Diaries of a Pulsator

We migrated to Pulse8 fibre a week ago from Sky ADSL / BT phone line

On the day of the migration, we seemed to lose incoming calls on the phone line, as the Kelly engineer who was running late, couldn't call us to tell us he was running late. So our morning install appointment commenced at 1:15.

The engineer didn't have the vDSL faceplate for the new NTE5 boxes so fitted the mk3 interstitial faceplate to the existing NTE5.

After a minor glitch all was up and running, at least on the BB. The engineer expedited the phone migration, or at least half of it, so we had outgoing calls but not incoming calls until some time in the early hours of the following morning when the bots must have completed their stuff.

How's the connection?

Pretty good. The fibre cabinet is the other side of our front garden wall, so it was just about guaranteed that we would get full 76/19 sync.

I've started to run the TBB Quality Monitor and the connection looks pretty good (see here).

TBB speed test speeds are a bit variable (different browsers / hardware and the flash and non-flash versions give slightly different results), but obviously much quicker than the Sky ADSL connection.

Latest test here.

The results tend to less linear in the evenings though.

Pulse8 have been great throughout the process.

Hey jaydub - can you post a link to your live BQM graph?

And maybe another speed test result...

Pulse8 are at the top of my shortlist.

Yep I would like to see a speedtest at peak hours, that looks like possible congestion at 09:05AM. Are you using ethernet when testing?

I am running the quality monitor, but probably shouldn't be as I am not on a fixed IP.

Just started one on my current IP address, so give it time.

http://www.thinkbroadband.com/ping/share/2f79faab44e...

All speed tests give pretty much the same result, night or day. I tend to run the speed tests at midnight or first thing in the morning, so here's a couple:

Mid-evening on Monday
http://www.thinkbroadband.com/speedtest/results.html...

Yesterday morning
http://www.thinkbroadband.com/speedtest/results.html...

TestMyNet speeds are in the same ball park as the single thread results.

I have been in dialogue with Sam about it and he has reported that the only issue that is being returned is the connection to the router, however it is very vague as it could be internal or external. I've moved the connection to the master socket and both his results and mine are no different.

A router swap would be the next stage, but I'm slightly loathe to go out and buy another to prove it's not the router. So unless there is someone within spitting distance of Marple who wants to loan a VDSL router to an almost complete stranger....

The problem is that I ace the Pulse8 Ookla speed test which is a multithread test (and possibly running on port:8080), so the line is well above the threshold speed where they can raise a fault.

I may just jump to their business product for a few days (if they can offer that) to see if it improves things, which should hopefully tell us whether it is congestion or not.

Please note that I had similar speed issues on Sky ADSL. Very similar day and night, so my suspicion is it's not specific to the Pulse8 TTB connection.

In reply to a post by jaydub:

Please note that I had similar speed issues on Sky ADSL. Very similar day and night, so my suspicion is it's not specific to the Pulse8 TTB connection.

Test wired or wireless?

I did not see that kind of speedtest on my Sky, TalkTalk or BT connection (apart from when Sky did genuinely have congestion). Are the Sky tests wireless (they have 2.4Ghz WiFi on most routers)?

Here are a couple of mine that I have saved down, they're like this 24/7:

http://www.thinkbroadband.com/speedtest/results.html...
http://www.thinkbroadband.com/speedtest/results.html...

My TalkTalk line is also the same.
TBH the single threaded speed is absolutely fine to do everything e.g. stream, download at a decent speed, skype call, facetime etc so I would not worry too much. It is when it drops to 3Mbps that it becomes an issue (which does happen with bad congestion!!)

EDIT: I also test over WiFi using default ISP supplied router and a macbook pro 2014 retina.

Edited by ukhardy07 (Fri 18-Nov-16 11:38:15)

I am always running the quality monitor (see sig).

Speed tests can vary somewhat but I rarely see the x1 match the x6 test.

This morning
Speed Test Result - 18/11/2016

Spoken to Sam and Adam and moved up to their business package about two weeks ago to see if it helped, made no difference from what I can see. Have tried two Asus routers, dedicated pfSense box and now running an Edge Router Lite, all with same result. All tests with direct connection to the master socket with a Mk 3b faceplate, hard wired not wifi.

The Pulse8 Ookla speed test shows 64mb (same as the x6 test here) as it is multithreaded. Modem line rate right now is 68651 down and 20000 up. FEC total stands at 3,100 and HEC or 3,000, BT OR modem has been up for 5 days.

Based on my performance on the residential package compared to the business one I would put this down to congestion.

In reply to a post by RobertoS:

In reply to a post by jaydub:

Please note that I had similar speed issues on Sky ADSL. Very similar day and night, so my suspicion is it's not specific to the Pulse8 TTB connection.

Test wired or wireless?

Wired

The fact that the FTTC is "the other side of our front garden wall" does not guarantee maximum speed.

The local FTTC here is probably a similar distance, about 10 Metres from my front door; but the link cables to the PCP travel about 50 Metres to the PCP, followed by about 250 M of original phone cabling, to get to my front door, still only 10 M from the FTTC.

(I can see inside the FTTC when opened, from my lounge window.)

I forget the site which gives forecast BB speeds; but when I used it to extract the forecast speeds for all 77 houses on this estate, it showed that my house and the others in the immediate group, were likely to get the slowest BB speed, whilst geographically the nearest to the FTTC; and second nearest geographically to the PCP.

The most remote houses on the estate, about 500 M from the PCP, are forecast to get faster speeds, although not maximum.

The estate was fully phone-wired during build on a greenfield site in 1967, so no obvious reason for the extended wiring.

So the geographic locations of the FTTC and of the PCP are not reliable indicators of likely Broadband performance.

I am always running the quality monitor (see sig).

Speed tests can vary somewhat but I rarely see the x1 match the x6 test.

That just indicates some congestion somewhere. As I've said the X1 is fast enough not to slow things down majorly, a vast majority of things you do online are single threaded so multi threaded speedtests can be misleading of the actual performance you see.