Yes it certainly would - configure RSL to give the UPSTREAM SNR and Attenuation on a single graph. When you host the image and provide a link, can you also add the full stats for the router.

Even just a single snapshot view of the attenuation and noise margin and connection speed data would help.

On site again today

I plugged in a D834GT and rang the BT business helpdesk

They tested the line got me to a speedtest (expected 5-6MB)and would not entertain ANY notion that the 0.2 upload was slow

Confirmed it is an ADSL2+ line and Samknows suggests 16Mb download ??

Pics:

Speedtest

Detailed telnet stats

Bits tone first

Bits tone second

The bin plot suggests something is wrong, but not seen an issue like this often enough for my brain to recall a reason why.

The plot confirms that the downstream attenuation looks sensible, as still seeing bits used in the high frequency bins, it is the bits for the upstream that look wrong.

http://www.thinkbroadband.com/images/reviews/fritzbo... shows a longer line with the attenuation causing higher frequencies to fade away. Notice how the green upstream is a tidy shape, rather than sawtooth you have.

The 42dB attenuation for upstream also points to there being a low frequency issue. Given unlimited resources it would be send someone to the street cabinet and test the line from there.

Alas BT Retail and BT Wholesale tend to take the view that almost any connection is a good connection. Rather than the unquisitive try to understand why your line is different to the norm.

17 errored seconds in 15 minutes isn't great.

Is there anywhere that confirms it's ADSL2+ on the router lke G992.5 etc, not familiar with the bit plots (though look like a few bits per bin over many tones - rate capped perhaps ?)

I would be asking for a lift and shift in case it's the line card. As it's a Market 3 exchange you have LLU as the ultimate sanction to threaten, or a more motivated ISP.

26dB downstream attenuation and only 6Mb ... and 42dB upstream.

Are you 100% certain that those have not been transposed by the router or software. Do you have a BT 2700/2701 to try - the stats on those are certainly the right way round.

If the attenuations are correct 26dB would normally give a lot better than 6Mb and 42 dB upstream is way higher than I would expect to see - somewhere from 12 to 18 dB upstream would be more in keeping. So it points to a fault somewhere.

Also, upstream SNR is 5.7 dB and that will be the reason the upstream sync is only 320k and 0.2Mb which could be 250kb ties in so BT are "correct"


You need to find out what is causing the upstream attenuation ... Have you tested the connection without filters and no phones connected?

On the first page had a Belkin N+ plugged and that reported similar attentuations

No phone

an yes have already tried just the master socket

In reply to a post by MHC:

26dB downstream attenuation and only 6Mb ... and 42dB upstream.

Are you 100% certain that those have not been transposed by the router or software.

If the attenuations are correct 26dB would normally give a lot better than 6Mb and 42 dB upstream is way higher than I would expect to see - somewhere from 12 to 18 dB upstream would be more in keeping. So it points to a fault somewhere.

Also, upstream SNR is 5.7 dB and that will be the reason the upstream sync is only 320k and 0.2Mb which could be 250kb ties in so BT are "correct"


You need to find out what is causing the upstream attenuation ... Have you tested the connection without filters and no phones connected?

So, assuming the attenuation figures are correct.

Downstream is lower than expected for the attenuation but the 7.4 dB SNR indicates it is about right for the line. It points to the line being very noisy. Is it in an industrial environment? Close to factories &c?

The Upstream attenuation is well above what should be expected with 30dB of excess attenuation and that suggests almost a short circuit or very low impedance path at some point. That is what you have got to find - or get BT to find for you. Provided you have tested everything cleanly (no filters, phones, excess cable, extension wiring) then it has to be outside your facility. Unfortunately getting this across to a BT Technician* or Tech Support agent will be very difficult.


Looking at both points together does suggest that the noise cause the low downstream is being picked up close to you and not the exchange.


* One reason why I differentiate between Technicians (or engineering Technicians) and Engineers.

Thanks for all the pointers

Environment is a garage in an otherwise residential area

Both routers used on site have been tested at my home and give the normally expected upstream

I have now got they're account managers details and am formulating an email - watch this space!

By Garage do you mean a petrol station and car repair type facility?

If so, there is likely to be a lot of RFI (noise) floating around from petrol pumps, control links and other equipment. Is there a time when everything is off and you could try again to see if the Downstream SNR changes or it resyncs higher.

You will still need to look at the Upstream attenuation.

Have you followed the wires all the way back from the master to the incoming location to make sure there is not a device hanging off it somewhere - a modem or alarm?