ADSL2+ bit-loading gaps (moved from technical)

My line has recently been migrated from 20CN to WBC. Downstream attenuation is now 33db and syncs at 14,000 Kbps (which I think is pretty good) with a 6db margin. Line is rock steady and stays up for weeks at a time.

I have a few questions:

1. I'm on fast path. Previously on ADSL1 I was getting perhaps 200 error seconds a day, now on ADSL2+ I am getting 2000 errors a day. They don't seem to be impacting on throughput - does 2000 error seconds a day seem excessive ?

2. When I first sync'd (during the day) I didn't have any gaps

http://www.flanz.plus.com/bit_loading_on_sync.PNG

but over time a large number of gaps in the bit loading graph have appeared:

Gap1 1376kHz-1380kHz (Tone319)
Gap2 1423kHz-1427kHz (Tone330)
Gap3 1440kHz-1445kHz (Tone334)
Gap4 1656kHz-1660kHz (Tone384)
Gap5 1669kHz-1678kHz (Tone387-Tone388)
Gap6 1721kHz-1734kHz (Tone399-Tone401)
Gap7 1747kHz-1751kHz (Tone405)
Gap8 1837kHz-1841kHz (Tone426)
Gap9 1846kHz-1850kHz (Tone428)
Gap10 1885kHz-1889kHz (Tone437)
Gap11 2057kHz-2061kHz (Tone477)
Gap12 2074kHz-2079kHz (Tone481)
Gap13 2087kHz-2100kHz (Tone484-Tone486)
Gap14 2117kHz-2126kHz (Tone491-Tone492)
Gap15 2139kHz-2143kHz (Tone496)
Gap16 2148kHz-2161kHz (Tone498-Tone500)
Gap17 2165kHz-2204kHz (Tone502-Tone510)

http://www.flanz.plus.com/bit_loading_days.PNG

Is this normal ? This looks like quite a lot of noise to me ? Is there anything I can do about this - my setup is a NTE5 master socket (in the hall) which is extended via 2 meters of CAT 5 cabling to a face plate filter (on a slave NTE5) in my study.

3. My upstream is currently capped at 448Kbps - if I uncap the upstream is that likely to make the errors/gaps any worse ?

Thanks for any help,

Mike

Upstream should have no impact on downstream.

Nothing odd probably radio stations the router can see

ADSL2+ is more susceptible to noise than ADSL. So although it isn't nice to see those gaps, note that they are predominantly in the additional higher, more sensitive frequencies that ADSL2+ adds.

On the other hand 2000 error seconds a day does seem high to me. We can't scroll the graph of course in a screen shot.

Re the slave NTE5, that works but it is not recommended to have more than one on a LAN. The slave should be an xtf-68 or 85. I doubt if this is causing the issue, but I would fix it anyway.

In reply to a post by flanzm:

now on ADSL2+ I am getting 2000 errors a day.

I'd find that slightly high. I'd expect no more than 1 per min on Fast Path or 1 per 5 mins on Interleaved.

In reply to a post by flanzm:

Is there anything I can do about this - my setup is a NTE5 master socket (in the hall) which is extended via 2 meters of CAT 5 cabling to a face plate filter (on a slave NTE5) in my study.

Is there anything plugged in to the master socket in the hall?

No nothing at all - it's a simple slave NTE5 with face plate from the master.

I'm going to try replacing the slave NTE5 with an XTF-85 as suggested and replace the cheapo RJ11 lead with a screened modem cable to see if that makes any difference.

If that doesn't work I'm going to have a go at trying to find out where the noise is coming from - although I'm beginning to suspect it's outside the house.

Edited by flanzm (Wed 29-Jan-14 16:37:06)

Uncapping the upstream should have little to no effect on the downstream, it'll just use up some of the high SNR to pile more bits per tone into the existing frequencies.