A Cautionary Tale

[Tacitus]

A while ago my standard router, a Zyxel 662H ceased to make a DSL connection from power up.  I assumed it had failed and moved over to a spare 2700 on the same line.  Some 3 weeks later this also failed with exactly the same symptoms - power cycle it and it failed to gain a lock on the DSL signal. 

Needless to say I contacted support and we went through the, 'try this that and the other' routines to no avail.  It seemed reasonable to me that for one router to fail was unfortunate, for two to fail in the same way in a relatively short space of time was too much of a coincidence.

I took the Zyxel to my sister's and tried it on her iDNet line.  From power up to sync took just over a minute and I left it functioning for an hour or so before swapping it back.  I've since had it functioning on her line for 3 weeks with no problems and swapped it for the 'failed' 2700 which has now been going for 3 days also without apparent problems.

So, we now have two routers - different makes and chip - both of which function perfectly well, but not on my line.  I should add that both mine and my sister's lines are iDNet and both run Max, not 21CN.

After discussion with support I elected to get BT out.  Apparently I was risking £160+vat.  I didn't like it, but by this time I had, not unreasonably, become convinced the problem was at the exchange.  BT guy arrives plugs in his modem and hey presto it works.  He performs a few more tests and goes away - I should add that I plugged in another borrowed modem whilst he was there and that also worked.

Now to the tragic bit.  Apparently unannounced BT have raised their call-out charges and I have been hit with a bill for £195+vat (=£229.12).  Just what I need in the run up to Xmas!  According to support it would have cost less if I pulled a wire out of the faceplate - circa £30 apparently.

I have now been told by support that they have seen occasions when modems suddenly become incompatible with the line.  Apparently it is rare but it does happen. It might have been useful if they'd mentioned this earlier, although I doubt they imagined it would happen twice in the same place with different modems. 

So, the moral of this tale is that when iDNet want you to jump through hoops then do it.  Although it didn't save me, it might save some of you a serious bill.

[Rik]

Sorry to hear about that, Tac. It's perhaps worth mentioning that the bill will rise to £239 from January, due to VAT increases. At the same time BT is moving to external demarcation points, which remove the ability to 'prove' a line at the test socket. This will mean BT are even more likely to raise a charge.

Coupled with Simon's messages that, after four weeks, BT have still not given IDNet an explanation for the October outages, makes me think that they treat small ISPs with contempt.

[Tacitus]

Coupled with Simon's messages that, after four weeks, BT have still not given IDNet an explanation for the October outages, makes me think that they treat small ISPs with contempt.

I think BT treat everyone with contempt.....

[Rik]

True. In that respect at least, it was better when it was still State-owned. When you consider their failings for Scotland, Northern Ireland and the north of England, and the capacity problems we've seen for early FTTC adopters, just as we saw for early WBC adopters, you are left to ponder what it is exactly we have Ofcom for.

[Steve]

Aren't the 2 Wire and the Zyxel 662H not built around the same chipset i.e the TI AR7?

I know the Zyxel is but I'm unsure of the 2 wire however there is the odd reference to a chipset via google

http://forum.kitz.co.uk/index.php?action=printpage;topic=5505.0   post#2 mentions the TI AR7

If correct it explains why two different routers ended up with the same result.

[Tacitus]

Aren't the 2 Wire and the Zyxel 662H not built around the same chipset i.e the TI AR7?
I know the Zyxel is but I'm unsure of the 2 wire however there is the odd reference to a chipset via google
http://forum.kitz.co.uk/index.php?action=printpage;topic=5505.0   post#2 mentions the TI AR7
If correct it explains why two different routers ended up with the same result.

TBH Steve I don't know.  I'm aware the Zyxel uses the AR7, in fact most of the Zyxel range do, but I wasn't aware the 2700 did.  In any event the post you refer to mentions the 2700 on 21CN and both mine and my sister's line are using Max. 

Sis can go to 21CN, but since she has a solid setup, I'm inclined to leave well alone.   

[Steve]

It was just the mentioning of the chipset in the 2nd post i.e TI AR7 that raised the possibility of why two routers should be incompatible with your line's Exchange DSLAM.

[Rik]

Sis can go to 21CN, but since she has a solid setup, I'm inclined to leave well alone.   

Probably wise, Tac. I have gained 1M from 21CN, but I was lucky enough to be in the trial which allowed IDNet to override DLM. Before that, it was all over the place.

[Tacitus]

It was just the mentioning of the chipset in the 2nd post i.e TI AR7 that raised the possibility of why two routers should be incompatible with your line's Exchange DSLAM.

Both myself and Sis use Zyxel routers, although at present we both have 2700s.  I have one with V6 firmware whilst she is currently testing my 'failed' 2700 with V5.

I might hop over to the DSLReports forums and see if anyone there has any idea.

The thing that bugs me about all this is that despite being hammered for £200+, I still have no idea what caused the problem in the first place.  From what Rik said, BT are turning their lack of maintenance and shoddy service into a business opportunity.

You couldn't make it up....

[Tacitus]

Probably wise, Tac. I have gained 1M from 21CN, but I was lucky enough to be in the trial which allowed IDNet to override DLM. Before that, it was all over the place.

Her attn is around 41dB, so she might gain a little speed.  Given her current sync is usually around 6700, although it has dropped to nearer 6000 with the bad weather, there is little to gain and, potentially a lot of grief.  Frankly not worth the bother, since she doesn't really need the extra bandwidth allowance.

[Rik]

The thing that bugs me about all this is that despite being hammered for £200+, I still have no idea what caused the problem in the first place.  From what Rik said, BT are turning their lack of maintenance and shoddy service into a business opportunity.

You couldn't make it up....

Curious that BT don't get engineers to give the customer a form to sign, agreeing to and detailing the work to be done. When my alarm system is inspected every year, I get a form to sign indicating what work has been done or is necessary.

Maybe we should start a petition to Ofcom that they insist that BT do this, which would also prove that they had actually turned up at the premises. After all, they only raise these charges for work done after the test socket, so if they fix the problem at the exchange or along the way, they don't need to attend the customer's premises do they? I can't remember, in the past, being charged for a voice engineer to plug his phone in and do a line test - what's the difference with ADSL. BT say the demarcation point is the test socket, so establishing whether there is a fault at that point should, surely, be part of the service we pay for. Just as it used to be.

[Rik]

Frankly not worth the bother, since she doesn't really need the extra bandwidth allowance.

I agree, Tac, it's why I'm beginning to think that going to fibre just isn't worth it to me.

[Tacitus]

Maybe we should start a petition to Ofcom that they insist that BT do this, which would also prove that they had actually turned up at the premises. ........ BT say the demarcation point is the test socket, so establishing whether there is a fault at that point should, surely, be part of the service we pay for. Just as it used to be.

I agree Rik.  BT turn up and perform some tests - which for all the customer knows could be complete rubbish - and then go away and do a report.  Nothing is agreed by the end user and the bill is unknown until the time it arrives.  Given it can vary between £30 at one end and £230 (ish) at the other, these are the actions of the average cowboy outfit.  There should be some system whereby the cost is known before any work is done.

I was, albeit reluctantly, prepared to accept the risk of £160 or so, but somewhere along the line it should have been made clear that the actual bill was likely to be significantly higher.  Now whether this is iDNet's fault or BTs is a moot point.  All I know is that I've been stung for £200+ and still don't know the cause of the problem.  In practice of course the answer is keep buying modems....

TBH given my kit seems to work on other BT lines, I can't really see how BT can claim my kit is faulty.  If as Steve says above, it is an AR7 problem, then why does the current 2700 work on my line without problem?

[Rik]

I wish I could give you a sensible answer, Tac, but I can't. Possibly the engineer fixed the fault that didn't exist?

[davej99]

Aren't the 2 Wire and the Zyxel 662H not built around the same chipset i.e the TI AR7?

According to ASL24 the 2-wire is STMicroelectronics and the Zyxel P660 series is TrendChip. This rather rules out the chipset.

[Steve]

I can argue about the P662HW

http://en.wikipedia.org/wiki/TI-AR7

but I've no idea about the 2 Wire  BT 2700HGV chipset. Although the STM is used on the 2 wire home portal.

[zappaDPJ]

The thing that bugs me about all this is that despite being hammered for £200+, I still have no idea what caused the problem in the first place.  From what Rik said, BT are turning their lack of maintenance and shoddy service into a business opportunity.

You couldn't make it up....

I suggested the very same thing elsewhere today and what makes it worse in my mind is that BT are using the ISP as a collection service and a buffer between themselves and the customer. It's an abysmal system.

[kinmel]

The 2700 uses it's own ADSL modem chipset, blown onto an ARES chip.

[davej99]

The 2700 uses it's own ADSL modem chipset, blown onto an ARES chip.

Sorry to take the thread a bit of course, but chipsets are increasingly of interest. I wonder Kinmel, if you have some links we could steal to get more info in this area. Maybe we should all be opening up out of warranty boxes; or maybe posting pics after we have stamped on rubbish routers. Cheers, Dave.

[Den]

Just taking up Riks point about prices going up in Jan due to the VAT rise. The cost then would be £234 not £239 as Rik said, he was using MK maths 

[kinmel]

Sorry to take the thread a bit of course, but chipsets are increasingly of interest. I wonder Kinmel, if you have some links we could steal to get more info in this area. Maybe we should all be opening up out of warranty boxes; or maybe posting pics after we have stamped on rubbish routers. Cheers, Dave.

I opened up a V5 and used Google to identify and rule out each and every chip on the board except one, that chip has a paper label marked "589-005-045 ( ARES ROHS )  and under the label the 2Wire Logo is etched onto the chipface

You can download a large scale photo of the chip and it's markings from  HERE


and this is what the 2700HGV looks like inside

[Tacitus]

The 2700 uses it's own ADSL modem chipset, blown onto an ARES chip.

That bears out what I found here.

[MisterW]

So, we now have two routers - different makes and chip - both of which function perfectly well, but not on my line.  I should add that both mine and my sister's lines are iDNet and both run Max, not 21CN.

I really can't believe that 2 different routers can be faulty in a way that causes them not to work on your line but to work fine on another line. Common sense would say that its an exchange fault, but I guess BT have a distinct lack of that!. Going back to your earlier post regarding the 'Suspicious VXCO offset' report on the 2700, my guess would be that there is possibly a fault on the line card affecting the frequency and that some modems (and BT's test gear?) can cope with the fault better than others...

[Rik]

Just taking up Riks point about prices going up in Jan due to the VAT rise. The cost then would be £234 not £239 as Rik said, he was using MK maths 

I was putting my markup on, Den.

[Tacitus]

I really can't believe that 2 different routers can be faulty in a way that causes them not to work on your line but to work fine on another line. Common sense would say that its an exchange fault, but I guess BT have a distinct lack of that!.

Agree entirely, which is why I went for the engineer visit.  I wouldn't have believed it, but that's what the tests show.  Given iDNet support never mentioned 'incompatibility' until post the engineer visit, they presumably discounted it on the grounds that one might fail but it's unlikely that it would apply to two different routers.

Going back to your earlier post regarding the 'Suspicious VXCO offset' report on the 2700, my guess would be that there is possibly a fault on the line card affecting the frequency and that some modems (and BT's test gear?) can cope with the fault better than others...

Oddly enough the current 2700 (V6) which has been pretty stable for a while now, reports the VCXO offset as being -112.6 ppm.  The old (failed) one reported it as -120 so I wonder if that represents a trigger point.  According to Zyxel their modems don't use VCXO offset, so that would discount it as the source of the problem for the 662H.

The failed 2700 has been running on Sis line since Sunday lunch with no apparent problems - at least I've not had an anguished phone call yet....