Changed ISP But Speed Still Poor

At the start of October I raised an issue with my ISP about a drop in speed. On an 8meg service I was getting speeds below 1meg. Eventually it evened out at 3/4mb on average against a potential of 6mb. At the time my line was checked and indicated a slight problem between the exchange and my home. BT fixed this but explained that it was only minimal dampness problems on a connection beneath ground. They felt that the problem was with the ISP and net congestion around peak times.

I checked all my own equipment to satisfy the ISP - changed filters, removed all equipment from the network other than the router, connected the router to the test socket, ran a clean line test - but nothing really improved matters. Eventually I believed that the fault lay with the ISP and moved to o2 20 meg service a few days ago. This speed has been very dissapointing returning only 8mb against a potential of 14mb. Again I checked all internal equipment without any improvement. In fact the speed actually dropped when connected to the test socket.

I have run noise tests on the phone line and everything is OK. Now I really am stuck. The router is a new o2 Pro router (Thomson) so I have also eliminated the possibility of a dodgy router.

Any ideas on what my next move should be?

If you post your line stats (Attenuation/Sync/SNR) it may help others advise...

Thanks, hope this is what you need:

Uptime: 0 days, 23:35:14

DSL Type: G.992.5 annex A

Bandwidth (Up/Down) [kbps/kbps]: 1,317 / 10,101

Data Transferred (Sent/Received) [MB/MB]: 48.64 / 461.42

Output Power (Up/Down) [dBm]: 12.0 / 18.5

Line Attenuation (Up/Down) [dB]: 14.0 / 26.5

SN Margin (Up/Down) [dB]: 7.0 / 5.5

Vendor ID (Local/Remote): TMMB / BDCM

Loss of Framing (Local/Remote): 0 / 0

Loss of Signal (Local/Remote): 0 / 0

Loss of Power (Local/Remote): 0 / 0

Loss of Link (Remote): 0

Error Seconds (Local/Remote): 0 / 0

FEC Errors (Up/Down): 121 / 191,741

CRC Errors (Up/Down): 121 / 176

HEC Errors (Up/Down): 222 / 141

Are you connected to the router via wireless or wired?

If wireless, have you tried connecting to the router via cable?

Matt

I have ! Mac and 1 PC connected by ethernet and a Mac laptop connected by wireless/airport. Initially when I set up the laptop I had it on ethernet connected to the master socket and the speed was the same as it is now.

You have a lot of noise on the line. Have you tried in the test socket?

Yes and the performance was the same......or worse!

How can you tell that there is a lot of noise? I did a phone test recently to check for noise and it seemed fine.

You won't hear it, it's RF. You could use Routerstats to graph it.

Thanks, I'll try RouterStats. I had a look to see if IPNetMonitorX did the same thing on the Mac......but it seems not.

Can't get RouterStats to work. I'll have another go tonight.

Any alternative?

Routerstats Lite is a bit more plug and play for a TG585.
DMT also allows display of SNR fluctuations.

I'm sure it was the Lite version but I will seek out DMT and try that.

I have found that a Mac networking app that I have seems to measure the noise that I am experiencing. It returns a result of -94. Does that mean anything to anyone?

I was surprised when you said you were only getting 8Mbps against a potential of 14Mbps on O2 because my experience of O2 is that (on LLU) you get download speeds in line with sync at all times.

Your stats show that you are getting download speeds in line with sync because you are only syncing at 10Mbps and slightly north of 8Mbps is all you can expect after overheads.

The real question is why are you only syncing at ~ 10Mbps when you have 26.5dB attenuation - I get faster than you with 37dB attenuation.

You say that you have tried in the test socket - I'd suggest that it would be informative for you to do that again and post up the stats with it connected like that.

Do you have a burglar alarm that dials a monitoring service?

In reply to a post by erskine:

Yes and the performance was the same......or worse!

Performance from the hidden test socket of the NTE should never be worse than from the front socket or any extension.
It would be interesting to see two sets of stats taken at around the same time - One from the test socket and one from your normal setup.

These are the latest set of stats from an ordinary extension. A few days ago, I went over the process again and tested against the test socket but the results were pretty similar to the original set of stats that I provided. I then removed all superfluous wiring in the sockets other than the connections to terminals 2 and 5. The reading below is what was taken after I had done that.

Uptime: 2 days, 1:51:31
DSL Type: G.992.5 Annex M US 56
Bandwidth (Up/Down) [kbps/kbps]: 1,343 / 12,057
Data Transferred (Sent/Received) [MB/GB]: 372.43 / 6.04
Output Power (Up/Down) [dBm]: 13.0 / 18.5
Line Attenuation (Up/Down) [dB]: 15.0 / 27.5
SN Margin (Up/Down) [dB]: 6.0 / 6.5
Vendor ID (Local/Remote): TMMB / BDCM
Loss of Framing (Local/Remote): 0 / 0
Loss of Signal (Local/Remote): 0 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): 0
Error Seconds (Local/Remote): 0 / 0
FEC Errors (Up/Down): 207 / 199,743
CRC Errors (Up/Down): 207 / 157
HEC Errors (Up/Down): 372 / 88

Things are a bit better but I will check against the test socket again and post back.

In reply to a post by erskine:

... I then removed all superfluous wiring in the sockets other than the connections to terminals 2 and 5.....

Using the test socket of the NTE has the effect of removing all internal wiring, including that to any extensions. When using the test socket, you are in effect, connected only to the incoming pair.
If you use a filter for physical compatibility, ensure that it's a proven good filter. Use only the shortest length of RJ11-RJxx cable to connect the router to the socket/filter and if possible, use a laptop on battery power to monitor the stats.

In reply to a post by erskine:

DSL Type: G.992.5 Annex M US 56
Bandwidth (Up/Down) [kbps/kbps]: 1,343 / 12,057

Well, downstream isn't going to go too much faster, but you may get more on Annex A

Yes, that's been suggested but then I would lose my Pro router and some other services...........but it would be cheaper!

In reply to a post by erskine:

Yes, that's been suggested but then I would lose my Pro router and some other services...........but it would be cheaper!

Funny how it changed from before http://forums.thinkbroadband.com/technical/t/3763041...

Yep, this happened after I removed the extra wires from the connections in all the sockets. I thought that I would eliminate a bit of noise but it may simply have been a coincidence.

In reply to a post by BatBoy:

Well, downstream isn't going to go too much faster, but you may get more on Annex A

No, he should get about 16 Meg Sync. I did from O2 @ 27dB attenuation.