On 14/01/2022 10:12, Tweed wrote:

> There’s so many poorly designed switched mode power supplies

Most are well designed, but the components that implement the
interference are removed during the production engineering, after the
original design has been certification tested.

On 14/01/2022 10:16, Mark Carver wrote:

> Well, overhead phone cables were never intended to carry RF either, but
> they do thanks to ADSL etc.

Overhead phone cables are designed to suppress audio frequency
interference, which means they are well balanced, including using
twisted pairs.

On 14/01/2022 10:37, David Woolley wrote:
> On 14/01/2022 10:16, Mark Carver wrote:
>
>> Well, overhead phone cables were never intended to carry RF either,
>> but they do thanks to ADSL etc.
>
> Overhead phone cables are designed to suppress audio frequency
> interference, which means they are well balanced, including using
> twisted pairs.

Except a lot are not twisted. There's an awful lot of 'twin flat' style
still in use. That said, it should still offer some protection in a
balanced situation

NY <me@privacy.invalid> wrote:
> "Woody" <harrogate3@ntlworld.com> wrote in message
> news:srrao7$89i$1@dont-email.me...
>
> [snip: my account of spectacularly poor data transfer dates with Powerline]
>
> > I could only comment that you must have been using the wrong type of
> > powerline unit. I have a couple of 600Mb TP-Link units. One is connected
> > to my source router adjacent to the VM superhub configured as a modem -
> > this on the downstairs ring; I have tried a 'receiver' on the upstairs
> > ring at the other end of the house and from an incoming source typically
> > about 110Mb I was getting well in excess of 50Mb upstairs. I then tried
> > connection to my shed: this is from a different power box in the
> > (built-in) garage so the meeting point of the rings is actually at the
> > meter termination, whence it runs by about 30m of 1.5mm armoured cable to
> > the shed and onto a spur feed there - I recorded 37Mb at that location.
> >
> > I too am a radio amateur and have found no RFI on my receivers whatsoever.
>
> I had been using a pair of WD Livewire (*) devices at our old house (2 ring
> mains, one CU - nice and simple) to get Ethernet from the router to a TV or
> PVR which only had Ethernet and no wifi. It gave perfectly good results,
> though I forget what speed.
>
> I tried those in our new house, and also a pair of Dlink DHP-500AV devices
> for comparison, and in both cases the results were dramatically worse. In
> both cases I was using like with like, rather than Dlink talking to WD.
>
> I can only think that there is something about the wiring in our house which
> spoils the Powerline signal. I can understand that going from one ring main
> to another *might* cause extra denaturing of the signal, and going from one
> CU to the other will probably cause more of it, but I don't see why two
> sockets about 20 feet apart on the same ring main should be *so* bad.
>
> It's the first time I've had such severe problems with Powerline.
>
I've tried a few over the years in our house and they have always been
next to useless, i.e. even when close enough together to get a
reasonable WiFi connection over the same distance they only manage,
maybe, 5Mbs. On any link where they might be useful (we do have some
quite solid walls and it's a large house) they dribble along,
unreliably, at 0.5 to 1Mbs.

I'm not sure what causes this, I know in my daughter's modern holiday
let a pair of powerline devices works perfectly across the top floor
(though it's moderately pointless as it's open plan).

It *may* be because we have an 'all RCBO' consumer unit and it is
quite a big house so cable runs are quite long. We're also a TT
installation, though I really can't see how this would affect it.

(If I was being cynical I'd suggest that what they actually do is use
WiFi and just pretend to make the connection via power line. In terms
of how well they work that's *exactly* what I see!)

--
Chris Green
·

Mark Carver <mark.carver@invalid.invalid> wrote:
>
> Back to power adaptors. Never used them myself, but in sorting out
> others I've discovered they don't work well when plugged into 'flying
> lead' extension blocks.
> Try to use them plugged into the actual wall socket.
>
The instructions with them do always say that they should be plugged
in direct to a wall socket. Putting my electronics/electrical
engineering hat on I can't think of any good reason why this should be
so though.

Thinking aloud....

Extension leads use flex rather than solid cable, modern mains
cable (i.e. the house wiring behind the wall socket) is just about
universally 2.5sqmm single solid conductor. Maybe multistrand
flex radiates more than single solid, though in principle stranded
should be better for RF as higher frequencies travel on the
surface of the wire (hence the use of Litz wire for coils etc.)

House wiring cable keeps the conductors at a pretty much fixed
distance apart which will present a fairly consistent impedance.
Do powerline devices depend on this (approximately) fixed
impedance? Extension leads will have a different (and less
consistent) inter-conductor spacing.

It might be interesting to try this out. Simply get a pair of
powerline devices and connect them via a pair of sockets on the ends
of lengths of different types of wire. Do they only work well when
all of the cable from one to the other is 2.5sqmm T&E?

--
Chris Green
·

In article <srrbvm$eub$1@dont-email.me>,
Tweed <usenet.tweed@gmail.com> wrote:

> Just as a point of information, my mesh network sustains my Virgin
> Media 200Mbit/sec data rate all over my house.

May I ask what that mesh network is?

Thanks.

Bob.

"Brian Gaff (Sofa)" <briang1@blueyonder.co.uk> wrote in message
news:srri4i$ks1$1@dont-email.me...
> Another oddity of powerline networks is that at least around here, some
> manage to also work two houses down the street but not in the in between
> houses. I would imagine this means that the signals are not getting to the
> phases of the mains on the intervening properties. When there used to be
> analogue wireless intercoms plugged into the mains, the same effect could
> be seen, so watch it with older baby monitors.

Yes I think successive houses are often wired on successive phases, so
houses 1, 7, 13 will all be on one phase, and 3, 9, 15 will be on another,
and 5, 11, 17 will be on the third. Assuming houses are numbered odds on one
side, evens on the opposite side.

I remember our server-room at work was (for some reason) wired with
different phases on each leg of the capital E layout of benches. There were
big signs on each bench saying which phase it used, and even bigger signs
warning people not to connect equipment (eg by RS-232) on two benches -
which was a real nuisance if you brought in a piece of equipment to be used
temporarily and the only spare mains sockets were on the wrong bench.

I suppose they were terrified of two pieces of equipment of different
benches both developing faults that made a data pin live, and those pins
then having 415 V (*) between them rather than both being at the same
potential.

(*) Is it 415 V between phases when there is 240 V phase-to-neutral?

"Chris Green" <cl@isbd.net> wrote in message
news:do18bi-7k9l.ln1@esprimo.zbmc.eu...
> The instructions with them do always say that they should be plugged
> in direct to a wall socket. Putting my electronics/electrical
> engineering hat on I can't think of any good reason why this should be
> so though.
>
> Thinking aloud....
>
> Extension leads use flex rather than solid cable, modern mains
> cable (i.e. the house wiring behind the wall socket) is just about
> universally 2.5sqmm single solid conductor. Maybe multistrand
> flex radiates more than single solid, though in principle stranded
> should be better for RF as higher frequencies travel on the
> surface of the wire (hence the use of Litz wire for coils etc.)
>
> House wiring cable keeps the conductors at a pretty much fixed
> distance apart which will present a fairly consistent impedance.
> Do powerline devices depend on this (approximately) fixed
> impedance? Extension leads will have a different (and less
> consistent) inter-conductor spacing.
>
>
> It might be interesting to try this out. Simply get a pair of
> powerline devices and connect them via a pair of sockets on the ends
> of lengths of different types of wire. Do they only work well when
> all of the cable from one to the other is 2.5sqmm T&E?

Did that once. I compared:

- two sockets on the same extension block
- one socket on extension, other on adjacent wall socket, both connected to
same dual-socket A
- one on extension, one on nearby socket B on same ring main
- one on wall socket A and one on wall socket B

This was for a four-way extension with about 6 feet of cable.

I didn't notice a significant difference with a simple extension, but there
was a noticeable reduction for all but the first case if the extension had
mains-spike suppressors.

"Chris Green" <cl@isbd.net> wrote in message
news:8218bi-7k9l.ln1@esprimo.zbmc.eu...

> It *may* be because we have an 'all RCBO' consumer unit and it is
> quite a big house so cable runs are quite long. We're also a TT
> installation, though I really can't see how this would affect it.

Ah, I wonder if RCDs and MCBs can shunt the powerline signal live-to-neutral
at the CU.

Our old house had one RCD, between the CU and the meter, and used fuses on
all circuits. This house has one RCD per ring main (ie they are within the
CU) and uses MCBs on all circuits.

On 14/01/2022 11:16, Bob Latham wrote:
> In article <srrbvm$eub$1@dont-email.me>,
> Tweed <usenet.tweed@gmail.com> wrote:
>
>> Just as a point of information, my mesh network sustains my Virgin
>> Media 200Mbit/sec data rate all over my house.
>
> May I ask what that mesh network is?
>
>
> Thanks.
>
> Bob.
>
Bumped in the hope of an answer.

I am happy to swop my powerline for mesh but need to know which
particular brand/model is the best.

Bob Latham <bob@sick-of-spam.invalid> wrote:
> In article <srrbvm$eub$1@dont-email.me>,
> Tweed <usenet.tweed@gmail.com> wrote:
>
> > Just as a point of information, my mesh network sustains my Virgin
> > Media 200Mbit/sec data rate all over my house.
>
> May I ask what that mesh network is?
>
My ad-hoc WiFi APs (two in the house, one 50 yards away in a cabin) do
better than that. I just checked, I'm using my laptop in the lounge
at the moment, I have a 300Mbs connection.

Trying 5GHz gives me slightly less, 180Mbs. I've never found 5GHz to
offer any advantage anywhere in our house except when the client is
virtually on top of the AP. We are quite lucky in being fairly well
away from any other WiFi signals, everything I can see is 'ours'
except one fairly faint BT router signal from a house across the road.

--
Chris Green
·

On 13/01/2022 18:54, Tweed wrote:

> Given the OP says wiring will be a pain I assume Ethernet cable is not an
> option. I’d second the recommendation for a mesh network. You need a 3
> radio per unit version. 2.5 and 5GHz plus an additional hidden 5GHz that
> links the units together. I have a Linksys Velop system and it works well.
> If the house has a wooden upstairs floor place the units upstairs and let
> them shine down into the ground floor. Wood is pretty transparent to RF.
>

How many Linksys Velop units would you expect to need to cover a 2,000
sq ft Victorian apartement?

On 14/01/2022 11:44, Tony Gamble wrote:
> On 14/01/2022 11:16, Bob Latham wrote:
>> In article <srrbvm$eub$1@dont-email.me>,
>>     Tweed <usenet.tweed@gmail.com> wrote:
>>
>>> Just as a point of information, my mesh network sustains my Virgin
>>> Media 200Mbit/sec data rate all over my house.
>>
>> May I ask what that mesh network is?
>>
>>
>> Thanks.
>>
>> Bob.
>>
> Bumped in the hope of an answer.
>
> I am happy to swop my powerline for mesh but need to know which
> particular brand/model is the best.
>

Well I've heard of Velop, Orbi and Amplifi and Meraki (part of Cisco).

I don't use any mesh networking but I do use Ubiquiti's Access points in
my home.... so no complaints about Ubiquiti

https://www.linksys.com/gb/velop/
https://www.netgear.com/home/wifi/mesh/rbk50/
https://www.amplifi.com/
https://meraki.cisco.com/solutions/next-gen-wifi/ and
https://meraki.cisco.com/technologies/mesh-routing

https://www.techradar.com/uk/news/best-wireless-mesh-routers

https://www.trustedreviews.com/reviews/ubiquiti-amplifi-mesh-wi-fi-system

On 13/01/2022 in message <xn0ncthcs19svj5001@news.individual.net> Jeff
Gaines wrote:

>Any thoughts/comment on powerline networks? There was a suggestion they
>can cause RFI and I thought the combined knowledge in here would be able
>to comment on that.

Just to say thank you to everybody for all the input/discussion :-)

I have some kit that will only work if wired but a solution might be a
sort of reverse "WiFi" that picks up a signal wirelessly and outputs it
though an RJ45 (is that right?) socket - is there such a thing?

--
Jeff Gaines Wiltshire UK
The facts, although interesting, are irrelevant

On 14/01/2022 12:03, Jeff Gaines wrote:
> On 13/01/2022 in message <xn0ncthcs19svj5001@news.individual.net> Jeff
> Gaines wrote:
>
>> Any thoughts/comment on powerline networks? There was a suggestion
>> they can cause RFI and I thought the combined knowledge in here would
>> be able to comment on that.
>
> Just to say thank you to everybody for all the input/discussion :-)
>
> I have some kit that will only work if wired but a solution might be a
> sort of reverse "WiFi" that picks up a signal wirelessly and outputs it
> though an RJ45 (is that right?) socket - is there such a thing?
>

yes there is, I believe its called a Wi Fi bridge.

https://arstechnica.com/gadgets/2021/08/point-to-point-wi-fi-bridging-between-buildings-the-cheap-and-easy-way/

just need to check that if you buy a Wi Fi AP that it can be put into
Bridge mode then you could plug a network switch into that bridge and
then hang several ethernet devices on subject to wi fi bandwidth
limitations.

On Fri, 14 Jan 2022 10:16:00 -0000 (UTC), Tweed
<usenet.tweed@gmail.com> wrote:

>> When I can find an AM radio and power it up, I'll plug the devices in and
>> see whether I can detect any interference in the LW or MW band that is
>> attributable to Powerline. I can't test the effect on amateur radio bands.
>>
>>
>
>The problems I’ve found with power line adapters is interference from other
>devices plugged in. A simple test will show a respectable throughput, but
>long term you get periodic dropouts, when fridge motors, washing machines
>etc start up. The link does recover, but it’s not ideal for the likes of
>video calls.

Same here. I eventually went to the trouble of installing an ethernet
cable run between my router in the hallway and my main computer in the
living room. It was a lot of hassle but worth it in the end. Powerline
would occasionally lose its connection, not often but enough to be
annoying, but the cable connection is rock solid and speed tests give
results close to sync speed every time.

I never saw any evidence suggesting that my powerline devices might be
causing any interference to anybody else though, depite the bad
reputation that some people seem determined to give them. Using a
small portable radio, I've tried setting it to blank space between
programmes at various frequencies including shortwave, and moving it
close to the powerline devices and mains cables around the house, but
never heard anything more than a few inches away from the devices
themselves. Other devices like digital clocks and the TV set appeared
to radiate more RF rubbish over much greater distrances, but nobody
complains about them. If my radio can receive broadcasts and ham
transmissions from round the world but can't pick up anything a foot
away from an electronic gadget, it's a bit difficult to believe that
said gadget is radiating enough rubbish to be a real problem.

Rod.

On Fri, 14 Jan 2022 11:12:39 -0000, "NY" <me@privacy.invalid> wrote:

>(*) Is it 415 V between phases when there is 240 V phase-to-neutral?

Multiply by the square root of three, so yes, more or less.

Rod.

NY <me@privacy.invalid> wrote:
> "Chris Green" <cl@isbd.net> wrote in message
> news:do18bi-7k9l.ln1@esprimo.zbmc.eu...
> > The instructions with them do always say that they should be plugged
> > in direct to a wall socket. Putting my electronics/electrical
> > engineering hat on I can't think of any good reason why this should be
> > so though.
> >
> > Thinking aloud....
> >
> > Extension leads use flex rather than solid cable, modern mains
> > cable (i.e. the house wiring behind the wall socket) is just about
> > universally 2.5sqmm single solid conductor. Maybe multistrand
> > flex radiates more than single solid, though in principle stranded
> > should be better for RF as higher frequencies travel on the
> > surface of the wire (hence the use of Litz wire for coils etc.)
> >
> > House wiring cable keeps the conductors at a pretty much fixed
> > distance apart which will present a fairly consistent impedance.
> > Do powerline devices depend on this (approximately) fixed
> > impedance? Extension leads will have a different (and less
> > consistent) inter-conductor spacing.
> >
> >
> > It might be interesting to try this out. Simply get a pair of
> > powerline devices and connect them via a pair of sockets on the ends
> > of lengths of different types of wire. Do they only work well when
> > all of the cable from one to the other is 2.5sqmm T&E?
>
> Did that once. I compared:
>
> - two sockets on the same extension block
> - one socket on extension, other on adjacent wall socket, both connected to
> same dual-socket A
> - one on extension, one on nearby socket B on same ring main
> - one on wall socket A and one on wall socket B
>
> This was for a four-way extension with about 6 feet of cable.
>
But in none of those cases does the powerline network offer any
advantage does it, they'd all work virtually perfectly using WiFi.

> I didn't notice a significant difference with a simple extension, but there
> was a noticeable reduction for all but the first case if the extension had
> mains-spike suppressors.
>

To be any use to me powerline would have to work between sockets that
are ten metres or more apart, that's the sort of length of Cat5 cable
I'm using to feed my APs.

I tried a pair of powerline adapters between garage and cabin
(probably a bit less than 10 metres, but a straight mains connection
on the same CU) and it was quite useless. WiFi was a lot better but
I've buried a Cat5 cable now.

I really can't find a case where powerline works for me.

--
Chris Green
·

On Fri, 14 Jan 2022 10:35:19 +0000, David Woolley
<david@ex.djwhome.demon.invalid> wrote:

>On 14/01/2022 10:12, Tweed wrote:
>
>> There’s so many poorly designed switched mode power supplies
>
>Most are well designed, but the components that implement the
>interference are removed during the production engineering, after the
>original design has been certification tested.

I would expect the ones in powerline type devices are better designed
than most, because they'd have to be, otherwise they'd interfere with
their own working. My own (admittedly limited) experiments with a
portable shortwave radio suggests that anything with a switch mode
power supply, or a digital display, or a screen, or any combination of
the above, is likely to be a worse offender than any powerline device.

Rod.

On 14/01/2022 12:35, Roderick Stewart wrote:
> On Fri, 14 Jan 2022 10:35:19 +0000, David Woolley
> <david@ex.djwhome.demon.invalid> wrote:
>
>> On 14/01/2022 10:12, Tweed wrote:
>>
>>> There’s so many poorly designed switched mode power supplies
>>
>> Most are well designed, but the components that implement the
>> interference are removed during the production engineering, after the
>> original design has been certification tested.
>
> I would expect the ones in powerline type devices are better designed
> than most, because they'd have to be, otherwise they'd interfere with
> their own working. My own (admittedly limited) experiments with a
> portable shortwave radio suggests that anything with a switch mode
> power supply, or a digital display, or a screen, or any combination of
> the above, is likely to be a worse offender than any powerline device.
>
> Rod.

Therein lies one of the big issues with power line networking.....

all those SMPS wall warts all over the house, flourescent lighting,
kitchen appliance motors, central heating pumps, internal SMPS in
domestic appliances like TV's, mains clocks, computers etc and laptop
PSU's all throwing out RF which then gets into all that twin and earth
cables all acting as recieve antennas!

On 14/01/2022 12:03, Jeff Gaines wrote:
> On 13/01/2022 in message <xn0ncthcs19svj5001@news.individual.net> Jeff
> Gaines wrote:
>
>> Any thoughts/comment on powerline networks? There was a suggestion they
>> can cause RFI and I thought the combined knowledge in here would be able
>> to comment on that.
>
> Just to say thank you to everybody for all the input/discussion :-)

But was your OP answered? As far as I can see, only Woody and I had any
practical experience of powerline use and RFI and neither of us had a
problem with it, whether MW or SW. There were many more posts which
stated there *might* be a problem, but were any of those other than what
they had read somewhere else, or was that what they had experienced
themselves?

FWIW, several years ago I had a variable-speed remote-controlled DC fan
installed in a conservatory, and it wiped out MW reception in it. That
was fixed by putting a delta filter on the PS leads, and putting a
couple of turns of those leads through a large ferrite ring. So I do
have experience of interference which can affect MW!

--

Jeff

In article <srq3gu$qr0$1@dont-email.me>, NY <me@privacy.invalid> wrote:
> "Bob Latham" <bob@sick-of-spam.invalid> wrote in message
> news:59aa4f40dcbob@sick-of-spam.invalid...
> > In article <59aa15e7eenoise@audiomisc.co.uk>, Jim Lesurf
> > <noise@audiomisc.co.uk> wrote:

> >> Yes, they will radiate RF. What we can't tell you is if you *will*
> >> notice it having an impact on something else. Nor if any near
> >> neighbours will encounter interference from you using it.
> >
> >> Bascially, it should never have been legalised.
> >
> > +1

> Or if it was, there should have been tighter regulation of technical
> standards so it avoids any fundamentals or harmonics that affect
> broadcast comms or any other application which can't tolerate powerline
> interference I believe some devices do claim to have notch filters to
> avoid MW radio 530-1300 kHz approx.

The basic flaw is that many of the devices connected to mains present a
non-linear load that varies with time. Hence even when the 'power line
network' box makers 'notch' their output, some of this gets converted
*into* components in the spectral 'notches'. Depends entirely on the
details of a given home setup. Given the time-variations in the presented
loads and the multiple-component output, all kinds of other intermodulation
components then get created. Basically, a wall of crap.

The fact that they were 'legalised' simply shows the reality that OffCrap
have zero interest or understanding of even RF 101, and are just there to
rubber stamp. Their main aim is to maximise Gov income from flogging off
spectrum, etc. Consequences b'damned!

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
biog http://jcgl.orpheusweb.co.uk/history/ups_and_downs.html
Audio Misc http://www.audiomisc.co.uk/index.html

In article <srrao7$89i$1@dont-email.me>, Woody <harrogate3@ntlworld.com>
wrote:

> I could only comment that you must have been using the wrong type of
> powerline unit. I have a couple of 600Mb TP-Link units. One is connected
> to my source router adjacent to the VM superhub configured as a modem -
> this on the downstairs ring; I have tried a 'receiver' on the upstairs
> ring at the other end of the house and from an incoming source typically
> about 110Mb I was getting well in excess of 50Mb upstairs. I then tried
> connection to my shed: this is from a different power box in the
> (built-in) garage so the meeting point of the rings is actually at the
> meter termination, whence it runs by about 30m of 1.5mm armoured cable
> to the shed and onto a spur feed there - I recorded 37Mb at that
> location.

> I too am a radio amateur and have found no RFI on my receivers
> whatsoever.

Which indicates that you were lucky and didn't have a home mains + consumer
devices, etc, arrangement that caused a significant problem for you. Others
- inc the neighbours of users - may be less lucky.

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
biog http://jcgl.orpheusweb.co.uk/history/ups_and_downs.html
Audio Misc http://www.audiomisc.co.uk/index.html

Jeff Gaines <jgaines_newsid@yahoo.co.uk> wrote:
> On 13/01/2022 in message <xn0ncthcs19svj5001@news.individual.net> Jeff
> Gaines wrote:
>
> >Any thoughts/comment on powerline networks? There was a suggestion they
> >can cause RFI and I thought the combined knowledge in here would be able
> >to comment on that.
>
> Just to say thank you to everybody for all the input/discussion :-)
>
> I have some kit that will only work if wired but a solution might be a
> sort of reverse "WiFi" that picks up a signal wirelessly and outputs it
> though an RJ45 (is that right?) socket - is there such a thing?
>
Yes, TP-Link make some reasonably priced outdoor APs which can be used
in pairs to talk to each other to provide a 'bridge'. Models such as
the TL-WA5210G and TL-WA7210N can do it. However these are rather old
models, I'm sure there are newer. I expect there are indoor devices
to to the same as well.

--
Chris Green
·

"Tony Gamble" <tonygamble@compuserve.com> wrote in message
news:j4d698Fq9v1U1@mid.individual.net...
> On 13/01/2022 18:54, Tweed wrote:
>
>> Given the OP says wiring will be a pain I assume Ethernet cable is not an
>> option. I’d second the recommendation for a mesh network. You need a 3
>> radio per unit version. 2.5 and 5GHz plus an additional hidden 5GHz that
>> links the units together. I have a Linksys Velop system and it works
>> well.
>> If the house has a wooden upstairs floor place the units upstairs and let
>> them shine down into the ground floor. Wood is pretty transparent to RF.
>>
>
> How many Linksys Velop units would you expect to need to cover a 2,000 sq
> ft Victorian apartement?

When you say "apartment", does that imply that it's all on one level?

If it's Victorian, its internal walls may be masonry rather than wooden
studs with plasterboard over. That may mean that the walls attenuate the
signal more. On the other hand, modern plasterboard can have a metal foil on
one side which attenuates even more :-(

Our house is L-shaped, with three consecutive rooms along one leg (that part
of the house dates back to the mid 1800s and so has thick external walls and
thinner masonry internal walls. The other leg has one room of Victorian
construction and then two large rooms consecutively beyond that which are
1990s construction (brick external, plasterboard internal).

The router and hence the master Linksys Velop node which is connected to it
by Ethernet, is half way along the first-mentioned leg. I've positioned the
primary node where it has good line of sight of the inside of the L on both
legs.

I found that we needed 6 nodes (primary and 5 secondaries) to give good 5
GHz backhaul connection between nodes. Unfortunately the shorter range of 5
GHz means that the nodes have to be a lot closer together than the range of
2.4 GHz (and we need 2.4 turned on) so the nodes are as far apart as 5 will
allow, but there is a *lot* of overlap of 2.4. The auto-negotiation of 2.4
channel means that some nodes often fail to connect to each other if they
are all turned on simultaneously after a power cut; by bitter experience
I've found that if this happens, I need to turn all the secondaries off and
then turn each on in sequence, waiting for it to connect before turning the
next on. This is a real PITA :-(

https://i.postimg.cc/BbqNwKLt/Scan-14-01-2022-1301-small.png

I've distinguished between a single line (brick external or plasterboard
internal) and stone (double-thickness, hatched). Windows are shown as
rectangles inserted into walls - all windows are modern uPVC and
double-glazed (*). Most of the house is ground floor only, with another
storey above just the modern brick part of the house.

The nodes and the router are the pink circles, with dotted lines showing how
they tend to connect to the primary node.

(*) They also have plastic inserts between the panes to simulate
old-fashioned small-paned windows - not our choice, done by previous owners!