cancel
Showing results for 
Search instead for 
Did you mean: 

Conflicting speed test

HaresEar
Grafter
Posts: 123
Registered: 02-10-2013

Conflicting speed test

Recently I had several del disconnections and today my speed dropped from a steady 9.5 and varies between 3 and 6.5.  The BTW shows 9.5 but PN is a lot lower.  Looking at the DSL info the SNR margin down is slowly going down and is now at 4.4.  Any ideas.
Uptime: 1 day, 5:22:35
DSL Type: ITU-T G.992.5
Bandwidth (Up/Down) [kbps/kbps]: 443 / 11.195
Data Transferred (Sent/Received) [MB/MB]: 271,33 / 345,94
Output Power (Up/Down) [dBm]: 12,9 / 0,0
Line Attenuation (Up/Down) [dB]: 19,3 / 34,5
SN Margin (Up/Down) [dB]: 23,1 / 4,5
System Vendor ID (Local/Remote): TMMB / ----
Chipset Vendor ID (Local/Remote): BDCM / IFTN
Loss of Framing (Local/Remote): 9 / 0
Loss of Signal (Local/Remote): 13 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): -
Error Seconds (Local/Remote): 7.017 / 0
FEC Errors (Up/Down): 0 / 0
CRC Errors (Up/Down): 9 / 4.515
HEC Errors (Up/Down): 0 / 106.025
7 REPLIES
Superuser
Superuser
Posts: 11,911
Thanks: 3,278
Fixes: 22
Registered: 22-08-2007

Re: Conflicting speed test

Hi,
These figures look reasonably sound to me - you do not indicate when you captured the information, but for the purpose of responding, I'll assume it was just before you posted.
SNRM is not static, it will go down in the evening and up during the day time from just MW radio transmissions alone.  How much will depend on the quality of your line.  Accoring to the data posted (note my assumption above) your last resynch occurred around 15:30 in the afternoon.  At that time, dusk has started to fall and therefore MW interference is just starting to kick in, thus your best synch speed (at 6dB) might have been a little degraded.
By the time you took the stats, night time interference would be well established, so a SNRM of 4.5 is not unreasonable.  You might normally expect to see night time SNRM fall by 2 to 2.5dB even in the best of conditions possibly even more on bad lines.
For the reported line attenuation you are getting 90% of the KITZ guide synch speed for your line, thus your synch speed obtained on the last early evening resynch is not unreasonable.  If you do another resynch during daylight hours on a bright sunny day, you might see a couple of 100kbps more, but not much.  Your DATA TRANSFER rate (IP PRofile) cannot be more than 88.2% of the synch rate, which for your current synch rate is 9.8Mbps.  If the 9.5 figure you mentioned is that data transfer rate then you are pretty much on the nail.  That said if the PN "current spped" figure is much lower, then there could be a problem with the automatic profile updates.  You will never see a DATA TRANSFER test deliver anything better than the PN profile figure.
If the PN profile is much lower than the BT profile, then the former needs a nudge.
Have you done the BT line speed tests and the "further diagnostics" option?
HTH?
HaresEar
Grafter
Posts: 123
Registered: 02-10-2013

Re: Conflicting speed test

Thanks for the info.  Latest speed test BT

FAQ
Results Image not loaded

1. Best Effort Test:  -provides background information.
Download  Speed
9.45 Mbps

0 Mbps 21 Mbps
Max Achievable Speed
Download speedachieved during the test was - 9.45 Mbps
For your connection, the acceptable range of speeds is 4 Mbps-21 Mbps.
IP Profile for your line is - 9.88 Mbps
2. Upstream Test:  -provides background information.
Upload Speed
0.33 Mbps

0 Mbps 0.45 Mbps
Max Achievable Speed
Upload speed achieved during the test was - 0.33Mbps
Additional Information:
Upstream Rate IP profile on your line is - 0.45 Mbps

We were unable to identify any performance problem with your service at this time
PN  Today 21:58 6667 kbps (833kB/s) 372 kbps (46.5kB/s)
Latest stats.
Uptime: 1 day, 6:40:39
DSL Type: ITU-T G.992.5
Bandwidth (Up/Down) [kbps/kbps]: 443 / 11.195
Data Transferred (Sent/Received) [MB/MB]: 282,74 / 492,15
Output Power (Up/Down) [dBm]: 12,9 / 0,0
Line Attenuation (Up/Down) [dB]: 19,3 / 34,5
SN Margin (Up/Down) [dB]: 23,1 / 4,5
System Vendor ID (Local/Remote): TMMB / ----
Chipset Vendor ID (Local/Remote): BDCM / IFTN
Loss of Framing (Local/Remote): 9 / 0
Loss of Signal (Local/Remote): 13 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): -
Error Seconds (Local/Remote): 7.076 / 0
FEC Errors (Up/Down): 0 / 0
CRC Errors (Up/Down): 9 / 4.577
HEC Errors (Up/Down): 0 / 106.176

£mb difference.Huh
HaresEar
Grafter
Posts: 123
Registered: 02-10-2013

Re: Conflicting speed test

Plusnet can you help with this issue.
Community Veteran
Posts: 38,460
Thanks: 1,027
Fixes: 62
Registered: 15-06-2007

Re: Conflicting speed test

I don't see the problem
The noise margin has dropped a bit from the start point - this is fairly normal on many lines where the higher frequency tones drop out and don't re-establish
BT Speed test result is as would be expected given the IP profile and the commensurate Plusnet profile of 9.8Mbps
What speed test are you referring to when you state PN Today
Have you considered asking for the upstream speed to be uncapped
HaresEar
Grafter
Posts: 123
Registered: 02-10-2013

Re: Conflicting speed test

I am referring to PN Bandwidth test in the members centre.
http://www.mybroadbandspeed.co.uk
Community Veteran
Posts: 38,460
Thanks: 1,027
Fixes: 62
Registered: 15-06-2007

Re: Conflicting speed test

OK
I just checked it on my connection and it is also coming up with a low value compared with the BT speedtest and the tbb test http://www.thinkbroadband.com/speedtest.html?utm_source=mainsite&utm_medium=navigation&utm_campaign=... so I think you can ignore the result as it obviously isn't giving the correct answer
Plusnet 2484kbps
BT 3140kbps
TBB 3120kbps
Superuser
Superuser
Posts: 11,911
Thanks: 3,278
Fixes: 22
Registered: 22-08-2007

Re: Conflicting speed test

Remember that these are end-to-end data transfer speed tests which are influenced not only by YOUR broadband connection speed, YOUR LAN speed but also by the capacity of the common bearer circuits AND the end host systems delivering the speed tests.
The key factors which need to be optimal for the best  end-to-end data transfer speed tests are...
1. You have the best possible LAN connectivity - usually that means ethernet cable
2. Nothing else - other computers or applications - must be using the internet service whilst the test is run
3. Your broadband connection is running at the best compromise between line synch speed and error rates (see below)
4. The rest is down to the loading / capacity of the internet services used to deliver the speed test
Broadband link performance
Broadband performance is a compromise between speed and errors - sometimes it can be better to go slower (lower synch rate) to go faster (better data transfer with fewer errors).  Though it seems illogical, a very fast synch rate generating a lot of errors will require a high level of retransmission and thereby take longer to move a unit of data.  That what the DATA SPEED TESTS are in effect doing - measuring how long it takes to deliver X mega bytes of data from the host to your PC and back again.  A slower synch rate which produces fewer transmission errors will over all deliver a better data throughput rate.
In addition to considering the error rate, line (connection) stability has to be factored in.  An unstable line (due to being unable to maintain synchronisation) will in addition to high retransmission rates also result in multiple resynchonisations thereby slowing down communications and causing disconnections from internet services. This can happen when (due to electical noise) the SNRM falls to ZERO for extended periods.  To counter that, the DLM will raise the TARGET SNRM in 3dB increments until a level at which a stable connection can be maintained is attained.  An increase in SNRM is at the expense of line synch speed.
The best possible SYNCH speed is ultimately limited by your line attenuation, which is in effect its electrical length.
Noise picked up on the line will degrade the signal, introduce errors, possibly giving rise to the need to retransmit the same data over and over again.
To counteract this, DLM can move the SNRM up, which reduces the SYNCH speed, thereby reducing the error rate, with the potential of reduced (wasteful) retransmissions actually increses the data throughput rate.
To guard for over-loading the synch rate, data is never sent faster than the IP profile rate, which is 88.2% of synch rate.  It is not unkown (due to issues with the BTw data feed to PN) for the PN IP profile to lag behind that set for the line by BTw.  If there is a difference between the two IP profile figures, then a request to PN can fix this.  Certainly the PN rate being lower than the BT rate will reduce DATA THROUGHPUT speed, however it being higher has been suspected of causing issues too!
After that your DATA THROUGPUT speed is potentially further degraded by YOUR LAN configuration / use (remember that WiFi is susceptiable to interference and can therefore suffer the over heads of data retransmission) and the concurrent activity on the shared systems.
As I said before and OldJim confirmed, your figures are looking fair and are within expectations for your line's attenuation.  Could more be squeezed of of it - well I don't know.  I do not know if the local LoS and Error Seconds and the DS HEC rates are of concern.  Someone else would have to comment on them.
HTH,
Kevin