by RouterStats (see the link below) but the BTHH6 does not.  Given that you are on ADSL, most old routers (listed on the router stats page) should work reasonably well for you and can be picked up on fleabay for very little.

[#5]

---

@Townman wrote:

Face plate - that ought not to case any issues, but it has been known for tarnished contacts on the faceplate plug to cause some issues.  See if they might be cleaned.

At the wrong end of the house - have you had the router located somewhere else?  If yes what kind of extension cable where you using?  Many extension cables will work fine for a telephone, but will totally trash broadband performance.

 @Baldrick1 - would you know if BT HH6 reports USABLE sync speed (net of error rate?) rather than the actual sync speed?

I would turn some attention to inspecting your wifi spectrum to see how many other wifi stations your router is "fighting" with.  Insidder is a useful free tool for this.

---

A few observations:

There is so much detail that my head hurts so please excuse me if the following has already been considered:

1. I have read of reports of the front part of the latest clip on type of master socket not seating and connecting properly, especially if the back box is screwed tightly to an uneven surface resulting in distortion. I would suggest that the hub is located at the master socket, a filtered faceplate fitted and the extensions fed from the filtered output, removing the need for the dangly filters

2. I interpret that the extension wiring is connected to the master socket and dangly filters (I assume) are used with all used outlets through the property. Therefore before anything else all testing should be done from the Test socket.

3. I'm not sure if the inconsistencies in the router statistics is a root issue. (Sorry @Townman  I don't know if BT HH6 reports usable sync speed (net of error rate?) rather than the actual sync speed.) I suspect that it's more a problem of getting data from one end of the property to the other.

4. Using Powerline can be problematic, especially if the two sockets are on different rings see https://community.plus.net/t5/Tech-Help-Software-Hardware-etc/Powerline-Unit-Performance/m-p/1607450.... I know that installing Ethernet cable can be a pain but with reasonable DIY skill, cable can be hidden.

5. If using wireless it's always sensible to us a WiFi sniffer, I use WiFi Analyzer on and Android phone to monitor other networks. Also don't rely on the router automatic channel selector. I get totally different results at the front, back and router location in my property so it's a case of finding the best compromise.

[#6]

Hi @Baldrick1 

Some useful points.  I too have concerns about the extension wiring.  Given that it is not connected via the faceplate, we cannot be sure that it is not a contributory factor.  It might well have been connected by the "BT" engineer - by were they a BT engineer or one of their questionable subcontractors?

Whilst connecting extension wiring in the back of the box might well work, such totally defeats the whole purpose of the NTU's purpose which is to isolate everything internal from the phone line.  The presence of a not by the face plate connection leaves the risk that the issue is caused by the internal wiring, especially if it is not current standard twisted pairs.

"3. I'm not sure if the inconsistencies in the router statistics is a root issue. (Sorry @Townman  I don't know if BT HH6 reports usable sync speed (net of error rate?) rather than the actual sync speed.) I suspect that it's more a problem of getting data from one end of the property to the other."

You have trodden in the the quagmire which so concerns me - you mention SYNC speed and net of error rate - the router reports the speeds as DATA speed.

With ADSL data speed is at best 88.2% of sync speed.

Max SYNC rate is the speed which might be attained it there were no errors on the line.  The BT HH6 is reporting a variety of figures which cannot be logically correlated.  Current speed and Max speed should not discernible relationship to the link establishment speed.  To the best of my knowledge in the UK we do not supply ADSL as rate adaptive.

[#7]

@Townman  In message #14 the OP states that the master socket has a clip on faceplate, this is the latest NTE5C type, there are no connections provided on the faceplate. Any extensions are connected to dedicated connections on the main part of the master socket. They are not connected to the incoming line until the faceplate is plugged in, there are bridging links on the faceplate that do this. The end result is the same as the older faceplate with extensions connected to it. If you are interested there are some pictures on this thread (message #27)

https://community.plus.net/t5/My-Router/replacement-router/td-p/1787875/page/2 

[#8]

RAM,

Thank you for the clarification - I’ve not had a NTU5c in my sticky mits. Will not make that mistake again!!

So it is entirely possible that the extension wiring might be isolated by removing the faceplate and could, with the faceplate in place impact the line’s performance.

[#9]

@Townman 

I take your point but my reference to sync speed was a direct copy and paste from your request for data.

@RealAleMadrid 

I'm not sure that it can be assumed that the master socket is correctly wired. We do not currently know if some bodger of a technician has wired up the extensions in parallel with the incoming connection rather than doing it correctly. @Prologica do the extensions work if you take off the master socket faceplate?

[#10]

@Baldrick1   I should have qualified my description with "if the master socket has been correctly wired" 🙂

@Townman  If a dual socket filtered faceplate is used the extensions are supplied by the filtered signal, a single socket faceplate connects the incoming line to all the sockets unfiltered which as you suggest could have a major impact on the line speeds.

[#11]

@Baldrick1 @Townman @RealAleMadrid 

Hi guys,

I just checked with the faceplate off. No dial tone on phone downstairs so sockets depend on the faceplate making the connection. 

Running ethernet is not a great option - would be long runs over two floors...maybe the next house! 

The new AV plugs did resolve the intermittent connection to the Humax box.

I’ll try a wired connection to my MacBook when the adaptor arrives at the weekend, both at the router and downstairs via an AV plug.

I’ve got a Zoom meeting tonight so I’ll check how stable the connection is on that (via WiFi)

i see what what you mean Baldrick1 re the wholesale speedtest on iPad. I got very different results with consecutive tests on iPad and Android. Android was far more realistic.

[#12]

@Prologica 

To get better WiFi at the far end of your property you could try configuring your 2704N as a Wireless Access Point and connect this to your Powerline unit via a Ethernet cable. To do this, connect to your 2704N in isolation, that is disconnected from the Internet. Make sure that the Hub 6 wireless is off as this might cause confusion.

Log into the special set up pages by putting http://192.168.1.254/expert_user.html into the address field of your browser. Search through the options and set DHCP to OFF, change the IP address to something within the fixed range of the Hub 6 such as 192.168.1.25 and that's it. Note that you will lose connection to it as soon as you change this as it's moved to the new IP address.

If you screw up or want to return to the standard settings simply push a paperclip or similar into the reset hole and keep the switch closed for 30 seconds or so. If you want more detailed instructions please ask and I will dig mine out of the junk box and post some screen shots.

Having configured your WAP connect one of the LAN sockets to your Powerline unit at the 'living' end of your property. This will then give you the 2704N wireless channel plus 3 spare Ethernet sockets for your use at the right end of the house.

[#13]

@Baldrick1 @RealAleMadrid @Townman 

Baldrick1 had previously advised not using auto channel allocation so when I split the bands I allocated ch 1 for 2.4ghz and 36 for 5ghz. I’ve just used WI-FI analyser and i can see are no other users of those channels in the various parts of the house.

I think possibly the poor fitting of the faceplate may have been the main issue as reseating that yesterday gave an immediate improvement over WI-FI standing at the router. The fall off in other parts of the house is no doubt due to obstructions so in the absence of Ethernet runs, I will have to play with AV plugs/ extender placement - We tend to use one part of the house in the day and the other part in the evening.

 I have the original Homeplugs (c.2008 vintage) I took off the Humax box. I’m not sure if they might still work ok for a wired Ethernet to MacBook connection. I believe they’re a different standard than the new TPlink 400s so will hopefully coexist on the ring main. It’s unavoidable that the pairs have to run across different circuits (different circuit breakers on the consumer unit) but the TPLink pair are working just fine now for catch up TV. 

[#14]

@Prologica See my last posting, this might be the answer to your problem.

[#15]

@Baldrick1  sorry Baldrick1 our posts crossed. Thanks yes I’m aware of possibly using the 2704n as an access point. I didn’t do that yet as I read conflicting info about whether it was effective or not...and I was dubious in case I needed to contact Plusnet again with an issue...resetting back to router, setting up credentials again etc.

but I’m happy to give it a go.