Do routers wear out?

Some older Linksys routers were good (I just got an older WRT54G v2 -- see my other post), but maybe it's a trend that the new stuff is not so good. We recently got a brand new Linksys wireless router at work, and we just couldn't get it to work correctly, so we sent it back and got a different brand, and the replacement brand worked fine (I don't remember what brand it was).
--

• Eric, http://www.InvisibleRobot.com/

linuxworks said:

I am using the freebsd 'm0n0wall' software on some embedded hardware.
I think monowall also runs on that 54g box (?).

Click to expand...

No 'm0n0wall' for WRT54G, but there are a number of other choices.

--

• Eric, http://www.InvisibleRobot.com/

Technology yes but sloppy English never seems to to improve. Modem/routers "slowing down" is slop speach for lots of problems. Even the cable used by some providers is so poorly manufactured and or installed that signal strength has been know to slowly drop to marginal levels over a period of months. After much badgering my former provider finally sent a tech out with a field strength meter. Presto new cable installed. More than a 12db jump. So lots of possible causes for marginal performance i.e "slow down".

I have had far better performance out of Lynksis units than any other I have used. So no one brand seems to stay reliable. Too bad.

Paul Stricklin
--
Club, Event Photographer to pay for the equipment
Nature, Landscape for the joy of life

LPS

Software based wireless routers in the long run are cheaper. Who would want to keep changing routers due to chip changes? I've had my DSL modem and wireless router running 24/7 for the last 8 mos and it's still working. It's also recommended to turn off/restart the modem if you feel the speed is slowing down.

tocar said:

it's still working. It's also recommended to turn off/restart the
modem if you feel the speed is slowing down.

Click to expand...

are the datacomm boxes in consumerland THAT badly done that they have memory leaks (when you allocate a chunk of memory for a task and then 'forget' to free it. this builds up over time in less than perfect software).

another thing that MIGHT be happening is that some isps are evil. they block user traffic, they insert their OWN throttling 'resets' in the line and do all kinds of 'traffic engineering' to lower your data rate ;(

they could VERY EASILY be lying to you just to get rid of you - to get you to stop complaining (the OP). if you start using the cable and start doing p2p, they will likely start to throttle you.

isps over subscribe. heavily. so they have to play dumb games on users to stop them from getting the b/w they actually bought! but in this case, its NOT the router's fault. in fact, its THEIR routers that play these throttle and NAK/reset games.

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

Let me speculate why I think that he might be right. When I was working I was in charge of research and one of my groups was working on the development of electronic materials that are used for the production of resistors and capacitors that are parts of most electronic gadgets. We have to carry out test aging of components produced from these materials (among others their value change when exposed to higher temperatures). Materials developed for military applications had to meet much more stringent requirements then those for consumer products.

It has also been my experience that a couple of years old laptops will never perform like when they were new, even if software is re-installed to the original stage. That same is NOT true for desktops - and we all know that heat management of laptops is quite tricky (compared to desktops). So I would speculate that if heat is not properly removed from a gadget its electronic components "move" out of its optimal values resulting in the gadget performing worse then when it was new.

Another example:most a few years old car radios do not sound as good as when they were new because their components are exposed to quite high temperatures and I would suggest their values are not anymore at their optimum (the same is not true for a desktop radio - even 10 years old radio can sound like a new one).

These are just speculations - let me know what you think about them.

mino

mino said:

It has also been my experience that a couple of years old laptops
will never perform like when they were new, even if software is
re-installed to the original stage. That same is NOT true for
desktops - and we all know that heat management of laptops is quite
tricky (compared to desktops). So I would speculate that if heat is
not properly removed from a gadget its electronic components "move"
out of its optimal values resulting in the gadget performing worse
then when it was new.

Click to expand...

I would buy that for analog (amps and preamps change over time) but not digital. I just don't buy that, sorry.

mino said:

Another example:most a few years old car radios do not sound as good
as when they were new because their components are exposed to quite
high temperatures and I would suggest their values are not anymore at
their optimum (the same is not true for a desktop radio - even 10
years old radio can sound like a new one).

Click to expand...

you are now talking analog, and so I now agree with you

mino said:

These are just speculations - let me know what you think about them.

Click to expand...

just did.

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

Thanks for your comments. I agree with you and understand that digital circuitry should not be affected by minor changes, and some major changes in values would probably shut the whole circuit off.

That still leaves me with the problem why my old 3 laptops cannot be brought to their original performance and my 7 years old Sony desktop, after re-installing OS works like it used to.

mino

Most of the consumer grade stuff like your off the shelf linksys or netgear router are bound to die. They are not made to last. I opened up my routers and was surprised at the poor quality of the components, some capacitors were bloated after a year of use, some sealants and glues had heated/leaked and had flowed.

I've never had much luck with linksys or netgear. The Buffalo router I picked up is doing fine, though, they seem to be much better made.

mino said:

Thanks for your comments. I agree with you and understand that
digital circuitry should not be affected by minor changes, and some
major changes in values would probably shut the whole circuit off.

Click to expand...

right. if the timing is so bad, it won't work at all. if the timing is close enough, it 'locks' and works. some jitter is allowed for, in reasonable digital designs. and component aging should also be allowed for (in quality designs).

mino said:

That still leaves me with the problem why my old 3 laptops cannot be
brought to their original performance and my 7 years old Sony
desktop, after re-installing OS works like it used to.

Click to expand...

have you quantified that? in which areas? memory speed? cpu speed? disk speed?

disks add errors over time. those DO get slower over time due to having to 'walk over' bad sectors and such.

ram does not go bad 'slowly' - its a working or non-working thing.

cpus also don't change speed. they might have higher error rates but that should be obvious to the user.

network cards, disk controller cards, etc - I can't think of those things 'slowing down'.

so I'm not sure what really slows down for you. have you isolated it?

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

Any device that depends in part on RC Time Constants will be affected by temperature related aging..Army Quarter Master Research shows that a semiconductor device is "more likely" to keep running the longer it has been running not that it will maintain optimum performance and that temperature has a direct effect on electronic components even if it is small at room temperatures.

Electronic Instructors were know to state emphatically that resistors do not short. So we kept a trophy board of shorted resistors in the repair shack.

Paul Stricklin
--
Club, Event Photographer to pay for the equipment
Nature, Landscape for the joy of life

LPS

Paul Stricklin said:

Any device that depends in part on RC Time Constants will be affected
by temperature related aging..

Click to expand...

in the analog domain, you have a smooth degradation curve.

in digital it works fine up until the point it doesn't work at all.

so again, I don't see any logic to 'computer slowing down' due to age, at least at the hardware level.

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

I have not tried to quantify it - all three laptops are quite old, but I know that Toshiba (1998) was quite fast and now it takes a few minutes even to start the OS. Two Dells are about 7 years old and even after re-installing OS are quite slow (definitely slower then I remember them to be). I am keeping them just to play with Linux.

On the other hand my Sony Vaio (I just realized that it is 10 years old) after re-installing the original software is still quite fast - certainly usable for some basic picture and video editing. I would not even try using Adobe Elements on any of the laptops and I used to use it.

I understand your points and my observations might be quite coincidental, and certainly unscientific. But in my experience the useful life of laptops is in general shorter than that of desktops, but it might be for different reasons then heat management.

mino said:

I have not tried to quantify it - all three laptops are quite old,
but I know that Toshiba (1998) was quite fast and now it takes a few
minutes even to start the OS.

Click to expand...

unless you timed it at 'birth' and also have an identical disk image now at 'death' - its just a gut feel and I'm still not convinced.

mino said:

Two Dells are about 7 years old and
even after re-installing OS are quite slow (definitely slower then I
remember them to be).

Click to expand...

still its human perception. 'remembering' startup times is not quite scientific, is it?

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

It is possible.

Also, I have had experience recently of a network card that went bad. It didn't completely fail, it just was slow, or periodically slow (can't remember). Maybe similar to your complaint.

I would try a different port, I would try a different computer, and a different cable (if it isn't wireless). I guess I would try a cable if it was a wireless issue.

Don Glenn said:

It is possible.

Also, I have had experience recently of a network card that went bad.
It didn't completely fail, it just was slow, or periodically slow
(can't remember). Maybe similar to your complaint.

Click to expand...

that's a different kind of slowness if you get high error rates you will back off and retry. if you get noise on the line (or cause noise) you will also back-off and retry, exponentially (by ethernet).

that's NOT slowdown due to 'aging' which is the thing I'm arguing about.

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

If there truly is a slowdown in the laptops, it could be the drives having to do more retries (Did you test with new drives, or at least run SpinRite 6 on the old drives?), or it could be a dirty or mis-functioning CPU cooling system, like bad fans, bad heat pipes, bad thermal compound, loose or misaligned thermal contact area, etc. causing the CPU speed to throttle due to heat issues. Or maybe a combination of those and other things.
--

• Eric, http://www.InvisibleRobot.com/

all really good points.

I didn't think about speed throttling. good!

disk retries - yup, could be that, too.

still not based on 'component aging' though fans get clogged and need replacing. drives need replacing.

Eric Carlson said:

If there truly is a slowdown in the laptops, it could be the drives
having to do more retries (Did you test with new drives, or at least
run SpinRite 6 on the old drives?), or it could be a dirty or
mis-functioning CPU cooling system, like bad fans, bad heat pipes,
bad thermal compound, loose or misaligned thermal contact area, etc.
causing the CPU speed to throttle due to heat issues. Or maybe a
combination of those and other things.
--

• Eric, http://www.InvisibleRobot.com/

Click to expand...

--
Bryan
(pic stream: http://www.flickr.com/photos/linux-works ) ~

I remember long time ago reading in an audio buffs magazine about audio speaker cables and how over time the electrons inside get lazy, and the sound quality no longer is what it used to be when the cables were brand new.

(of course no one but the the real sound experts would hear any difference. ordinary people were not skilled enuff to make any judgement whatsoever.)

the solution that was suggested to cure the lazy electrons was a device that gave the electrons a jolt of electricity, and flushed out the lazy electrons and replaced them with fresh ones. the procedure had to be repeated at specific intervals, to keep the quality of the sound at superior levels.

In the same magazine I was reading many advertisements about different brands of really advanced and quite expensive speaker cables. many of them had a water jacket, that is they basically consisted of a speaker cable inside a tube filled with some liquid.

when reading the finer print with contact information etc I concluded that these very expensive speaker cables were manufactured in mom's kitchen by some entrepeneaurs that had figure out a way of fitting a speaker cable into a water hose, filling it with some liquid and making the assembly water tight enough so they could be shipped to customers.
I have always wondered how many people bought these special made cables.

Tage

--
http://www.thisbeautifulearth.com

I think that there are a lot of people buying expensive HDMI Monster cables to connect digital devices, even that it is not necessary.