Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[father_ted]

Hi all,

I've just moved to an FTTC area and left my Hyperoptic connection behind...

Openreach installed new drop wire from pole today, new master socket 5C, no further internal wiring.

I'm about 300m from the cabinet as the crow flies, the Openreach man said the line was 600m.

Vodafone modem (and VMG8924-B10A, but switched to the ISP supplied router in preparation for the call centre script) both syncing just under 30Mbit. BT Wholesale checker predicted 60-75 for the address and that's what Vodafone quoted, the handback threshold is 52.

At the same time as the router switch I tried an old NTE5a with Mk3 filter plate - sync improved by 0.2Mbit.

Router reports the following:
Attenuation    
DS1 15.9 dB, DS2 45.8 dB, DS3 71.9 dB    
US0 5.7 dB, US1 35.8 dB, US2 56.3 dB
Power    DS 12.4 dBm    US 3.9 dBm
CRC Errors in last 85 minute(s)    DS 24    US3 (CRC errors at 1 minute were 17/0)

Is this in any way normal? Some of those attenuation numbers seem huge.

I can't imagine the speeds will improve that much without engineering intervention. My first thought was that the line is very long - but OR confirmed 600m. What's the best way to present this fault to Vodafone?

Thanks for any comments.

[mofa2020]

Can you show full line stats if possible? i.e sync. speed, attainable speed, SNRm, etc...

[father_ted]

Here's all the VF supplied router will give me:

DSL Status Information
DSL Mode
ITU-T G.993.2_Annex_B (VDSL2)
DSL Uptime
4 hours, 7 minutes and 1 second
Line Coding
DMT
Status
Connected
Number of Cuts
1 (this was me changing a cable)
Link Power State
L0
Line Quality
   Downstream   Upstream
Current Rate    29308 kbps    5917 kbps
Maximum Rate    29654 kbps    6069 kbps
Signal-to-Noise Ratio    6.2 dB    6.2 dB
Attenuation    DS1 15.9 dB, DS2 45.8 dB, DS3 71.9 dB    US0 5.7 dB, US1 35.8 dB, US2 56.3 dB
Power    12.4 dBm    3.9 dBm
CRC Errors in last 247 minute(s)    59    6
K (number of bytes in DMT frame)    0    0
R (number of bytes in RS code word)    0    16
S (RS code word size in DMT frame)    0.2606    0.9988
D (interleaver depth)    1    1
Delay    0 ms    0 ms
   Downstream   Upstream
Super Frames    3540892    913378
Super Frame Errors    42    4
RS Words    0    59343288
RS Correctable Errors    0    18
RS Uncorrectable Errors    0    0
   Downstream   Upstream
HEC Errors    105    0
OCD Errors    0    0
LCD Errors    0    0
Total Cells    835323682    0
Data Cells    27201    0
Bit Errors    0    0
   Downstream   Upstream
Total ES    63    1182
Total SES    10    120
Total UAS    512    9216045

[mofa2020]

I'm about 300m from the cabinet as the crow flies, the Openreach man said the line was 600m.

Well seems that OR man line estimation is right since the current stats confirms the line is giving max speed it can afford @ 6.2dB (both downstream / upstream) and technically it is the same speed expected for a 600m line length, though the checker estimation of 60-75 for the address suggest the line length is ~ 300m or less to get that speed.

check again with the phone number because checking by address is less accurate for speed estimation.

[father_ted]

technically it is the same speed expected for a 600m line length

Most of the guidelines I've seen suggest 50Mbit is expected around 600m, or have I been looking at the wrong data sources?

There was no phone number (and it's still unknown until I plug a phone in tomorrow)...

[mofa2020]

According to thinkbroadband FTTx guide, 600m FTTC line would be estimated around 35mb down 14mb up but other factors can increase or decrease the actual speed.

https://www.thinkbroadband.com/guides/fibre-fttc-ftth-broadband-guide

[parkdale]

Your in a very similar boat to me... my top line speed was 58Mb but has now down to 39Mb @ 566m I can't even get an IP profile now! the ECI line card is borked.

[tubaman]

I'd say those are not good numbers for a ~600m line. With that number of ES and SES in only 4 hours i'd also not be surprised if DLM takes action before very long.
I'm wondering if there is some aluminium cable in that route somewhere?
 

[j0hn]

The TBB FTTC chart is overly cautious in its estimates.

I got 44Mb on a 1024m line with over 600 subscribers on the cabinet.

30Mb at 600m is pretty poor.

[father_ted]

So conversation with Vodafone didn't go well today, Openreach are telling them the line isn't connected to the DSLAM and can't be because there are no ports available... I keep repeating that it's connected to something that will sync at 29Mbit, and the engineer told me which cabinet it was, so they just need to find the port and put the routing in for my PPPoE session - but nobody in the call centre understands that.

The solution is to send an engineer on Friday.

An interesting point on sync is that the noise drops off massively sometime between 2030 and 2230, then the noise returns before 0800. Looking at the line stats now the maximum sync jumps up to 40Mbit (still not great), it remains synced at 29Mbit but SNR is up to 13dB not 6dB. Forcing a resync does bring the link up at 40, and it stays stable all night until sometime before 0800.

What kind of noise on a line is constant (within 0.5dB) all day, then drops off by over 7dB overnight?

[father_ted]

Many hours on the phone later and my PPP session is now up.

However, sync has not changed from 29Mbit apart from overnight when some noise goes away.

I've now changed to a VMG8924-B10A to get better stats from the modem. The headline downstream attenuation is a fairly constant 22.4dB, which seems high for a 600m line (as advised by OR engineer) - isn't the constant around 13dB/km?

HLog seems fairly smooth to me and does not indicate bridge taps are present.

SNR per tone graph looks like the line is struggling at the higher frequencies.

SNR margin graph clearly shows interference starting at 0710 - I'll log for a few days but the time is variable and I can't figure out what it could be. It's definitely nothing in my house, nobody was awake that early and being constant all day, it's not the boiler/fridge compressor or anything that would cycle in and out.

I'm waiting for tonight to compare QLN with the quiet overnight period which should build a frequency profile of the interference.

Anything else I should be looking at?

[j0hn]

22.4dB attenuation suggests about 1km and not 600m.

Either you have a fault or the DSL checker is very wrong to give those estimates for you.

Do neighbours have similar estimates to you? Could you ask any of them what speeds they get?

[father_ted]

Do neighbours have similar estimates to you? Could you ask any of them what speeds they get?

My estimates are on the lower end of the street, which fits with the cable from PCP entering at the other end of the street, so I'm further away.

The other end of the street estimates at 61-80 on BT Wholesale, my end is 55-77.3. All fed from the same cabinet.

Vodafone don't even want to send an engineer, they've gone straight to 'you can have a discount or leave penalty-free'.

I'd rather pay full price and get the full service...

[burakkucat]

Vodafone don't even want to send an engineer, they've gone straight to 'you can have a discount or leave penalty-free'.

I'd rather pay full price and get the full service...

Take Vodafone's option to leave penalty-free and migrate to A&A. Once with A&A, ask them to investigate the low synchronisation speed problem.

[father_ted]

Take Vodafone's option to leave penalty-free and migrate to A&A. Once with A&A, ask them to investigate the low synchronisation speed problem.

That does seem to be the sensible option. I'll give them a call in the morning about their "we'll fix your line" offer. It does mention that it's not intended for people who are 'hoping [they] had a little more speed' - but I think sync below 50% of expected, and well below handback threshold should qualify as a real fault.