Print

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

[crgbt]

Hi all,

I have been experiencing an odd issue with DrayTek Vigor 130 modems for a while now and thought I'd ask the community for feedback.

I have a VDSL line from A&A, and I use a DrayTek Vigor 130 modem on the line with a FB2900 (or Netgate SG-3100). After around 5 to 6 months, the DrayTek's will start to fail by producing thousands of Error Seconds (ES) and hundreds of Serious Error Seconds (SES) every day, no matter what DLM throws at the line. Replacing them with another model of modem or a brand-new Vigor 130 resolves the issue right away.

I'm on my fourth Vigor 130 right now, as one just died today. I've purchased a VMG1312-B10A from eBay (judging by the sellers knowledge I'd not be surprised if they are a member here, too!) with the hope that it's more robust than the Vigor 130's.

The modems have always been on a UPS; originally on a cheap stand-by type UPS, now on a line interactive UPS (https://powerwalker.com/?page=product&item=10120080&lang=en), so I'm quietly confident it's not power related.

Does anyone have any suggestions as to what may be causing the modems to fail in the way they are? I do have a very noisy line, and with my limited knowledge I'm thinking it may be due to that, but I could well be wrong.

Thanks for looking
crgbt

[burakkucat]

Welcome to the Kitz forum. 

You have described a rather odd situation.

What have you done with the "dead" Vigor 130s? Returned them to DrayTek for anaylsis? Having installed three new replacements, as each predecessor has "died", I would be carefully checking the environment in which the device is operated. For example --

• Good ventilation?
• Spurious (abnormal) voltages on the VDSL2 circuit?
• Spurious (abnormal) voltages on the LAN port?

I am puzzled. 

[hopkins35]

I also had trouble with two Vigor 130 modems on two Andrews & Arnold bonded FTTC lines on a Hauwei cab, would be stable for a week or so before throwing a tantrum and disconnecting 20-30 times in a 24 hour period although I found that a reboot would stabilise them again for a period of time. Like you I switched to two VMG1312-B10A and they've been much better for the last six months. I concluded that the Lantiq chip in the Vigor 130 just couldn't handle a noisy line on a Huawei cab

[crgbt]

Welcome to the Kitz forum. 

Thanks

What have you done with the "dead" Vigor 130s? Returned them to DrayTek for anaylsis?

Yep. All of them (save the latest, I have a case open with DrayTek) have gone back for replacement. Each time I get the same housing but a new board inside.

Having installed three new replacements, as each predecessor has "died", I would be carefully checking the environment in which the device is operated. For example --

• Good ventilation?
• Spurious (abnormal) voltages on the VDSL2 circuit?
• Spurious (abnormal) voltages on the LAN port?

I did think about ventilation at first. In my old property, I had my first Vigor 130 (and FB2900/SG-3100) in far worse conditions - it was in the cupboard with my boiler previously, so quite a wide fluctuation in temperatures and they never skipped a beat.

Right now, all equipment (modem, router/firewall, rpi, UPS) is inside a kitchen cabinet which is rather small, but I have drilled ventilation holes and fitted a Noctua NF-A8 5V (https://noctua.at/en/nf-a8-5v) as an exhaust fan which is keeping the cabinet at ambient roof temperatures (18-23 degrees C depending on the weather).

Do you know of a way that I'd be able to measure voltage on the VDSL2 circuit? All of the other equipment is on the same line-interactive UPS which has a built in AVR, so I'd hope it wasn't voltage fluctuation on the LAN side but if you know if a way to measure this it's definitely something I'd look into doing, likewise with the VDSL2 circuit.

I am puzzled. 

Me too!

I also had trouble with two Vigor 130 modems on two Andrews & Arnold bonded FTTC lines on a Hauwei cab, would be stable for a week or so before throwing a tantrum and disconnecting 20-30 times in a 24 hour period although I found that a reboot would stabilise them again for a period of time. Like you I switched to two VMG1312-B10A and they've been much better for the last six months. I concluded that the Lantiq chip in the Vigor 130 just couldn't handle a noisy line on a Huawei cab

As much as I'd like this to be something I can bottom out and resolve as I do like the Vigor 130's features, right now I'd settle for the noise on my line to be the issue and using a different modem is the answer. The Vigor 130's are not cheap, and it's a hassle to keep RMA'ing them!

Thanks all

[burakkucat]

As for spurious voltages, I was thinking of the extreme error condition when 240V AC has been found to be present on the Openreach pair.

All modems (and modem/routers) have a high-pass filter on the input xDSL port. At the minimum this would be a pair of capacitors, one in each leg. For the AFE (analogue front end) to be "zapped", an alien voltage would have to be significantly excessive and prolonged. Such voltage could either be across the pair or between one leg and earth. The likelihood of that being the cause of your problem is quite minimal but it is something I would check. (A DVM set to a sufficiently large AC voltage settings, one probe to a good earth the other probe to each leg of the pair, in turn. Then with both probes, check across the pair. If that shows nothing abnormal, select DC voltage appropriate for c.a. 50V. Repeat, as before. One probe to a good earth, check each leg of the pair. Then check across the pair.)

[crgbt]

As for spurious voltages, I was thinking of the extreme error condition when 240V AC has been found to be present on the Openreach pair.

All modems (and modem/routers) have a high-pass filter on the input xDSL port. At the minimum this would be a pair of capacitors, one in each leg. For the AFE (analogue front end) to be "zapped", an alien voltage would have to be significantly excessive and prolonged. Such voltage could either be across the pair or between one leg and earth. The likelihood of that being the cause of your problem is quite minimal but it is something I would check. (A DVM set to a sufficiently large AC voltage settings, one probe to a good earth the other probe to each leg of the pair, in turn. Then with both probes, check across the pair. If that shows nothing abnormal, select DC voltage appropriate for c.a. 50V. Repeat, as before. One probe to a good earth, check each leg of the pair. Then check across the pair.)

Thanks for the suggestion, burakkucat. I'm waiting for confirmation of a DLM reset on my line, and I don't want to upset DLM until I know when the reset is going to take place so I'll wait for this to be confirmed and get the test done the day before. I've had a few suggestions from the person I bought the VMG1312-B10A from too, such as checking output voltage on my UPS, ensuring the power strip on the UPS is not surge-protected, and potentially using a different power adapter.

I do have higher than average voltage (248-252V) at my property but it's within limits according to Northern Power Grid, and as my UPS has an AVR built in I'd hope it would buck the power down to an acceptable range should it go over the 253V limit!

Also, I just noticed in your signature about your involvement in the ELRepo. I just want to say thanks - I've used that repo quite a lot in the past.

crgbt

[burakkucat]

Do keep us updated, please.

Also, I just noticed in your signature about your involvement in the ELRepo. I just want to say thanks - I've used that repo quite a lot in the past.

The kuro neko performs one of his best Japanese-style bows and blushes, whilst acknowledging your message. 

[crgbt]

Do keep us updated, please.

So the plot has thickened somewhat.

I've been monitoring the Vigor 130 stats since the first one went pop using a handy tool called telnetd (https://coherent-ascent-678.appspot.com/), as my scripting skills leave a lot to be desired.

As I mentioned, my line was stable (~30 ES/~1 SES per day) following the Vigor 130 being replaced. That was until around 03:00 this morning, when everything went haywire again and I was getting thousands of ES on the line. I watched the counters go up for about 30 minutes and it seemed like every second had an error, so I decided to go and carry out the steps burakkucat (and the eBay seller) advised, the results of which are:

Input voltage to the UPS: 248.1V
Output voltage from the UPS: 247.9V
Output voltage from the 12V modem adapter: 12.3V
AC voltage from either leg to earth on the telephone pair: ~0.5V
DC voltage across the legs on the telephone pair: 50.2V

The total downtime was ~6 minutes on the line. I didn't have another spare modem, so I plugged the Vigor 130 back in. Now this is where it gets weird - there have been no ES or SES since plugging the modem back in at 06:21. I feel like it's too much of a coincidence that a fault on my line stopped during those 6 minutes, and I'm leaning towards there being a certain fault condition (or just the condition of my line?) that sets the Vigor 130 off.

This also brings into the question: were my previous Vigor 130's actually faulty, or did I just leave them when testing long enough for the fault condition to reoccur? Also, how come they seem to start generating errors after ~6 months, when being rock solid up to that point?

I'm very confused! I have a case open with A&A anyway, so I'll see what they have to say on the matter, and if they can do a more thorough line check than that is available in the control panel.

Any other suggestions in the meantime would be most welcome!

Thanks,
crgbt

[jack21]

Though you had only recently put a new Vigor 130 into service, I recall you saying you had not replaced its PSU on this occasion....and now have experienced the fault after only a few days of service, rather than the 'usual' 5 or 6 months. I'm just wondering if the PSU could be a problem-causer, rather than the modem or line??

[crgbt]

Though you had only recently put a new Vigor 130 into service, I recall you saying you had not replaced its PSU on this occasion....and now have experienced the fault after only a few days of service, rather than the 'usual' 5 or 6 months. I'm just wondering if the PSU could be a problem-causer, rather than the modem or line??

I'm going out on a limb here, I'm going to say you are eBay Jack? 

You are absolutely correct - I've not swapped the PSU this time. I'm stuck with these two PSU's unfortunately, as looking in my the box from the last replacement the PSU wasn't replaced. As I mentioned via eBay message, I have a 12V DIN-mounted PSU somewhere in the loft which I'm going to try (making sure I have the +/- the right way around!), but I've not had a chance to dig this out yet.

DrayTek have also sent me 4 different firmware versions to try, each with differing modem code. I don't particularly want to upset DLM, but I'll be giving each one of those firmware's a try over the next 4 weeks.

Thanks,
crgbt

[jack21]

Yes, you're right Crgbt, we happily dealt on eBay a few days ago.
You'll be OK DLM-wise with modem-firmware swaps, as long as you keep the swap frequency low....and in any case you'll need to allow some reasonable time to check each one.
But do consider giving the B10A a whirl, together with DSLstats, to look for further performance evidence/clues.

[burakkucat]

Input voltage to the UPS: 248.1V
Output voltage from the UPS: 247.9V
Output voltage from the 12V modem adapter: 12.3V
AC voltage from either leg to earth on the telephone pair: ~0.5V
DC voltage across the legs on the telephone pair: 50.2V

I see nothing to be concerned about. The voltages present on the xDSL/telephony pair are exactly as expected.

Does the UPS have an adjustable output? If "yes", then a shift one increment downwards might be worth considering.

[crgbt]

Yes, you're right Crgbt, we happily dealt on eBay a few days ago.
You'll be OK DLM-wise with modem-firmware swaps, as long as you keep the swap frequency low....and in any case you'll need to allow some reasonable time to check each one.
But do consider giving the B10A a whirl, together with DSLstats, to look for further performance evidence/clues.

I'll certainly be giving it a try. As I said on eBay, I'm hopefully getting another line from A&A so I'll at least have a second line to keep working on whilst I'm testing modems on another.

I see nothing to be concerned about. The voltages present on the xDSL/telephony pair are exactly as expected.

Does the UPS have an adjustable output? If "yes", then a shift one increment downwards might be worth considering.

Good news, thanks. Unfortunately the UPS is on the cheaper end of things, and has no such feature. I do have a DIN-mountable 240V>12V power supply in the loft which I'd like to try, however looking at it's spec sheet it has active PFC and I'm not sure if this will interfere with the UPS? The manufacturer does say not to put surge protectors on either the input or output, but nothing specifically about an active PFC PSU.

Thanks again all
crgbt

[sevenlayermuddle]

Where are you keeping these routers, and how hot are they getting?

There’s many possible reasons for failure of modern electronics, but probably the most common single reason is failure of aluminium electrolytic capacitors.   There’s lots of nonsense talked on the web about a ‘capacitor plague’, of duff bootleg components from the early 2000’s. There may be a shred of truth in these reports but in fact, premature capacitor failure is nearly always caused by devices getting too hot.  Even the best caps have a rated life measured in months not years, if operated at or near their temperature limits...

Just a thought.

Edit... sorry, PS:  I see earlier, you answered my question, they are in a fan-cooled cupboard.  All the same, I’d be tempted to try them outside that of cupboard in open air for a while, see if failure rate improves.

[crgbt]

Where are you keeping these routers, and how hot are they getting?

There’s many possible reasons for failure of modern electronics, but probably the most common single reason is failure of aluminium electrolytic capacitors.   There’s lots of nonsense talked on the web about a ‘capacitor plague’, of duff bootleg components from the early 2000’s. There may be a shred of truth in these reports but in fact, premature capacitor failure is nearly always caused by devices getting too hot.  Even the best caps have a rated life measured in months not years, if operated at or near their temperature limits...

Just a thought.

I keep them in a kitchen cupboard, and when I first moved in to this property I had zero ventilation and things got hot, although not as hot as when I had the same kit running in an airing cupboard in my old property. Since then, I drilled an 80mm hole in the bottom of the cabinet and fitted a Noctua NF-A8 5V (https://noctua.at/en/nf-a8-5v) fan, along with ventilation holes in the top of the cabinet and little wedges to keep the cabinet door open a few mm. The cabinet inside walls currently measure 28.1 degrees and the Vigor 130's chassis measures between 30.4 and 31.7 degrees, depending on where I measure it. The ambient temperature in my house is 23 degrees, so I don't think that's too bad a temperature.

Edit: I should note that I've had Vigor 130's fail since putting in the fan

What is your take on this setup?

For what it's worth, I used to modify Netgear DG834GT's back in the day with more robust caps for friends/family as I'd had many of them go bad, so I get where you are coming from re: cap failure.

Thanks for the feedback!

crgbt