Getting up to speed

A port became available at my 'waiting list' cabinet and my 6Mb ADSL is now a faster fibre connection. Zen said they'd put me on the lower 'up to' (I think it was) 38Mb service because its easier to upgrade later if wanted but you can't downgrade. I would like the faster service for the 20Mb upload.

According the the Openreach DSL checker for my phone number:

Clean - 66.7 to 45.7 down with 18 to 12 up and a 39.8 handback

Impacted - 55.2 to 35 down with 15.5 to 8.3 up and a 29.5 handback

My line ought to be clean, its a new estate with new wires, no ancient unknown extensions in the attic.

First I'm worried about disconnecting my router too much as I try to get the best speed while the line is under test over the first couple of weeks. Having said that...

I have my own FritzBox 7490 router and my initial download speed was 26.8 with 7.1 up (tested here at TBB). Zen's supplied FritzBox 3490 gave me 31.1 and 7.6 but a setting change has got my own router's speed in line with Zen's.

That was connected to the Mark 4 faceplate (what I bought off eBay on advice given here...) but when Zen's router is connected to the internal 'test' socket I get 36.6 down and 7.5 up.

So the slower connection is being caused by...? The eBay Mk4 faceplate? The extension wired into the faceplate (no phone attached)?

What I see is that I'm getting the worst possible speeds from an impacted line. Which I doubt I have.

As I say I don't want to disconnect and reconnect too often in the first couple of weeks to work out how to get the best speed I can but I'm willing to hear what you think.

Are you talking about sync rates or throughput speed test results ?

Any other stats available. ?

In reply to a post by Zarjaz:

Are you talking about sync rates or throughput speed test results ?

Speed test results from the ThinkBroadband speed test.

Any other stats available. ?

What would you like to know?

One thing I noticed is that the Zen router said was syncing at 40000 and 10000 when connected directly to the test socket. Using the faceplate gave a lower sync. Currently with my router it�s 34591 and 9997.

Edit: think those sync speeds are wrong, see next post...

Edited by Woolwich (Sat 02-Jun-18 19:47:36)

I know you can�t format here but here are my current stats...


Negotiated Connection Properties
Receive Direction Send Direction
Max. DSLAM throughput kbit/s 40000 10000
Min. DSLAM throughput kbit/s 128 128
Attainable throughput kbit/s 35001 12278
Current throughput kbit/s 34591 9997
Seamless rate adaptation off off

Latency fast fast
Impulse Noise Protection (INP) 42 0
G.INP on off

Signal-to-noise ratio dB 7 7
Bitswap on on
Line attenuation dB 21 32

Profile 17a
G.Vector off off

Carrier record J43 J43

There is no training period with fibre. DLM is always active. Disconnect it too often and DLM will move the speed down (it'll think there is a fault). From my own experience DLM has improved of late. Last year I was stuck in a range of about 66 - 72Mb/s for a long time drifting up and down over a period of a couple of months or so but of late it has moved me up to 77.2Mb/s. Just don't keep fiddling.

I do know with Plusnet, their fault replies include info on such things as whether you're line is impacted, disconnecting a lot etc. I'd imagine Zen do much the same. Just because it is a new estate doesn't mean that everything will always be squeaky clean. There are other factors to take into account.

And do try other speedtests sites - preferably wired.

Edited by 69bertie (Sat 02-Jun-18 20:37:03)

If you perform too many resyncs in a 24 hour period then DLM is likely to apply banding to the line. This can take many months to remove, sometimes never.

I strongly advise you to setup the modem how you like then leave it alone. There is no 10 day training period on FTTC. Despite BT quoting a training period to everyone DLM starts up on your line within the 1st 24 hours.

If you have a separate modem and router then you can do as much rebooting/fiddling with the router you like as DLM is only interested in the xDSL link between the modem and the cabinet.

You should not be looking at running speed tests to determine how your line is doing. There are so many variables (IP profile %'s, congestion, peering) that can affect throughput that it's just not a reliable guide.

You ideally want to login to your modem, or modem/router if using a combined unit, and looking up what your sync speed is.
You should really be getting 39,999/40,000Kbps if you are on the lower package and looking to upgrade.

Can you login to the modem and post the full connection/dsl statistics.
A description of what equipment is connected to what would also help.

Edited by j0hn83 (Sun 03-Jun-18 00:14:41)

Extension wiring often causes a speed loss, even with the filtered socket. Point is the MK4 is minimising that as much as possible, the difference in speed is likely this.

New build does not really make a huge difference. Often times, new builds are wired by builders who have no idea, they will run the BT lines alongside power cabling, and often run things however is easiest rather than the shortest route. They sometimes install fancy looking sockets, which they wire up poorly, often with the bell wires connected, and star wiring is not uncommon in new homes even to this date. With a new build things looks nice, however often it's cheap, nasty and poorly done below the surface.

Then with new builds, the developer is really responsible for providing anything over basic broadband. If the developer doesn't fund anything, then you kind of get what comes. New builds are often times quite far from the cabinet, and capacity can be limited at the closest cabinet, so it's not uncommon to end up on a cabinet much further away as this one had availability.

A new estate near me, half of it has internet speeds of 1Mbps. This is because the closest cabinet was at capacity, so instead they routed all the lines to a cabinet, which is in the opposite direction, and actually goes all the way back to a further away telephone exchange. The new estate effectively became the longest lines on that exchange. The FTTC is so far away, they still only see around 2Mbps on FTTC.

Anyway, who is your ISP, are you using the ISP router? If so, we can help get router stats.

Zen, judging from the first post, and there�s some stats from the Fritzbox posted later.

In reply to a post by j0hn83:

You should really be getting 39,999/40,000Kbps if you are on the lower package and looking to upgrade.

Can you login to the modem and post the full connection/dsl statistics.
A description of what equipment is connected to what would also help.

Yes, see above or below where I posted the router stats. I do have a 40000/10000 sync. I also see a drop in actual speed when using the Mk 4 faceplate rather than the test socket.

I'm using a FritzBox 7490 connected to the Master socket's Mk 4 faceplate. The router also has the phone line connected (because that model can do DECT). There's an internal extension running from the back of the Mk 4 faceplate to an unused but connected phone point.

In reply to a post by ukhardy07:

Often times, new builds are wired by builders who have no idea

I had them run cable for me. There's Cat 5 Ethernet running from the incoming Openreach point to my Master socket under the stair. The OR Engineer connected his cable to the Cat 5 and installed the Master. There's also another Cat 5 running from the Master back to the incoming position for an extension. OR Engineer moaned that he wasn't down to wire an extension but did anyway. But he did use a second pair in the 'incoming' Cat 5 rather than the second cable. So I have a Cat 5 pair connected to the exchange line going to the Master with a pair in that cable 'returning' to an extension. If that makes sense...

A new estate near me, half of it has internet speeds of 1Mbps.

You must live near me! Some have FTTC, others not because the cabinet is full. Many others are on a cabinet without fibre and have less than 3Mb ADSL. But the developer is funding a new fibre cabinet. Real Soon Now.

Anyway, who is your ISP, are you using the ISP router? If so, we can help get router stats.

Zen. See previous post with stats.

In reply to a post by Woolwich:

I also see a drop in actual speed when using the Mk 4 faceplate rather than the test socket.

Correct me if I'm wrong, but I believe that when you use the test socket it completely disconnects the extension wiring.

Hence there's likely a fault with that wiring; if you're not using it then simply disconnect it (and/or test it in isolation).

Negotiated Connection Properties
Receive Direction Send Direction
Max. DSLAM throughput kbit/s 40000 10000
Min. DSLAM throughput kbit/s 128 128
Attainable throughput kbit/s 35001 12278
Current throughput kbit/s 34591 9997
Seamless rate adaptation off off

Oh no you don't (panto fashion)

You have a DSLAM configuration capped at 40000/10000 but the modem is synced at 34591/9997 and it thinks that 35001/12278 might be possible, but during negotiation phase (sync time) it did not manage to get that (reasons can be variations in noise over time)

In reply to a post by MrSaffron:

You have a DSLAM configuration capped at 40000/10000 but the modem is synced at 34591/9997 and it thinks that 35001/12278 might be possible, but during negotiation phase (sync time) it did not manage to get that (reasons can be variations in noise over time)

Ah, yes, I see.

So does this look like a line which is never going to achieve more than 35000 down? (Which according to my DSL Checker result is the worst possible case.)

Maybe a silly question, but have you checked which pair of wires in your Cat5e he used from the dropwire junction to the master and which pair for the return.

Always a possibility that they may be split, or not fully terminated at some point.

In reply to a post by MHC:

Maybe a silly question, but have you checked which pair of wires in your Cat5e he used from the dropwire junction to the master and which pair for the return.

Always a possibility that they may be split, or not fully terminated at some point.

I changed the front plate for a Mk4 so its always possible I put the wrong cable in the wrong connector. In fact I did because on the first attempt I had no dial tone. But if anyone can tell me which connection goes to which I could confirm if its correct.

Get some picture of what you have, host them somewhere and provide a link.

Until you confirm what the current wiring state is then its impossible to say.

We've seen people with bad wiring go from slower than you to well past the 40 Mbps mark, so totally confirming that your wiring is actually correct is important.

OK, here are some pictures of my wires

http://i63.tinypic.com/mai7gk.jpg This is the back of the extension point - are the correct cables in the correct positions?

http://i67.tinypic.com/4iiqg8.jpg The incoming OR cable and connections which are behind the extension point. You can see the blue OR cable pair connected with the blue pair in the Ethernet which runs to the OR Master socket under the stairs. The orange OR pair are not connected. The orange Ethernet cable is the return from the Master socket and feeds the extension. There's a bundle of cables from the second Ethernet which is not in use.

http://i65.tinypic.com/10wtso6.jpg At the Master socket where I have used my own (eBay) Mk4 faceplate. Are the cables in the correct connectors?

Meanwhile, I've connected the router directly to the 'test' socket. The router reports a better speed but its much the same when doing a TBB speed test.


Receive Direction Send Direction
Max. DSLAM throughput kbit/s 40000 10000
Min. DSLAM throughput kbit/s 128 128
Attainable throughput kbit/s 50537 14906
Current throughput kbit/s 39998 9997
Seamless rate adaptation off off

Latency fast fast
Impulse Noise Protection (INP) 51 0
G.INP on off

Signal-to-noise ratio dB 9 9
Bitswap on off
Line attenuation dB 21 30

Profile 17a
G.Vector off off

Carrier record A43 A43

I haven't seen an attainable throughput as high as this before. But the speed is still the same...

Date Tue 05/06/2018 11:10
Down HTTPx6 31.05 Mbps
Down HTTPx1 35.15 Mbps
Upstream 7.69 Mbps
ISP Zen Internet

So the router is happier in the test socket but it doesn't actually make any difference?

Edited by Woolwich (Tue 05-Jun-18 12:00:08)

Wires look ok

But the Mk4 faceplate is it the one with a DSL and phone socket or just a phone socket?

If just a phone socket then the orange pair represent a bridge tap, hence why speeds are better with the test socket exposed.

To mitigate the effect

1) Disconnect the orange pair
or
2) Fit a MK4 VDSL2 faceplate

In reply to a post by MrSaffron:

But the Mk4 faceplate is it the one with a DSL and phone socket or just a phone socket?

The one with both DSL and phone. Got it here https://www.ebay.co.uk/itm/BN-2018-Openreach-Telepho...

One thing which might improve it slightly - keep the twists in the pairs for as long as possible. Even a few inches of untwisted wires is enough to pick up noise. You will also note that each pair, is twisted at a different rate, that is to further minimise cross talk between them and I would twist together the brown and green pairs at both ends - in can sometimes make a difference.

As its the vdsl filter (if memory serves me right) there will be two connector less blocks

The A/B which is unfiltered, hence why removing faceplate improves things
A 2/3/5 one which is where the orange wires should go 2 and 5

In reply to a post by MrSaffron:

The A/B which is unfiltered, hence why removing faceplate improves things
A 2/3/5 one which is where the orange wires should go 2 and 5

Thanks, I'll try that shortly. Does it matter which orange wire goes to which connector? If it does I have a 50:50 chance...

With the extension line connected to the A/B block I had a clear line. No noise on the quiet line test. If it was unfiltered shouldn't I have been hearing broadband noise?

As the broadband is well outside the frequencies you can hear likely to have not heard anything

Also if you had a phone plugged in at the extension without a microfilter that would increase the problems with the connection

Which orange of the pair does not matter

In reply to a post by MrSaffron:

As the broadband is well outside the frequencies you can hear likely to have not heard anything

Slightly confused now!

I thought I had bought a Mk4 faceplate in order to do away with additional filters. I thought I needed to connect the extension to the Mk4 faceplate and then that would be suitable for a voice handset.

But actually the connectors on the Mk4 are for a data extension?

And the connectors on the 5C 'baseplate' for extension wiring are filtered for voice?

In otherwords I already had a filtered Openreach box, I was just buying a more convenient two port version, one for voice and one for DSL..

The aim is to have the FritzBox modem/router connect to the DSL and, to use it for DECT, its phone line to the phone point. Both on the Mk5 faceplate. Then there's a phone extension point in another room which will not have a handset attached but I do want for possible future use.

Currently I have

Receive Direction Send Direction
Attainable throughput kbit/s 50254 15338
Current throughput kbit/s 39998 9997

If I can get a 15Mb upload by moving to the 'faster' fibre package it would be worthwhiile for me. Lets see what happens when I reconnect the Mk4 faceplate - wired as advised!

I have retrieved the 5C I have

The telephone extension if you want it filtered should be connected be connected to 2 and 5 on the Master Socket 5C piece of plastic, i.e. the part that is screwed to the wall box.

The removable VDSL2 faceplate and its A/B connector is where you connect any data extensions.

So you connected the extension to the wrong connector, presume there was no instructions.

The 2/3/5 connector is filtered by the faceplate and when the faceplate is removed is disabled.

In reply to a post by MrSaffron:

presume there was no instructions.

Perfectly correct.

OK wired up properly with extension line working and Mk4 faceplate back on.

A drop in attainable

Attainable throughput kbit/s 48394 14880
Current throughput kbit/s 39998 9997

Maybe later I'll disconnect the extension completely then we'll know if its that wiring or the Mk4 faceplate.

A drop but the upload which you were wanting to chase is holding up, and with maximum sync currently there is scope to upgrade.

NOTE: Upgrading might see the line vary speeds more i.e. be less stable as you are likely to be utilising more marginal bits of the signal. So I'd advise once you have finally finished playing to wait a few weeks and confirm things are stable.