Computing desk
< September 3	<< Aug \| September \| Oct >>	September 5 >

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

September 4

Surge protector for WiFi hub

We have a household BT Wi-Fi router which has a tendency to reset itself occasionally. After much head-scratching, I came to the conclusion that it was doing this when the electric oven in the next room was switched on. Obviously, there is some sort of momentary fluctuation in the mains power supply. I wondered whether connecting the router power supply to the mains through a surge protector would be likely to help, or whether there were alternatives? This is for domestic use only, so an expensive uninterruptible power supply etc would be overkill. AndyTheGrump (talk) 11:32, 4 September 2011 (UTC)[reply]

DSL is pretty susceptible to electromagnetic interference. The extreme example is a thunderstorm, where I would expect it to reset many times. So it may not be the mains, but could easiliy be wireless interference. Do any of you extension phone lines go near the oven? Can you try temporarily disconnecting them at the master socket and see if matters improve? --Phil Holmes (talk) 14:00, 5 September 2011 (UTC)[reply]

You could have a loose wire in either outlet, or your breaker box, or even somewhere else. If you go looking for it yourself, do be careful not to poke anything that could possibly have an electrical current running through it with anything that could conduct that current to yourself, etc.. :) It's possible that plugging the router into an outlet in a different part of the building could rule out problems with certain connections (or confirm that it's in the breaker box). ¦ Reisio (talk) 12:07, 4 September 2011 (UTC)[reply]

Unfortunately, the router can't really be moved anywhere else, and the oven is permanently wired to the mains - so I doubt that there is a loose wire anywhere. I shouldn't have to pay more than about £5 for a single-socket mains surge protector, so I may as well give it a try anyway, I suppose. AndyTheGrump (talk) 20:23, 4 September 2011 (UTC)[reply]

What you actually want is a small uninterruptible power supply rather than a surge protector. I think daily electrical surges would have fried your router's power supply by now, whereas if it's a voltage dip or a current outage then a UPS will save the day. Comet Tuttle (talk) 22:50, 4 September 2011 (UTC)[reply]

Yes, you may well be right - I've read up on this a bit more, and as you say, it may be a dip or outage that is the problem. The cheapest UPS I've found is about £60 though, which is a lot to pay for a domestic system. Still, if it solves the problem, it might be worthwhile - it did it again a couple of hours ago... AndyTheGrump (talk) 23:25, 4 September 2011 (UTC)[reply]

One thing to try first is to run a long extension cord from the router to some outlet on the other side of the house, so the router is obtaining power from some distant room, and see if it still occurs when the oven switches on. It may be that the wiring from the oven to the "fuse box" also reaches to the outlet you're currently using for the router; but other wiring in your house might be on a different circuit and therefore may be unaffected by the oven's operation. If the router never fails again, you might try relocating the router, if possible. (Usually not a great idea to have extension cords habitually strung across the house.) Comet Tuttle (talk) 01:10, 5 September 2011 (UTC)[reply]

I could try that - though it doesn't always do it when the oven is used, so testing might be difficult. I can't relocate the router (a shame, as it would probably work better higher up anyway - this is a three-storey house - but the phone socket is on the ground floor) and extension cords trailing all around the house aren't going to be popular. I'll see if I can figure out how the house is wired though, there may be a simple solution. AndyTheGrump (talk) 01:28, 5 September 2011 (UTC)[reply]

An automatic voltage regulator like [1] would potentially be sufficient but may not be much cheaper. Another thought, presuming the PSU for the router is an external AC/DC adapter as many are, if it's one of the older heavy transformer type then it's possible switching to one of the newer switched mode PSUs that can handle 100-250V (or similar) may work since the PSU may be able to tolerate the voltage fluctuations itself (but this is nearly a complete guess). If you can borrow one from somewhere with the right plug (including polarity) and output voltage and able to supply the necessary current (although it shouldn't be too high, albeit that's less likely to be an issue for a switch mode PSU) you can test the idea. If you do have a SMPS but it can only do 200-250V (or similar) input then switching to one with more universal voltage input may work. However a major spoiler which just occured to me while doing some research for the answer, it seems a lot of ADSL modems use a low voltage AC input (usually from 9-22V or so) rather then DC input, [2] mentions that and this BT one [3] does too and I seem to recall many of the ones I've owned have as well (which is what got me thinking). Since it sounds like you have a wifi router with a built in ADSL modem, then perhaps yours does as well. Nil Einne (talk) 13:41, 5 September 2011 (UTC)[reply]

I'll have a look into that. Another possibly might be repetitive electrical impulse noise (What, no article? see [4]). Apparently you can sometimes track this down with an AM radio - I've got one somewhere... AndyTheGrump (talk) 14:19, 5 September 2011 (UTC)[reply]

Reading the linked ref suggests to me it will only affect the ADSL sync. If you connect to the routers web interface and perhaps ping it constantly, you should be able to see if it's simply losing the ADSL sync or the router itself is resetting/dieing. This will be more difficult if when it dies is unpredictable but should be doable. Alternatively setting up SNMP monitoring if your router supports it as the refs suggests or perhaps simply checking the uptime may work. Nil Einne (talk) 14:38, 5 September 2011 (UTC)[reply]

Er, yes, though I'm drowning in TLA's and ETLA's here. At least Wikipedia has articles on most of them. :) AndyTheGrump (talk) 14:51, 5 September 2011 (UTC)[reply]

Oven connection to mains (now no longer the issue, it seems)

"oven is permanently wired to the mains" does not sound particularly ideal... or, I would imagine, even up to code. I know your circumstances might make this observation irrelevant, just saying. ¦ Reisio (talk) 07:13, 5 September 2011 (UTC)[reply]

This is the UK, and as far as I'm aware it is usual to do it this way (see [5]) - to clarify, there is an on-off switch on the outlet between the mains and the oven, but no plug-and-socket connection. Unless the wiring is seriously wrong, it should be on its own radial circuit from the mains junction box, rather than on a common ring main - which rather implies that plugging the Wi-Fi router in elsewhere in the house may well make no difference... AndyTheGrump (talk) 13:13, 5 September 2011 (UTC)[reply]

In a modern wiring setup, a normal sized electric oven will typically have its own circuit back from the consumer unit, rather than being on a shared ring. Small electric ovens, fixed microwave ovens, and the spark mechanism of gas hob/oven appliances will typically be on the kitchen's plug ring and wired with FCUs. Old wiring schemes might have had the oven share a ring, which isn't a very good thing. 2.122.75.122 (talk) 13:22, 5 September 2011 (UTC)[reply]

(ec with AtG) It the standard UK ring wiring scheme, fitted appliances such as smaller ovens, microwave ovens, washers, dryers, waste disposals, and sometimes fridges and freezers, are wired to the wall with an FCU (fused connection unit) or a SFCU (a switched FCU). Electrically this is identical to the usual BS 1363 plug and socket setup, except that the appliance wire is fixed into the FCU terminals with retainer screws, meaning the appliance can't be unplugged. I can't seem to find a Wikipedia article about FCUs/SFCUs specifically, or their equivalents in other countries' wiring schemes. 2.122.75.122 (talk) 13:19, 5 September 2011 (UTC)[reply]

Also seems to be common or probably the norm here in NZ for the oven to be directly connect to the mains where we don't us BS 1363 (see [6]). Not sure but it may partially be because the normal plug we use is only for 10A which isn't considered sufficient. While there variants supporting 15A and higher, I don't think they're common in houses and may have postdated the installation of ovens. Nil Einne (talk) 13:30, 5 September 2011 (UTC)[reply]

I've now checked, and the oven is on its own circuit, rather than the ring the router uses. Evidently I don't need to worry about the house burning down... AndyTheGrump (talk) 14:23, 5 September 2011 (UTC)[reply]

Yes, that's the normal UK wiring, so this means that the cause is more likely to be the interference caused by the thermostat in the oven. It is probably worn and sparking. (You can possibly hear this, and easily pick it up on a long-wave radio.) You could try shielding the router with a metal sheet between the oven and the router, but the solution might have to be moving the router. You could also try complaining to BT and ask to try a different router. This interference is carried along the mains wiring, so the surge protector might help. Dbfirs 14:44, 5 September 2011 (UTC)[reply]

As I said above, DSL is very susceptible to EMI, and so it may be this that's the issue. --Phil Holmes (talk) 17:38, 5 September 2011 (UTC)[reply]

Yes, it does seem likely that Phil Holmes' explanation is the correct one. You could try covering your router with a metal cage (a metal box or foil will restrict airflow and may cause overheating). You could also try replacing your oven thermostat. If your router really has to stay near the oven, perhaps enclosing it in metal, then wiring (cheap cat5e cable) to a separate wireless router to restore wireless access might be a possible solution, but it seems over-complex. Do try a spike-protector first because the interference might be carried mainly by the wiring. Dbfirs 08:31, 6 September 2011 (UTC)[reply]

html5 web sockets and wirewalls?

how does html5 (supposedly "full duplex" instead of typical ajax clutches) get around the firewall problem? — Preceding unsigned comment added by 86.101.168.7 (talk) 12:30, 4 September 2011 (UTC)[reply]

I found this stackoverflow post with the same question: http://stackoverflow.com/questions/1967943/will-html5-websockets-be-crippled-by-firewalls -- one response says "I'm not an expert (so please check to confirm) but I beleive there will be an UPGRADE mechanism where a regular HTTP connection can be made, then upgraded to a WebSocket, so no existing firewall rules interfere unless they are doing aggressive application level packet inspection. Connections are still initiated by the browser." Could someone here please explain "upgrade" mechanism...? Is this always possible? Can someone describe in theory how/why an upgrade works or is expected to work. I don't know a lot about firewalls and port forwarding and all this, so please assume a low level of knowledge. However I don't understand why web sockets aren't blocked by windows firewall, router-level firewalls, company firewalls, etc... Is htis related to nat traversal? if not, wat is the relationship between this and nat traversal. this is not homework, just curious. Thanks.--86.101.168.7 (talk) 12:37, 4 September 2011 (UTC)[reply]

Imagine TCP/IP connections are like phone calls. One side has to start the connection (make the phone call), but once they're connected, both sides can talk to each other over the connection until one side closes the connection (hangs up the phone).

Firewalls and routers usually always allow outgoing connections (allow you to make outgoing calls), so there's no problem. Firewalls and routers are mainly a problem if a program on your computer needs to listen for incoming connections (needs to answer a ringing phone).

In a usual webpage request, the browser connects to the server, the browser asks for an item, the server sends it, they repeat as needed for other items, then one or both sides close the connection.

In a WebSocket request, the browser connects to the server, the browser sends a special message asking to switch to WebSocket mode, the server sends an acknowledgment, then the rest of the conversation over the connection is in the full-duplex WebSocket mode.

I also found WebSocket.org which links to WebSocket Protocol. --Bavi H (talk) 13:35, 4 September 2011 (UTC)[reply]

Why most of the time the voltage indicator remains on in Fridge?

I have a fridge running smoothly since 2004. Suddenly, I noticed that the voltage indicator (yellow light) remains on for several hours and then it comes back to the normal condition. During that time, ice and other groceries melts away. It usually happens at afternoon (1 pm to 4 pm) and at morning (4 am to 6 am). Thanks--180.234.97.221 (talk) 19:15, 4 September 2011 (UTC)[reply]

I think it will be a temperature indicator, not voltage, unless you live in an area with regular "brown-outs". If other electrical appliances are functioning normally, then the most likely fault is in the motor or the coolant of the refrigerator. Are these times when some heating switches on? If so, then the fridge is not able to work hard enough to compensate for the extra heat in the environment. Seven years is around the time when faults become more likely. You would probably get better answers on the Science help desk. Dbfirs 22:19, 4 September 2011 (UTC)[reply]

Is this a freezer/fridge with automatic defrost by any chance ? If so, that light may just tell you when it's in defrost mode. I hate automatic defrost, myself, as it destroys the food. I'd much prefer to defrost it myself once a year, or, better yet, have a design that never frosts up in the first place (I have an idea for how to do this on my invention list).

Some freezers have options to turn off the defrost mode or at least make it less aggressive. I had one that didn't, so removed the timer that triggers the defrost cycle. Unfortunately, the design for that refrigerator/freezer blows the cool air from the freezer down to the fridge via a narrow channel which quickly frosts over, making for an extra cold freezer and warm fridge, unless I defrost manually often. StuRat (talk) 16:06, 5 September 2011 (UTC)[reply]

I have had a variety of Whirlpool an Maytag devices in my 10 or so years of really paying attention to what was keeping my food cold, and I have never seen a unit defrost itself before. Maybe that's one of those highfalutin features? In any event, i would strongly suggest simply googling for the model number plus the brief description of your problem. You would be surprised how many appliance owners use the internet for problem-solving. In the few times I have had issues, this has resulted in a wealth of problem-specific knowledge. --144.191.148.3 (talk) 16:53, 6 September 2011 (UTC)[reply]

Some rather low end refrigerator/freezer combos tend to have this feature, although not so much on mini fridges. Note that there isn't normally much to tell you it has automatic defrost, except that frost doesn't build up and food doesn't last as long. StuRat (talk) 22:07, 6 September 2011 (UTC)[reply]

Chess app in IPad 2 ad "Learn"

Which chess app has been shown in Ipad 2 ad "Learn" http://www.apple.com/ipad/ — Preceding unsigned comment added by 182.178.174.157 (talk) 20:01, 4 September 2011 (UTC)[reply]

http://www.tchessgame.com/ ¦ Reisio (talk) 07:41, 5 September 2011 (UTC)[reply]