New drop wire, worsened linestats

Is it a bit odd that a new drop wire from the pole (changed because of an issue with a second line at the property on ADSL2+ - soon to switch to FTTC) would have worsened linestats?

It got changed today as a second line pair was damaged and the ADSL dropped out when it was windy, the previous drop it turns out was aluminium and the new drop is a 4 pair solid copper wire. (apparently BT don't do 2 pair drops anymore, only single pair and if somebody needs more they use a 4 pair...at least according to my engineer).

Anyways, on the new drop with my FTTC the attenuation has come down yet the SNR has worsened and the Pwr value has changed on the upstream.

This doesn't make much sense to me, a new drop which is solid copper and alumium anda shorter run (since the engineer didnt put a loop on at the house like the old wire was) has worse line stats?

Would that be due to an extra joint that is now in there to connect the old drop wire on the house to the new drop wire? Or is it bad termination?

I'm still getting max speed on the FTTC so it isn't too much to worry about at the moment, it's just curious as to why the stats would have dropped somewhat when I expected the opposite.

What were the exact stats before and after?

At the end you say you are still getting maximum speed, 79999kbps?, on the FTTC. An improvement on the line of the kind you describe, with a possible side-effect of removing noise on the ADSL2+ circuit, will have caused both your max attainable and noise margin to rise. In other words, what you see is good news not bad news.

If you go to my website linked in my sig there is a page about noise margin that explains it.

Previously Attn was 9.5 down
Pwr 12.8 down -8.1 up
SNR 13.6 down (down from the 14.6 i got a year ago which is much down from initially ) and 15.7 up

Current stats are...
Attn 9.0 down
Pwr 12.8 down, -9.6 up
SNR 12.8 down 14.9 up

Max attainable speed has also dropped somewhat although I don't have those figures to hand.

It is very confusing to me, changed out most of a run of aluminium for copper and slightly shorter shouldnt have a negative effect on it unless a) the engineers terminating is worse or b) the extra joint where he joined the old dropwire on the house with the new is affecting it?

On the contrary Roberto, the ADSL line (soon to be FTTC) will no longer have dropouts as its wires were sticking out and bent so that was damaged...the current FTTC line stats have actually worsened dropping the max attainable speed somewhat (although it does still get the max 80/40 sync currently).

See the other post I just made for attn/pwr/snr stats

Thanks

To be honest there is not much change at all and probably nothing to worry about.

Attenuation going from 9.5 to 9.0 - not really significant. Possibly due to the change from Al to Cu but that is all. SNR - 1dB change - again not significant, it will vary by the minute depending on what else is happening, what electronics are running locally, if there is a radio transmitter close by or a neighbour turning their modem on/off. Over time as more are connected the max atttainable and the above 6dB SNR will gradually degrade - very little you can do to alleviate that.

Provided the joint was made using the correct gel-crimps then than will not normally give any problems - they either work or totally fail and you would soon know if that happened.


In te time I have spent typing this my SNR as changed three or four times as has the max attainable

I tend to look at my SNR a lot to see the changes and it's been very stable at 13.6 for about a month now never really changing so the 1db drop did look very strange to me

On the flip side though there's been 0 FEC errors on the downstream in the 2 and a half hours it's been connected, that's not bad since i usually get at least a few an hour (although upstream has had 25 FEC's, 4 CRC's and 4 error seconds but no G.INP on the upstream).

I guess I'll monitor it, I'm expecting a drop again when the second line on this drop wire gets activated with FTTC too :|

In reply to a post by NBK:

and the Pwr value has changed on the upstream.

Remember that upstream has the "upstream power backoff" principle, which reduces your power so as to reduce the crosstalk you cause to other lines.

Upstream power can change at resyncs, as can the consequent SNRM and attainable values, sometimes in less than obvious ways.

Are you sure of te power levels on te upstream ... Are they actually -8.1 and -9.6 dBm ? I would not expect them to be negative.

Pwr(dBm): 12.8 -9.6

Copied straight from the modem running V100R001C01B030SP08

And from the dodgy web interface...

Line attenuation (dB) 12.8 -9.6
Output power (dBmV) 12.8 -9.6

It's always reported a negative upstream value as far as i can see
A paste I have from july 2015 shows...
Pwr(dBm): 12.8 -6.2

Interesting from my HG612:

Software version V100R001C01B030SP08
Firmware version A2pv6C038m.d24j

and

SNR margin (dB) 16.6 0
Line attenuation (dB) 13.6 7
Output power (dBmV) 13.6 7

So exact same Software however my Line Atten and output power are positive and have NO decimal places. What Firmware does yours have?

As you already know, In the GUI the figures given as SNR margin are wrong - they are the Attenuation, The Attenuation figures given are the Output power. It is better to use DSLstats or Telnet in to get the correct figures:

Interesting that the older data you refer to has dBm whereas the newer is dBmV - they are distinctly different. Certainly accessing through Telnet brings it up in dB whereas the GUI (incorrectly?) refers to dBm. There is no easy direct conversion.

Never use the web interface for any stats other than the sync and max attainable rates. You can see by comparing with the telnet ones taken at the same time which are the dodgy ones on the web interface and which of the duplicated ones is valid.

Post deleted by RobertoS

I did not say they were right, although I believe they are, They are direct from the GUI ...

I'll PM a link.

My firmware is
Firmware version A2pv6C038m.d24j
Same as yours.

I've also dug up some linestats from Jan 2015 not long after I got FTTC and...

Max: Upstream rate = 43516 Kbps, Downstream rate = 134804 Kbps
Path: 0, Upstream rate = 19999 Kbps, Downstream rate = 79999 Kbps

Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 21.6 24.0
Attn(dB): 0.0 0.0
Pwr(dBm): 12.8 -6.9

Which quickly changed to this by Feb 2015...

Max: Upstream rate = 38966 Kbps, Downstream rate = 120440 Kbps
Bearer: 0, Upstream rate = 19999 Kbps, Downstream rate = 79987 Kbps

Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 17.5 23.4
Attn(dB): 9.1 0.0
Pwr(dBm): 12.8 -6.9

So yeah upstream Pwr has always been a negative value since day one

In reply to a post by MHC:

Are you sure of te power levels on te upstream ... Are they actually -8.1 and -9.6 dBm ? I would not expect them to be negative.

My upstream power levels are negative (taken from a telnet session into a Billion 8800NL) - both the single aggregate value and the 3 individual bands too.

That's with an attenuation of under 9dB, and a line of around 100m (though probably 0.4mm copper).

Down            Up
SNR (dB):        11.3            14.8
Attn(dB):        8.7             0.0
Pwr(dBm):        13.1           -9.2

and

VDSL Port Details               Upstream                Downstream
Attainable Net Data Rate:           30489 kbps             102330 kbps
Actual Aggregate Tx Power:         -   9.2 dBm               13.1 dBm
====================================================================================
        VDSL Band Status                U0              U1              U2              U3              U4              D1              D2              D3
  Line Attenuation(dB):  1.9     8.5     11.2     N/A     N/A    4.7     10.8    17.0
Signal Attenuation(dB):  1.9     8.0     10.9     N/A     N/A    6.1     10.7    17.0
                SNR Margin(dB):  15.2    14.8    14.8     N/A     N/A    11.3    11.3    11.3
                 TX Power(dBm): -22.8   -34.1   -9.4      N/A     N/A    9.8     7.8     7.0

Edited by deleted (Mon 15-Aug-16 15:56:23)

I am just suspicious of such low power levels. In your case -22.8 and -34.1 dBm are down in the micro and 10s of nano watt region.

This is a common (but still irritating as it's repeated so often) misconception. Router/modems generally report the SNR noise margin (which is a "safety" buffer), not the actual SNR. If you have maintained line speed yet the noise margin has increased and your attenuation has decreased then your actual SNR has got better.

There are other reasons why the noise margine might get increased, typically because a line has become unstable, and that does decrease sync speed but that doesn't seem like the case here.

So please, please everybody repeat before posting. SNR/noise margi.n is not the same as SNR. They are related, but they are not the same thing.

In reply to a post by TheEulerID:

So please, please everybody repeat before posting. SNR/noise margi.n is not the same as SNR. They are related, but they are not the same thing.

I gave up trying to explain that about 10 years ago!

It appears however the OP's SNRM fell when the line was replaced. Contrary to my expectation (triggering my explanation to him early on), when he said it had got worse, which I assumed was the misconception you describe.

It had got worse - it fell. Though it now looks as though it wasn't by a significant amount.

It may not be significant (at the moment, i'm expecting a drop when the 2nd line on this dropwire gets activated with FTTC!) but it is very perculiar indeed and goes against all expectations

Misconception or not the fact that modem/router stats refer to it as SNR means it's easier to refer to it as SNR for whatever purposes they have in quoting the figures.

Additionally a drop in the SNR value on the linestats along with a lowering of max attainable speed indicates it has indeed worsened

the previous drop it turns out was aluminium

There has never been aluminium dropwire.

Line attenuation actually went down, which generally indicates improvement as well. So it's not all bad

Just wondering, are you very close to your cabinet, because the power levels can and do get attenuated if its not required to get a good signal and not to create crosstalk to other lines.

Because -9.2 dB is rather low, so I would home the noise level is lower than that.

A friend of mine who has VM told me that he had to have attenuators installed to his connection to reduce his power levels due to his line was causing issues to other customers on his cabinet.

So maybe that's what's happening here.

Paul

Yes, about 100m away (I think 0.4mm copper, so equivalent to 150m of 0.5mm copper) - and it's the "upstream power backoff" principle I mentioned earlier that performs this "attenuation". Modems on short lines agree on a lower power level to transmit at.

My QLN shows that U1 is pretty quiet, but has no data for U2. The downstream bands are noisy.

My bitloading puts 9 bits in U0, 6-7 bits in U2, but only 3 bits in U1 - showing it is really holding back in U1. I don't know if it truly is down in the nano-watts, but it is low!

the new drop is a 4 pair solid copper wire. (apparently BT don't do 2 pair drops anymore, only single pair and if somebody needs more they use a 4 pair...at least according to my engineer).

The engineer is right . Was ment to save on copper but most guys just use the 4pr all the time now ! Well you can only carry so much stuff.

Was maent to save on copper but most guys just use the 4pr all the time now ! Well you can only carry so much stuff.

I have been told you can get a 10 pointer for using CAD55 when only Drop11 was required.

And you would be right .

Kinda daft really, all that time spent developing new racking in vans, and yet forgot the simplest part, a decent bit of space to store all the other random gear we have to carry ...... And thus making engineers just carry CAD55, risk an audit failure, because their can has not enough room.

BS is what it is.

not suprising its a 10 pointer, dropwire15 (cad55) is not as good for broadband when ive been testing compared to drop11,

if I had space I would be carrying dropwire12 0.9cu,

Time to convert the passenger seat...

I think its got more to do with the price !

Would that explain why my stats have worsened then?
I've gone from an old 2 pair aluminium drop wire to a new 4 pair copper wire and I'm still seeing speeds and stats worsening :| Dropped another 0.4 over night on the down and upstreams :|