Print

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

[spitfire]

Greetings Old Friend ... if you still remember me. 

I'm involved in a debate regarding the old chestnut of the pros and cons of shutting down your router nightly prior to going to bed.

As usual, some say Yes; some say No; some say Occasionally; some say that the longevity of routers can be affected by routinely shutting down nightly; some refute this; some say that shutting down only means a marginal increase in boot-up time; others disagree.

If you can spare the time, I'd be grateful for your personal opinion about this. I won't give you my own modus operandi lest it appears I'm fishing for my side of the debate.
It's likely that your members have historically discussed this many times, so you may prefer just to point me to a previous thread.

No rush Kitz ... just when it's convenient for you.

Happy New Year to you and yours.

Spit

[bob.gas]

Hi Spit.
I beat you too it   

[PhilipD]

Hi

Greetings Old Friend ... if you still remember me. 

I'm involved in a debate regarding the old chestnut of the pros and cons of shutting down your router nightly prior to going to bed.

As usual, some say Yes; some say No; some say Occasionally; some say that the longevity of routers can be affected by routinely shutting down nightly; some refute this; some say that shutting down only means a marginal increase in boot-up time; others disagree.

If you can spare the time, I'd be grateful for your personal opinion about this. I won't give you my own modus operandi lest it appears I'm fishing for my side of the debate.
It's likely that your members have historically discussed this many times, so you may prefer just to point me to a previous thread.

No rush Kitz ... just when it's convenient for you.

Happy New Year to you and yours.

Spit

The light bulb argument often comes into play here, where people point out a light bulb nearly always fails when it is turned off then back on again, therefore it will last longer left on.  This is because tungsten bulbs have an inherent problem being turned on, as the tungsten when cold has a lower resistance, so it takes a hit when turned on as surge of current flows until it lights up white hot where the resistance drops.  The bulb analogy is then applied to all manner of other electrical products.  (LED bulbs often fail at switch on as well, but this is due to a surge of current caused by the cheap methods used to drop the voltage in a lot of LED bulbs, notably the cheapest ones.)

Modern electronic gadgets don't experience this surge of current as the power supplies prevent it, as otherwise the electronics would be fried the first time they were switched on!

Another argument is the heating and cooling of turning a router/modem on and off cause some sort of mechanical fatigue over time due to expansion and contraction, so best left on.  Again this isn't really relevant in our modern world.  Circuit boards and components get soldered on the board by the whole thing being baked at 250 degrees or more, sometimes more than once, so a product warming up and down between 20 degrees and say 60 degrees when on is nothing.  However, there is one component that hates being warm and that are capacitors, so leaving a device on that is running warm as most routers do, means the capacitors could be at their end of life after a year or two.  The irony here is that a failing capacitor will often cause problems turning a device on, which perpetuates the myth that a device is more likely to fail being turned on and off.

As far as modern routers are concerned, turning them on and off doesn't age them, it may even prolong their life given electrolytic capacitors age quickly when warm running 24/7.

The main issue is with DSL quality management, as people turning their VDSL modems on and off mean other users see time-varying crosstalk, which could cause loss of service and varying performances and download speeds.

You will often read of people knowing when someone has gone on holiday for a couple of weeks as their margin and sync speed goes up, imagine if 150 users in a VDSL cabinet were all randomly turning their modems on and off, it would be a mess for the DSLAM to manage and would make technologies like vectoring impossible to implement as conditions would be constantly changing.  This is why they say keep the modem connected and turned on all the time.

Regards

Phil

[tubaman]

I'm afraid I can't agree that thermal cycling is not relevant to modern electronics.
It absolutely does effect solder joints - not just those visible on the PCB but also within the components themselves.
Just Google "solder thermal cycle failures" or similar and look at some of the research papers that come up.

In a domestic setting the effect is likely to be small, but it does exist and it does lead to failures.
A thread in the Computers and Hardware section (http://forum.kitz.co.uk/index.php/topic,17454.msg318783.html#msg318783) highlights where a graphics issue has been fixed by baking a PCB in oven. The root cause here is most likely solder cracks due to thermal cycling.
 

[Chrysalis]

thermal cycling was actually the prime cause of the xbox 360 RROD's.

[PhilipD]

Hi

I'm afraid I can't agree that thermal cycling is not relevant to modern electronics.
It absolutely does effect solder joints - not just those visible on the PCB but also within the components themselves.
Just Google "solder thermal cycle failures" or similar and look at some of the research papers that come up.

In a domestic setting the effect is likely to be small, but it does exist and it does lead to failures.

It is relevant but shouldn't be a problem unless there is a design flaw, and it is simply something that people don't need to worry about and can't be used to decide if leaving an appliance on 24/7 or turning it off is better or worse.  A lot of modern electronics will be turning on and off various circuits as required anyway in order to reduce power consumption so even if left on they are constantly cycling various components between being hot and cold.  What about a TV left in standby, it is off essentially and it completely cools down, to some people though because it is plug in and turned on that to them is leaving it on 24/7.

thermal cycling was actually the prime cause of the xbox 360 RROD's.

No it wasn't thermal cycling that caused the problem, it was a manufacturing and design defect which meant it could fail during normal use, which involves heating up and cooling down.  Look at this way, there are billions of CPUs and chips soldered to circuit boards in all manner of products which never fail in that manner.  Billions of Smartphones have CPUs that are constantly going from room temperature up to 60 or 70 degrees then back down again, soldered using the same methods as that X-Box in question that don't fail. 

There is no one answer fits all in terms of longevity of the equipment in question if it is better to leave it on 24/7 or turn it off and on, but in terms of VDSL, it is better for the everyone if the modem is left on to avoid moving goal posts in terms of cross talk and spectrum management. 

Regards

Phil

[Chrysalis]

Yes I know it was poor hardware design, but it was thermal cycling that stressed that design to fail point.

Thermal cycling increases stress of components.

Its the same way that the original comet 1 planes failed in that air pressure cycling triggered those failures in mid air.

[PhilipD]

Hi

Of course thermal cycling stresses components, but my point is all electronic products are designed so they aren't stressed enough to fail from it, as otherwise it costs the manufacturer money and lost sales.  The X-Box, like any PC really, experiences thermal cycling, regardless if left on 24/7 or not, so the fact it failed wasn't because of thermal cycling (that's a given, you can't remove that from the equation), it failed because of a design flaw, and that flaw can't be extrapolated to other electronic products, well it can, and if it is there is born an urban myth.

Generally devices best fair based on their target usage, so a modem/router expected to be left on 24/7 is tested for that scenario, but the irony here is because they can't wait 1-2 years to see if it would last past the warranty period running 24/7, means they are tested in ovens and exposed to big swings in thermal cycling to accelerate the aging.

So turning off a router/modem each night, it's a poor and weak argument to say don't do it because it will fail sooner, and conversely, the electrolytic capacitors which usually fail in domestic routers/modems will last longer by virtue of being exposed to less heat, so you could argue the opposite.

VDSL modems should be left on 24/7 to avoid problems with time-varying cross talk, and to allow the DLM to manage user connections based on a stable cross talk environment.

Regards

Phil

[Chrysalis]

well they will all eventually succumb to it, they more designed that just outlast the warranty period.