Broadband Related > FTTC and FTTP Issues
Improved latency (8ms) by changing from HHG2500 to Archer VR2800
mw2870:
I have been with Vodafone approx 12 months (of an 18 month contract). I generally tend to ping-pong between BT and VF, each time to contract is due for renewal.
Recently I found this forum, which is fantastic by the way! Great resource!
I decided to get rid of the Vodafone router and switch to something else, and went for an Archer VR2800.
I managed to get hold of my DSL username and password from VF, originally they told me the username @broadband.vodafone.com, which didn't work, reading some posts I found out that it should have been @broadband.vodafone.co.uk
Anyway, prior to switching to the Archer VR2800 I had connected a Raspberry Pi to the ethernet socked on the VF router, and written a python script to use speedtest-cli to run every hour, writing the output into a CSV file and then uploading it to Google Drive.
I am trying to collect some evidence of heavy contention in the evenings (high latency and slow throughput), so thought that this would help, I would quite like to break my contract early and move to another provider.
Anyway what was interesting, was that when I switched from the HHG2500 to Archer VR2800, I instantly noticed a lower latency in my results.
It looks like on average about 8ms lower.
Date,Time,Ping (ms),Download (Mbit/s),Upload (Mbit/s)
06/10/19,01:00,21.937,36.23,8.96
06/10/19,02:00,22.024,36.71,8.97
06/10/19,03:00,22.202,36.62,8.86
06/10/19,04:00,22.403,36.31,9.02
06/10/19,05:00,22.076,36.64,8.49
06/10/19,06:00,22.13,36.30,8.85
06/10/19,07:00,22.516,31.35,8.87
06/10/19,08:00,22.203,36.15,8.95
06/10/19,09:00,22.287,35.63,9.03
06/10/19,10:00,21.701,31.03,8.92
06/10/19,11:00,21.726,36.29,8.91
06/10/19,12:00,22.251,22.95,8.88
06/10/19,13:00,22.188,35.93,9.04
06/10/19,14:00,21.768,33.81,8.88
06/10/19,15:00,22.205,27.75,8.72
06/10/19,16:00,23.323,34.23,8.92
06/10/19,17:00,22.677,33.73,8.71
06/10/19,18:00,22.151,36.30,9.04
06/10/19,19:00,15.986,10.99,0.00
06/10/19,20:00,19.358,28.68,9.30
06/10/19,21:00,25.255,15.81,9.06
06/10/19,22:00,16.588,36.57,9.50
06/10/19,23:00,14.845,31.34,9.34
06/11/19,00:00,14.89,31.97,9.62
06/11/19,01:00,15.321,35.69,9.57
06/11/19,02:00,14.399,33.79,9.47
06/11/19,03:00,14.241,35.43,9.43
06/11/19,04:00,14.58,36.47,9.73
06/11/19,05:00,14.561,35.05,9.30
06/11/19,06:00,17.576,36.67,9.44
06/11/19,07:00,14.102,36.41,9.42
06/11/19,08:00,15.234,36.58,9.39
06/11/19,09:00,14.071,36.14,9.51
Router was swapped around 19:00 on the 10th June.
This got me thinking, as the speedtest-cli is done from the Pi, connected via Ethernet, there is no Wi-Fi to worry about.
Could the improvement in latency by down to better processing in the TP-Link vs the Vodafone router, or could this be due to more favourable line parameters?
I don't have any line stats from the VF router, but again using some fantastic articles on this forum, I managed to enable proper telnet on the Archer VR2800.
I now have DSL stats running on my Pi as well as my speed test script.
Line stats are:
Stats recorded 11 Jun 2019 09:58:18
DSLAM type / SW version: BDCM:0xb12d (177.45) / v0xb12d
Modem/router firmware: AnnexA version - A2pvbH042j2.d26m
DSL mode: VDSL2 Profile 17a
Status: Showtime
Uptime: 10 hours 14 min 0 sec
Resyncs: 0 (since 10 Jun 2019 23:54:46)
Downstream Upstream
Line attenuation (dB): 13.8 0.0
Signal attenuation (dB): Not monitored
Connection speed (kbps): 40000 9997
SNR margin (dB): 10.3 9.4
Power (dBm): 14.0 6.7
Interleave depth: 16 1
INP: 43.00 0
G.INP: Enabled Not enabled
Vectoring status: 5 (VECT_UNCONFIGURED)
RSCorr/RS (%): 0.0000 0.0001
RSUnCorr/RS (%): 0.0000 0.0000
ES/hour: 0 6.03
If the VF router was really not behaving very well, could I potentially see an improvement in my line over time using the new router?
j0hn:
Low Interleaving adds 8ms delay.
I can see from your stats you don't have Interleaving on the line. You have G.INP instead.
--- Quote ---G.INP: Enabled
--- End quote ---
It's possible G.INP wasn't being applied with the Vodafone router.
That would account for the 8ms.
Odd though as it should really be G.INP compatible.
mw2870:
Interesting, thanks for the reply John.
Would the DSLAM detect a different modem immediately and check to see what it is capable of?
I believe from reading the forums that by just changing my hardware the DSM will not be reset.
That would require a change of service or provider.
Last time I was with BT I used to be on Fast Path and had very low latency.
This time round the Vodafone router always reported Interleaved and never ever showed Fast Path.
--- Quote from: j0hn on June 11, 2019, 02:12:15 PM ---I can see from your stats you don't have Interleaving on the line. You have G.INP instead.
--- End quote ---
So if interleaving is not already on, only G.INP am I right in saying that conditions can not be improved any more?
I am on a 40/10 service, so am already synced at full speed for the downstream.
I asked Vodafone if I could upgrade to the higher 80/20, but they would not allow it without taking about another 18 month contract, which I don't wish to do.
I am on CHELTENHAM Exchange, Cabinet 71.
Downstream Handback Threshold(Mbps) is 56.4 (Clean), 54.6 (Impacted)
So I guess they only have to offer me another 16Mbps for the extra money.
j0hn:
--- Quote ---Would the DSLAM detect a different modem immediately and check to see what it is capable of?
--- End quote ---
That's my understanding.
If it detects the modem doesn't support G.INP then Interleaving is applied.
Changing to a modem that supports G.INP will see it activated.
--- Quote ---I believe from reading the forums that by just changing my hardware the DSM will not be reset.
That would require a change of service or provider.
--- End quote ---
Correct.
--- Quote ---So if interleaving is not already on, only G.INP am I right in saying that conditions can not be improved any more?
--- End quote ---
As far as DLM and latency goes yes.
adslmax:
mw2870 - I have send you PM about VR2600 how to enabled telnet for working with dslstats. Thanks
Navigation
[0] Message Index
[#] Next page
Go to full version