Switching Off Router When Not In Use

[#1]

In the past I always switched my router off when I was not using my computer. Then one day my SNR shot up to 15dB and my sync speed correspondingly dropped. One of the early suggestions by the CSC was that the cause could have been caused by my switching on and off being interpreted by BT as the line being dropped because of a poor connection.
Having now got my SNR back to a sensible level I have been leaving my router switched on 24/7 for fear of it all starting over again. However, I have measured the power consumption of the router and it takes 12W continuously. Not a lot I agree but it represents 3% of my total daily consumption, which I am trying to minimise.
I have read elsewhere that one switch off a day is good. Does any-one actually know what the truth is? Is it for example safe to switch a router on and off say twice a day without the SNR turning to rats. Common sense says it should be OK else surely every-one using a USB modem would be permanently stuck with a low SNR.

[#2]

I think the general rule is that one or two power-offs of a router per day is fine.  Any more than 2-3 per hour is definately NOT fine.
B.

[#3]

Stick to one power-cycle per day if you must, i.e. turn it off overnight, but don't make it any more frequent than that.  Apart from risking a downwards re-profiling, the router's PSU is designed for continuous running and too many power-cycles can shorten it's life.

[#4]

In the past I left my router on all the time, but over the summer, I found that I must have been having sync problems (mainly over night), which caused my profile to drop (often to 500k), so I've started trying to remember to turn the router off at night, and that seems to help.
The problem goes back years, on dial up I often found that I couldn't get internet access between about June and September!  BT, as you would expect, fell back on the "we only guarantee voice" argument.  I suspect a dodgy connection in the copper wire, which flexes as the ground dries out in the summer, and as it cools overnight.
Anyway the point is, I turn my router off over night, and get a better profile as a result.
John

[#5]

I always switch mine off overnight and its never affected my sync speed. Once or twice a day is fine, tho' I've heard that too often could be interpreted as a line fault.
I agree with the poster who said the main long term danger is PSU related. Its best to avoid too many power surges caused by switching on/off. (As with your computer's PSU).
But you have to balance that against the minuses of  leaving unnecessary electrical gear turned on: cost & environmental...

[#6]

You realise that your router is costing you 0.18p per hour?
or about £15 if you leave it on all year.
See http://www.guardian.co.uk/money/gallery/2008/aug/01/energy.householdbills?picture=336141363

[#7]

The point that baldrick makes about the USB connection is one that has puzzled me. When I first went onto broadband and used the Voyager 105 (which I very much liked), I was always switching it on and off (didn't know any better) and can't remember having progressive loss of speed - indeed, it was extremely stable and quite satisfactory.
However, the speed is approximately double what it was now that I have a wireless router and leave it on all the time. Only thing is, the comparison is probably not sound because there has been some tweaking at PN towers and if I understand it correctly, USB is not as good as ethernet/wireless.
BTW is it true that it is better not to switch the router on and off to save wear and tear on parts?

[#8]

Quote from: poppy

.......... BTW is it true that it is better not to switch the router on and off to save wear and tear on parts?

With one exception, the electronic components of most modern products should last fot over 100,000 Power On Hours once they have survived the first 1000 hours.
The exception comes from a manufacturing problem in the mid to late '90s affecting electrolytic capacitors made in mainland China. The problem arose from  the defection of a Tiawanese chemist who took with him the "recipe" for the composition of the Electrolyte but got it wrong!
The most common failure today is Electrolytic Capacitors dislaying blown cans. You can see the problem if you examine the tops and bottoms of the cans where the expansion and venting of the electrolyte is visible.
So. whereas it is true that it is better NOT to thermally cycle electronic equipment , the Capacitors may survive longer if switched off when not in use.
..... and then there is the "Tin Whiskers" phenomenon ..... but that is another story....