Print

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

[Ixel]

This may seem pretty minor, so apologies if it is, but I appreciate any comments.

In my time of having FTTC, more than 2 years, I've noticed on an unlocked HG612, a Fritz!Box 7390, a Cisco 887VA and finally an ASUS DSL-AC68U, that my CRC error count increments by 1-2 every minute or two on the downstream. You could almost count it, it's somewhat predictable, with exception to the less often spike of usually 10-20 CRC errors.

I had an engineer out unintentionally quite a while back (some people may remember me complimenting him for his efforts), he found a battery contact fault and from my understanding he fixed most of the problem but something still remained. At the time I didn't notice a problem with the line, although since having FTTC my CRC has always followed the above pattern.

Things I've tried
- Various microfilters instead of the MK2 SSFP
- Various modems (as mentioned above)
- Quite line test: There's no noise on 17070 option 2
- Keeping the modem off for one night until 9am~
- Unplugging the DECT phone
- Plugging the phone and modem into a power conditioner
- Changing the power supplies of the phone and modem

Other things I've noticed
- On the HG612 I noticed a fair amount of bitswapping, sometimes hundreds per minute on the downstream if my memory serves me well

On another thread it was suggested that I should ask for a line test to be done. As my estimates from BT's tool are 74.1Mbps to 64.3Mbps - clean, I get an attainable of 91/30, and actual sync of 74/20 (from the ASUS DSL-AC68U and ECI /r), I don't feel I really have a leg to stand on regarding requesting a test or reporting a fault regarding regular CRC error blips. I currently have the Zen Unlimited Fibre Office package, which is the highest business FTTC package they do (and comes with critical care for a prompt repair time in the event of downtime ).

My question is do you think I should open a ticket with Zen (or phone them?) and ask them to run a test?

If you require me to post some stuff from the HG612 then I'll plug it in briefly and post things like the QLN.

[NewtronStar]

My question is do you think I should open a ticket with Zen (or phone them?) and ask them to run a test?

If you require me to post some stuff from the HG612 then I'll plug it in briefly and post things like the QLN.

If it's crosstalk then I don't think any ticket with Zen is going to solve it and a possible Openreach  visit charge if no fault is found on there network, it's your call but a connection speed of 74/20 in my books is excellent.

[Black Sheep]

Concur with above. 

[Ixel]

Yeah, 74/20 is good but what concerns is me the way the CRC's are (particularly in comparison to some other ASUS users of the same device), but you're probably right - perhaps it's crosstalk. If DLM does push me onto interleaved again then I'll have to go back to the ECI /r and either sell the ASUS or just store it for a possible future use (e.g. when vectoring is rolled out, if it ever will be that is).

It's indeed a risk if BT Openreach come out and then find no problems. A risk I'll definitely not be up to.

[Chrysalis]

a fault for 1-2 crc errors a minute?

No.

And your speed also seems within spec.  If your line passes GEA I feel I wouldnt bother with engineer.

Whilst a pair swap can improve things, it can also make things worse and now days engineers it seems have to find a reason to justify a pair swap.

[Ixel]

a fault for 1-2 crc errors a minute?

No.

And your speed also seems within spec.  If your line passes GEA I feel I wouldnt bother with engineer.

Whilst a pair swap can improve things, it can also make things worse and now days engineers it seems have to find a reason to justify a pair swap.

Fair enough, well with the CRC's being so minor this unfortunately pushes the ES value up regardless of how minor the CRC errors are, even just having 1 CRC = 1 ES .

Given everyones feedback so far I won't bother with asking if it's possible for a test to be done just to see if a fault is detected. I'll see what happens with the ASUS and go back to the ECI /r if it goes interleaved. Thanks everyone .

[HighBeta]

If it was aaisp I'd say yes with out hesitation. Also the clueless charts would pick issues up in an instant.

Not so sure with Zen. Unless the line goes interleaved as you mention.

Edit - sorry for the caps
aaisp will literally have openreach diging up the street with in days of a fault being reported.
Didn't quite find that with zen as Chrysalis highlights below

[Chrysalis]

yeah with plusnet or aaisp I would consider it as plusnet' charge is lower plus they have a better attitude than zen, aaisp will always fight the customer's side disputing charges etc.

2 errors per minute tho DLM territory.  As even if its 1 error per ES 2 per minute is 2880/day.  So I respect its not as low as it sounds, but if these are spread out evenly rather than coming in bursts I think it will be shrugged off.

[Ixel]

I see, thanks guys. Yeah AAISP does have its advantages, although to use the internet with their plan would've been very expensive for me as I use 300GB+ bandwidth each month. I use AAISP for their mobile phone SIP service, works superb!

I'll see if DLM does anything in the morning, if it does then I'll return to the ECI /r and put the ASUS in storage until there's a potential fix for the newer firmware. Fiddling with firmwares and the settings (trying to fix the latest firmware spiking CRC's by the thousands), including the problems I had with the latest firmware in the early hours of this morning, may well be the thing that triggers DLM.

[Ixel]

My suspicions are confirmed, just posting to say that it looks like DLM has speed banded me, fortunately no interleaving though . I suspect the speed band was caused by the re-syncing in the early hours of the morning on version 2072 of the ASUS DSL-AC68U firmware. If I wait a couple of days to just over a week and limit my re-syncs to one per day, or none at all, I imagine my speed will return to normal again (hopefully).

I'm mainly posting a reply to mention a theory I have regarding DLM, unless it's already understood to be this way anyway.

An interesting note is that I'm wondering if DLM has two forms of capping, meaning:
- Speed bands are used depending on attainable rate, and also used if re-syncs occur too often
- Interleaving/INP/delay are used if the connection has a high ES (that exceeds the threshold defined by the DLM profile)

My current speed is 49/15. My speed band was 74/20.