My electricity is being switched off next Thursday for 8 hours. I can't
think of any reason why my powerline adapters shouldn't continue to
work, but will they? I've got some battery-powered Raspberry Pis in a
part of the house where the WiFi doesn't work.
When a local village had it's planned electricity shut down for the day
I installed the wiring for a 3 phase genny for the local pub.
As I had swapped the power from mains to the genny before 8am someone
asked how will we know when the power is turned off.
My answer was just wait for the all the alarm bell boxes to go off that
have dead batteries in the control panel.
Also how will the wifi work and indeed the router if you are connected to
the internet that way.
You could set up a mobile as a hot spot and do it over mobile data, but
then you would need to reset all your passwords and oog in stuff as well.
If the weather is good go to the coast instead!
This newsgroup posting comes to you directly from...
The devices need mains to power them so they will act as comms devices - ie
to convert between Ethernet and the modulation that is applied to the mains
voltage on the house wiring, and the convert back to Ethernet in the other
I wonder if they would extract power from the Ethernet if both powerlink
devices were plugged into computers that provided PoE (power over Ethernet).
I think you may be out of luck.
I presume you've got a way of powering your router by battery, to access the
internet, and to hand out IP addresses to new devices (via DHCP).
Thinking *really* laterally - and I'm not for a minute suggesting that you
try this - in theory you could connect one powerlink device and your router
to a 12V-to-mains converter (powered from a car battery) and then run an
extension cable to another place where a powerlink device is plugged in. But
I'd bet the 12V-to-main converter will output such a horrible approximation
of a sine wave that the noise it generates would swamp the powerlink signal.
DON'T TRY IT - it's only a thought experiment.
I did once earn a few brownie points when the small business where I was
working suffered a power cut (JCB through high voltage cable somewhere in
the area) and the company couldn't even make/receive phone calls because
their phone system was entirely by VOIP. They couldn't even look up their
list of customers and their work diary to see which customers they were
booked for us (PC repair engineers) to visit, or the customers' phone
numbers to explain why we may be late. Luckily I had a 12V-to-mains
converter in my car (I bought it so I could charge my laptop while I was
driving etc), so they were able to rig it up to the company van (engine
ticking over so battery didn't go flat) and power a VOIP-to-POTS interface
so all incoming phone calls came through to a backup hard-wired phone, and
they managed to boot up their diary server and a router, and access it from
a laptop. It kept them going until the mains came back. I didn't see how
well they coped because I had to go out to see one of the customers whose
details we could remember because we had a paper copy of the job sheet, and
by the time I got back, normal service had been resumed. The buggers still
made me redundant a few months later :-(
If you think about it, they already need an enormous notch filter at 50Hz,
and they need to compensate for all kinds of electrical noise that's on the
mains from whatever electrical gear happens to be connected (SMPSUs
especially). They're already adaptive enough to scan for good frequencies
they can use to transmit, and I suspect the harmonics from the sine wave
converter are relatively predictable. The bit rate may be well down, but I
would expect them to still work.
If one is determined to continue to produce wideband RF interference
over several miles one could do this. But the simpler solution would
run a bit of Cat5e instead of the mains extension cable and get a much
better data connection without the inverter and powerline adaptors.
I agree that Cat 5 is always the best solution: the one that will "just
work" without any intermittent loss of connection, failure to reconnect
after power cut (*), sudden drop in speed when a microwave is turned on, or
bizarre interactions between wifi and bluetooth (**).
But it almost always involves trying to route the cable between one room and
another, buried under the edge of the carpet, fed under the metal carpet
strips in doorways, or else drilling through walls or ceilings to feed the
cable and plug through. I did initially think of laying Cat 5 in the loft,
to feed a wifi access point for the part of the house where wireless devices
would be used, but it would have meant drilling through the ceiling into the
loft, crawling along in very confined spaces near the eaves, and finding a
way of hiding a cable going vertically from floor to ceiling. I quickly
dismissed powerline because the house has two separate "fuse boxes"
(although on the same meter) and the signal strength gets very much worse
when you cross from one ring main to the other; it was pretty dire even a
few sockets away on the same ring main. Simple wifi (even 2.4 GHz) from the
router was woefully inadequate, so we had to invest in several mesh
devices - which work beautifully most of the time, until the problem when
the power goes off and the devices don't reconnect once it comes back.
(*) We have a mesh network to get broadband from one part of the house to
cover a "wing" at right angles. Getting the devices to connect after a power
cut (as happened in the middle of last night) is a problem, because if the
power to all the nodes is restored simultaneously, the child nodes don't
connect to the parent; instead the children need to be turned off and then
back on in sequence after the central parent node has started.
(**) On my old phone, I couldn't listed to streamed radio programmes (eg via
BBC Sounds and its previous iPlayer equivalent) via bluetooth headphones
because the phone didn't like bluetooth and 2.4 GHz wifi transferring data
at the same time. The solution was to use wired earphones when streaming
over wifi, or else transfer the file so it was held locally on the phone and
then listen using bluetooth headphones.
If you don't have cat 5, but have decent aerial cable (ie CT100, not the
horrible brown stuff) you can use MoCA to run ethernet over coax. I have
some bonded MoCA adapters that got approaching gigabit over the TV coax.
You simply put a MoCA adapter between the TV access point on the wall and
the TV and tap off ethernet.
I never got around to measuring them in a controlled test, but internet
connection was 200Mbps and it handled that without breaking sweat, over a
fairly sprawling house.
In theory you're supposed to have a MoCA-capable splitter (one that handles
satellite frequencies is fine) and a blocking filter to prevent the signal
leaking upstream, but all the points came off an existing aerial booster
and I just ran the MoCA across ports of the booster without touching that
setup at all.
I used it to run a WAP and office at the other end of the house from the
FTTH incomer, and it worked very well.
(I have the parts for sale if anyone wants them)
What was the reason that thin and thick coax for Ethernet had an impedance
of 50 ohms, whereas coax for TV is 75 ohm?
I remember the joys of T pieces and terminators for thin Ethernet, and the
problems that could be caused to everyone else on the LAN if an extra TV
piece had to be inserted for an additional PC, breaking the continuity of
the LAN by un-terminating it temporarily, or even if a T piece was
disconnected from a PC's network card even though the continuity of the LAN
was not disturbed (*). Structured cabling, with a separate cable back to a
central hub or switch, uses a lot more cable, but it's a lot more resilient
to connecting/disconnecting devices to/from the LAN.
(*) TCP seemed to be a lot more tolerant of a T-piece on an intact LAN being
disconnected from a network card, than OSI (ICL's OSLAN) was - the latter
almost always threw its toys out of the pram, and maybe even caused a UNIX
server to core-dump.
75 ohm is lower loss and generally matches antenna impedances. 50 ohms
can carry more power before dielectric breakdown and is used for higher
Though is a rather simplistic view.
Mists of time, I'm afraid.
Last coax ethernet I worked on had some really weird things happening.
There were certain pairs of devices that simply could not see each other
- despite seeing every other device on the network. Swapping the order of
one in the ring fixed it ... but introduced weirdness elsewhere.
Well, I think I see the problem. And I'm astonished it worked at all -
Ethernet should not be configured as a ring...
Elsewhere in the thread - Manchester encoding is used because it has no
net DC component.
And a story. I once lost complete faith in my engineering manager when
he plugged a bit of coax into a T-piece, took the entire office network
down, and didn't even connect him doing this with the way the entire
office erupted in expletives.
Perhaps that's why he made me redundant a year or so later!
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.