|
|
|
Hi all,
My FTTC service syncs at about 45 down 18 up. It used to be 65/18, however it dropped 20 mbps overnight about 8 months ago. I complained to my then ISP, Sky, who sent an Openreach engineer around who blamed it on crosstalk. I understand that it has an effect as more lines are connected to the cabinet, however I'm still perplexed as to why the sudden drop overnight?
I've spoke to my new ISP (Zen) today who said they can't raise a fault as the Openreach estimate is showing 51 mbps. I'm just curious how it can drop so much overnight and if there's anything I can do to improve things?
I'm on an ECI cab, I'm using an Openreach ECI modem and an ASUS RT-N66U router.
Any help appreciated.
|
|
|
It could literally be one line that got activated in the day, people came home from work and plugged in the modem and viola you got the cross talk.
Cross talk can go from nothing to very bad, if the wrong pair in the bundle is close to yours.
Its why so much effect went into vectoring, alas this has not worked out that well for ECI kit.
|
|
The author of the above post is a thinkbroadband staff member. It may not constitute an official statement on behalf of thinkbroadband.
|
|
|
Its why so much effect went into vectoring, alas this has not worked out that well for ECI kit.
You seem to be implying that vectoring has been rolled out for Huawei kit, which was not my understanding.
Kevin
plusnet Unlimited Fibre - sync approx 64500/20000 at 450m - BQM
Using OpenDNS
Domains and web hosting with TSOHOST
|
|
Register (or login) on our website and you will not see this ad.
|
|
|
You seem to be implying that vectoring has been rolled out for Huawei kit, which was not my understanding.
It has. Just far, far, far from all of the Huawei kit.
|
|
|
As MrS says - one line can give significant cross talk.
On one line when a family moved out from a house a few doors away their modem was obviously turned off. Sync increased by around 10Mbps, then about 10-15 days later when te new owners finally moved in and started using their service, the additional speed was lost. That is all it takes.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
M H C
taurus excreta cerebrum vincit
|
|
|
|
I really didn't think it was that bad, so thanks for the replies!
Is there a better modem I could be using?
|
|
|
Its why so much effect went into vectoring, alas this has not worked out that well for ECI kit.
You seem to be implying that vectoring has been rolled out for Huawei kit, which was not my understanding.
I believe that the comment by MrS meant that BT put 2-3 years of development into Vectoring i.e. a lot of effort. What has been rolled out is G.INP/G998.4 which is not vectoring they are different but complimentary technologies for handling crosstalk etc. Vectoring has been rolled out to a limited number of exchanges from what I understand of the B press releases.
With BT now looking to roll out G.fast vectoring appears to be less attractive to them from a financial perspective.
plusnet Fibre > Sky Fibre Pro > Pulse8 Fibre XL - 14ms Ping, Sync ~ 65.78/18.73Mbps - BQM
|
|
|
You seem to be implying that vectoring has been rolled out for Huawei kit, which was not my understanding.
It has. Just far, far, far from all of the Huawei kit.
That's not a roll-out then.
|
|
|
|
The others have mentioned the direct impact of crosstalk. This is where there is an impact directly on the noise measured by the modem, so it chooses to use fewer bits, resulting in a lower sync speed.
Note that, while crosstalk impacts a line gradually, the consequent loss of speed is only felt when a re-sync happens - so can appear suddenly. Especially if you've left your line running for a few months, and can't see the line statistics.
There is also an indirect impact: errors. The additional noise from crosstalk can also be seen as an increase in the error rate on the line.
An increasing error rate will be monitored by DLM, which will eventually intervene. Before G.INP was used, a DLM intervention could also result in you losing an extra 10-20% of your bandwidth, and add 8-16ms of latency to your ping times.
In the case of this kind of DLM intervention, the crosstalk will have been building gradually, and the error rate increasing gradually. But intervention will result in a sudden, overnight loss of speed.
The now-standard use of G.INP downstream, on Huawei cabinets, results only in a hit in latency of 0.2ms, and a notional increase in bandwidth of only ~5% - but many lines have seen speed increases instead. Much better than the old-style of intervention.
Unfortunately, G.INP isn't in use on ECI cabinets, and isn't a solution for all lines on Huawei cabinets either. For these cases, old-style DLM intervention remains a burden.
You have probably felt the impact of both the direct increase in noise, and the indirect intervention of DLM in losing your 20Mbps.
Better modem?
I don't know that there is one that will give you better performance. But getting one that reports statistics could be beneficial, even if it only lets you understand how things are working. The usual alternative is an HG612, unlocked.
|
|
|
A few months ago I got a Vigor 130.
Current stats:
| Text | 1
23
45
67
89
1011
1213
1415
1617
1819
2021
| > show adsl
---------------------- ATU-R Info (hw: annex A, f/w: annex A/B/C) ----------- Running Mode : 17A State : SHOWTIME
DS Actual Rate : 47408000 bps US Actual Rate : 20000000 bps DS Attainable Rate : 47762600 bps US Attainable Rate : 23501486 bps
DS Path Mode : Fast US Path Mode : Fast DS Interleave Depth : 1 US Interleave Depth : 1
NE Current Attenuation : 16 dB Cur SNR Margin : 6 dB DS actual PSD : 4. 8 dB US actual PSD : 5. 0 dB
NE CRC Count : 33 FE CRC Count : 206 NE ES Count : 33 FE ES Count : 203
Xdsl Reset Times : 0 Xdsl Link Times : 4 ITU Version[0] : b5004946 ITU Version[1] : 544e0000
VDSL Firmware Version : 05-07-06-0D-01-07 [with Vectoring support] Power Management Mode : DSL_G997_PMS_L0
Test Mode : DISABLE -------------------------------- ATU-C Info ---------------------------------
Far Current Attenuation : 18 dB Far SNR Margin : 6 dB CO ITU Version[0] : b5004946 CO ITU Version[1] : 544eb206
DSLAM CHIPSET VENDOR : < IFTN > > |
Am I getting the right downstream for the stats?
|
Pages in this thread: 
Print Thread
