Hi,

I've been wrestling with the problem of my 2862 dropping synch almost daily since switching to BT FTTC last October.

I started troubleshooting with BT and Draytek support from around mid-November, but still haven't fully resolved the issue seen.

Draytek support suggested latest firmware and to try the modem variants which I have done, leaving a good number of days to find the same problem occurs on all. I tried the suggestions I've found about switching off the option G.Vector to see if it helps, but it had zero effect.

BT have continued dialogue with me, but always say that without a BT Hub they can't support or guarantee anything; that said, the BT contact continues to listen and is willing to send in an OpenReach engineer... again...

What have I found out is that the BT Hub will maintain synch and has been solid for a fortnight. I can't see any of the link stats, but I am seeing similar values for Line Attenuation, SNR Margin and Attainable Speed measurements.

I have created a simple Python script to telnet to the Draytek and run a number VDSL status commands every 15 seconds, writing a CSV output.

This has led to finding out that after a period of time (measured in hours/days), when no downstream CRC/ES/SES errors are reported they start to increase relentlessly until the BT DLM drops synch, after which the downstream CRC/ES/SES errors return to zero (as expected) for another period of time before they start increasing again, usually within 24 hours.

As you might imagine, BT suspect REIN as the problem, but I haven't been able to identify any particular source - of course I know it could be occurring somewhere in the wiring from the cabinet to the house, but somehow I don't think it is given what I have found out.

Until earlier this evening I haven't been able to investigate while the downstream CRC/ES/SES increases is in progress. I've spent hours isolating and removing potential interference sources, adjusting routing of wires and nothing has had an impact on the CRC/ES/SES errors.

The Quiet line test didn't yield anything either - there is no noise, no fizzes, no buzzes or pops - it's as silent as it can be.

In the end I applied the standard IT approach to fixing problems and switched the Draytek off and on again; like the synch reset triggered by the DLM, all of the downstream CRC/ES/SES errors reset to zero and for the couple of hours there have been no increases - synch speed, attentuation and SNR margin are all the same as before the reboot.

Given REIN doesn't just 'go away' on its own I am suspicious of an issue with the Draytek, especially given the BT Hub is rock solid with the same speed profile.

Any thoughts? I'm at my wits end and determined not to switch to using the BT Hub

Cheers!
Andy

Well, after the cascade of downstream errors that occurred yesterday evening (1,600 CRC, 400 ES and 200 SES in a period of 2 hours 45 minutes) the DLM triggered a synch reset and the reconnected speed was reduced again.

I've now turned off the Draytek 2862 and put the BT Smart Hub 2 back into service - I'm debating whether to raise yet another case with Draytek support or just give in!

FWIW I use a 2862 and never have any problem whatsoever. Although they are a few quid, you might consider buying another one and trying that. At least then you can show if it is the router you have.

After being over a year in the same predicament as you are (with my 2762), I conclude that Draytek routers don't handle bad lines well. Also, if you use them, powerline adapters (even Devolo) will result in more CRC errors and frequent drops because of the overlapping frequencies between powerline and VDSL signal, especially if your line is near an electrical socket. DSL equipment will adjust the line's parameters to compensate for the problems, hence disconnections. You're not telling us whether there's a pattern to disconnections in that whether they occur at specific times etc.

I've switched over to Fritzbox and although I continue to have disconnections every few days, I have fewer of them. You're lucky that Openreach in your area will listen and co-operate. Where I stay, they couldn't give a *** and we even complained to the CEO (without effect), and Plusnet want rid of us. We had 14 Openreach visits and all denying a problem on their poorly maintained network. More competition should be brought in as Openreach have become very very complacent over the years.

it'd be helpful if you posted stats. 1,600 CRC is quite high within a couple of hours, but it's the SES you should worry about. Are you connected to the master socket? Have you tried different firmware types (for example, more conservative)? Is your filter/cable ok? Is your Draytek reporting fastpath or interleave? I considered my 2762 to be faulty too, but then had a period of 2 weeks of stability with it. On my Fritzbox I note more errors when there's windy weather but Openreach claim their network is the best you can get, even though it's p*** p***.

Just wanted to add that my case is also similar to yours in that I have no errors, then a flurry of them, culminating in a dropped sync.

From my research, I've established that we have a high resistant fault that openreach won't fix. It's to do with jelly crimping.

However, I also think that there's some electrical equipment in the area that causes this flurry of errors on my line. I thought it was our neighbours' boiler but I'm unsure now.

I've also seen an impact of power line adaptors (by Devolo) on our line, despite the compatibility mode being on. My Fritz box has 3rd party power line setting integration but I think this is only the case with Devolo magic 1 and 2, whereas I have Devolo 1200.

Hi,

Thanks for the comments - I ran over the message size limit in my first post, so deleted some of the pertinent bits that you raised.

Before the switch to FTTC I had a pretty good ADSL link which did drop occasionally, but never quite as often as it does on VDSL.

I have no powerline adapters, in fact the router is connected to a BT Master Socket 5C that is reasonably new (18 months I think) compared to the original that was there for more than 20 years.

I have tried all modem code variants and they all seem to behave the same - after a variable period of time following a line resynch the error cascade starts until the synch is dropped by the DLM and reconnected or the router is restarted. There is no discernible pattern that I can find. The cascade events have started at any time during the day, but perhaps slightly more during the evening - typically late.

The investigation I carried out the other evening was to try to find and isolate any noise sources to the point where most things in the house were switched off. None of which helped. If it's something in the cabling from the cabinet to the house I can't honestly comment, except to wonder why after the line resynch the cascade stops. I would have expected it to continue.

One thing I haven't tried is replacing the PSU because there are no DRAYPSU17 spares available in the UK which is the one required for the 2862Vac router I have. I will source an alternative though - CPC seem to have a decent variant.

I'll drop the capture I took before I switched off the 2862 in the next reply...

Cheers!
Andy