I won't need to pinch the frontplate; it's quite loose and a gentle tap will get it off

I wasn't there at the time, but I don't think the engineer did anything outside except possibly a reset.

With the previous, but not very old socket, the extension was wired into the terminals inside the frontplate, that is, not into the back of the socket.

Anyway, I'll take your advice and leave well alone for now - but we'll see how it goes!

Thanks to all for your help.

Update... On Wednesday, the speed hit 69Mbps, and I was well pleased. Unfortunately, it didn't last; each day after that, the speed dropped by about 6Mbps, and by Saturday it was down to around 51.

On Sunday afternoon, though, it had fallen to under 1Mbps! At that point, I thought I would test using an ethernet cable and plugging the HH5 into the test socket. (The extension wiring was definitely isolated as there was no dial tone on it.) That, of course, involved disconnecting it temporarily, but did have the effect of bring the speed back to 43Mbps. However, the line seemed to drop completely shortly after that, so I powered the HH5 down and back up again - and was rewarded with a speed of 49Mbps. I then put everything back, and connected using wifi; on testing the speed was still 49Mbbps.

Later on Sunday evening I was again getting less that 1Mbps, as well as drop outs, so I turned the hub off and on for a second time, and the speed is again 49Mbps (on wireless).

This is exactly what happened a couple of weeks ago following the first engineer's visit - really good initial speeds, followed by a daily decline and ending with ridiculously low speeds which were partially fixed by restarting the HH5.

I'd really like to understand what is going on. Is there any way of telling what my line should be capable of?

The consistent 49Mbps sounds like DLM has reapplied banding. (That's a capped speed (line profile, which us not the same thing as IP Profile).

Post deleted by RobertoS (duplicate)

Edited by RobertoS (Mon 20-Mar-17 01:36:04)

Thanks - so presumably, 49Mbps is the best it thinks my line is capable of, and I'm stuck with it?

Are those speeds all from speed tests? Or from sync speeds? If the latter, then you almost certainly need to report it to the ISP, with the aim of getting an Openreach engineer.

If the former, then you need to start figuring whether the download tests match your real experience.

As for the capability, there are 5 things we can use to cross check for consistency
- The attenuation from the line statistics
- The Max attainable from the line statistics
- The current line speed and current SNRM from the line statistics
- The physical distance to the cabinet
- The two estimated ranges A and B from the BTW checker at https://www.dslchecker.bt.com/

Have you tried the Quiet Line Test?

Particularly when the speed has dropped.

Dial 17070 etc, in both the NTE Test Socket and also on extensions etc when the NTE is closed up in normal operation.

------

Also, the VDSL side of the NTEs and of "dangly" Splitters, still has the ordinary Phone signals on it, as the VDSL side is actually "straight through", no filtering etc.

Only the PHONE side of the NTE and the "dangly" Splitters, is truly "filtered", with conventional Phone signals such as speech etc.

I have confirmed that from taking a Splitter apart, also testing with several in series configuration and consulting circuit diagrams.

The Splitters in the series testing were a variety, clearly not identical, so the tests were not biassed as they would have been if all from one maker, batch etc.

The speeds I have been quoting are all speed test results, either tbb or BT Wholesale. The reported speed has again dropped below 1Mbps a couple of times in the past two days (I've been at home) and when that happens, it is very noticeable - web pages take for ever to load, for example. As suggested, I tried the quiet line test both with an extension phone and one plugged into the test socket - and both seemed, well, quiet, to me. There was a very faint hum on the extension phone, but it was the same when the speed was higher.

The BTW checker shows the following (sorry, can't get the figures to line up ):

Downstream Upstream Downstream Handback Threshold
High Low High Low
VDSL Range A (Clean) 76.8 57.8 20 20 51.6
VDSL Range B (Impacted) 66.5 39.8 20 11.6 32.4

The stats from the HH5 (are there others I can get?) this morning were, firstly when I was getting <1Mbps:

3. Firmware version: Software version 4.7.5.1.83.8.222.1.1 (Type A) Last updated 21/02/16
4. Board version: BT Hub 5A
5. DSL uptime: 0 days, 01:14:38
6. Data rate: 18979 / 52512
7. Maximum data rate: 23357 / 70001
8. Noise margin: 5.9 / 9.6
9. Line attenuation: 18.8 / 15.2
10. Signal attenuation: 18.7 / 15.4

and a little later, after I'd restarted it and was getting about 48Mbps:

3. Firmware version: Software version 4.7.5.1.83.8.222.1.1 (Type A) Last updated 21/02/16
4. Board version: BT Hub 5A
5. DSL uptime: 0 days, 00:07:39
6. Data rate: 18979 / 52512
7. Maximum data rate: 25001 / 73767
8. Noise margin: 7.9 / 11.2
9. Line attenuation: 18.8 / 15.2
10. Signal attenuation: 18.6 / 15.4

What can I learn from these? Is it significant that powering the HH5 off and back on again seems to help - could I have a dodgy hub?

Just to muddy the waters further, I'm moving to Plusnet next week. I haven't requested a hub from them, as I can use the HH5, but maybe I should?

I do appreciate everyone's help.

Given the difference in speed test results but relatively similar connection speeds, it could be an issue with the hub, but rare to see Ethernet side issues like that (if testing over wireless time to flip to Ethernet to double check next time).

If testing over Ethernet what seems more likely may be an oddity with your session in the core network and by reconnecting this session is rebuilt or now avoiding a busy part of the network

The fixed data rate in both directions with a greatly increased maximum is worrying. Have you been rebooting or re-sync'ing frequently recently?

Edit: Ah, I see earlier I suggested it was banded at 49Mbps. Presumably a bit higher now.

I advise you not to mess with reconnections or re-syncs. Just concentrate on the throughput and reasons for it being poor. You could easily trigger more severe banding.

Edited by RobertoS (Fri 24-Mar-17 14:48:42)