info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Mesh systems
23 views
Skip to first unread message
Tim+
Nov 30, 2021, 3:00:47 PM11/30/21
to
We have a BT mesh system to provide seamless Wi-Fi throughout our house and
it works very well indeed. My daughter is having a lot of Wi-Fi issues
with the walls in her house so I was wondering about getting her one.
When we bought it, there weren’t a lot of options on the market at the time
but a quick squiz at Amazon shows lots of new kids on the block from £49
for a three node system. The BT system it much pricier at £199.
Anyone know if these cheaper ones are any good or have personal experience
of non-BT mesh systems? Must be fairly idiot proof to set up and decently
reliable. ;-)
Tim
--
Please don't feed the trolls
it works very well indeed. My daughter is having a lot of Wi-Fi issues
with the walls in her house so I was wondering about getting her one.
When we bought it, there weren’t a lot of options on the market at the time
but a quick squiz at Amazon shows lots of new kids on the block from £49
for a three node system. The BT system it much pricier at £199.
Anyone know if these cheaper ones are any good or have personal experience
of non-BT mesh systems? Must be fairly idiot proof to set up and decently
reliable. ;-)
Tim
--
Please don't feed the trolls
Mark
Nov 30, 2021, 4:36:34 PM11/30/21
to
backhaul bandwidth within the mesh isn't shared with 'user access'
bandwidth, and the available bandwidth in any case should be greater.
However if the walls are very absorbing, its possible three BT 'discs'
may not be enough. You might need to be prepared to return it.
If that turns out to be the case powerline might be a better way to go -
I have used five Netgear DD-WRT routers in repeater bridge mode in a
student house with concrete walls and coverage is barely good enough. In
a less severe environment the same setup was exceptionally good.
Graham J
Nov 30, 2021, 4:53:21 PM11/30/21
to
Mark wrote:
[snip]
>
> I believe the pricier systems have more radios and more antennas so that
> backhaul bandwidth within the mesh isn't shared with 'user access'
> bandwidth, and the available bandwidth in any case should be greater.
> However if the walls are very absorbing, its possible three BT 'discs'
> may not be enough. You might need to be prepared to return it.
>
> If that turns out to be the case powerline might be a better way to go -
> I have used five Netgear DD-WRT routers in repeater bridge mode in a
> student house with concrete walls and coverage is barely good enough. In
> a less severe environment the same setup was exceptionally good.
Ultimately the only fully reliable arrangement is to have all the Access
Points connected by Ethernet cable to their master controller and the
internet router. That way the backhaul is much more reliable and
potentially much faster - possibly up to 10 Gbit/sec to every AP. Then
site as many APs as are needed to guarantee coverage - certainly one in
each room - perhaps several. The controller then manages all the APs
and instructs which one is to communicate with each client, changing the
assignments as the client moves around.
This is really only practical in an industrial setting where flood wired
Ethernet would be the norm. But in a domestic environment, whenever you
redecorate it is always worth putting in Ethernet cables, so that in
time you have a sensible flood wiring arrangement.
--
Graham J
[snip]
>
> I believe the pricier systems have more radios and more antennas so that
> backhaul bandwidth within the mesh isn't shared with 'user access'
> bandwidth, and the available bandwidth in any case should be greater.
> However if the walls are very absorbing, its possible three BT 'discs'
> may not be enough. You might need to be prepared to return it.
>
> If that turns out to be the case powerline might be a better way to go -
> I have used five Netgear DD-WRT routers in repeater bridge mode in a
> student house with concrete walls and coverage is barely good enough. In
> a less severe environment the same setup was exceptionally good.
Points connected by Ethernet cable to their master controller and the
internet router. That way the backhaul is much more reliable and
potentially much faster - possibly up to 10 Gbit/sec to every AP. Then
site as many APs as are needed to guarantee coverage - certainly one in
each room - perhaps several. The controller then manages all the APs
and instructs which one is to communicate with each client, changing the
assignments as the client moves around.
This is really only practical in an industrial setting where flood wired
Ethernet would be the norm. But in a domestic environment, whenever you
redecorate it is always worth putting in Ethernet cables, so that in
time you have a sensible flood wiring arrangement.
--
Graham J
Martin Brown
Dec 1, 2021, 3:29:16 AM12/1/21
to
On 30/11/2021 21:52, Graham J wrote:
> Mark wrote:
>
> [snip]
>
>>
>> I believe the pricier systems have more radios and more antennas so
>> that backhaul bandwidth within the mesh isn't shared with 'user
>> access' bandwidth, and the available bandwidth in any case should be
>> greater. However if the walls are very absorbing, its possible three
>> BT 'discs' may not be enough. You might need to be prepared to return it.
>>
>> If that turns out to be the case powerline might be a better way to go
>> - I have used five Netgear DD-WRT routers in repeater bridge mode in a
>> student house with concrete walls and coverage is barely good enough.
>> In a less severe environment the same setup was exceptionally good.
Fixed point to fixed point powerline is OK provided you are not a radio ham.
> Mark wrote:
>
> [snip]
>
>>
>> I believe the pricier systems have more radios and more antennas so
>> that backhaul bandwidth within the mesh isn't shared with 'user
>> access' bandwidth, and the available bandwidth in any case should be
>> greater. However if the walls are very absorbing, its possible three
>> BT 'discs' may not be enough. You might need to be prepared to return it.
>>
>> If that turns out to be the case powerline might be a better way to go
>> - I have used five Netgear DD-WRT routers in repeater bridge mode in a
>> student house with concrete walls and coverage is barely good enough.
>> In a less severe environment the same setup was exceptionally good.
The other much cheaper option for a cut price solution is one of the
Wifi repeaters (potentially loses half your bandwidth in noddy mode). It
is effectively a plug and play repeater in its default mode.
I have one wired into the TV feed in pass through and rebroadcast as a
new SSID mode (hell to set up but reliable once you manage it). Problem
is that once configured you cannot get back to the settings page without
a factory reset to all defaults so the slightest error is fatal.
It requires power cycling every couple of months when it stops talking
to the network but for £20 it is cheap cheerful and very effective.
> Ultimately the only fully reliable arrangement is to have all the Access
> Points connected by Ethernet cable to their master controller and the
> internet router. That way the backhaul is much more reliable and
> potentially much faster - possibly up to 10 Gbit/sec to every AP. Then
> site as many APs as are needed to guarantee coverage - certainly one in
> each room - perhaps several. The controller then manages all the APs
> and instructs which one is to communicate with each client, changing the
> assignments as the client moves around.
use it. I discovered that short lengths <30m of nominal 100M wiring will
quite happily take 1G signals without any noticeable loss in speed.
> This is really only practical in an industrial setting where flood wired
> Ethernet would be the norm. But in a domestic environment, whenever you
> redecorate it is always worth putting in Ethernet cables, so that in
> time you have a sensible flood wiring arrangement.
the most difficult blackspot in the home and/or streaming TV. More than
that is overkill in a domestic setting. Most clients barely require any
significant bandwidth compared to what modern Wifi can support.
--
Regards,
Martin Brown
Ken
Dec 1, 2021, 4:44:56 AM12/1/21
to
On Wed, 1 Dec 2021 08:29:14 +0000, Martin Brown
<'''newspam'''@nonad.co.uk> wrote:
>On 30/11/2021 21:52, Graham J wrote:
>> Mark wrote:
>>
>> [snip]
>>
>>>
>>> I believe the pricier systems have more radios and more antennas so
>>> that backhaul bandwidth within the mesh isn't shared with 'user
>>> access' bandwidth, and the available bandwidth in any case should be
>>> greater. However if the walls are very absorbing, its possible three
>>> BT 'discs' may not be enough. You might need to be prepared to return it.
>>>
>>> If that turns out to be the case powerline might be a better way to go
>>> - I have used five Netgear DD-WRT routers in repeater bridge mode in a
>>> student house with concrete walls and coverage is barely good enough.
>>> In a less severe environment the same setup was exceptionally good.
>
>Fixed point to fixed point powerline is OK provided you are not a radio ham.
I don't exactly count as a radio ham. I had zero DAB reception and
<'''newspam'''@nonad.co.uk> wrote:
>On 30/11/2021 21:52, Graham J wrote:
>> Mark wrote:
>>
>> [snip]
>>
>>>
>>> I believe the pricier systems have more radios and more antennas so
>>> that backhaul bandwidth within the mesh isn't shared with 'user
>>> access' bandwidth, and the available bandwidth in any case should be
>>> greater. However if the walls are very absorbing, its possible three
>>> BT 'discs' may not be enough. You might need to be prepared to return it.
>>>
>>> If that turns out to be the case powerline might be a better way to go
>>> - I have used five Netgear DD-WRT routers in repeater bridge mode in a
>>> student house with concrete walls and coverage is barely good enough.
>>> In a less severe environment the same setup was exceptionally good.
>
>Fixed point to fixed point powerline is OK provided you are not a radio ham.
really poor FM in my house - until I removed my powerline stuff and
replaced it with a mesh. Now both FM and DAB are just fine.
By the time I discovered this I was already in the habit using getting
all of my radio from streams but I don't miss the powerline.
My setup is an Asus Lyra Mini with three APs. Since I bought it there
have been quite a few negative reviews but it works fine for me, the
RJ45 sockets on the APs supporting the things I have that don't
support wifi.
My latest usb dongle on my old Surface is connecting at 866Mbs.
>
Martin Brown
Dec 1, 2021, 5:05:24 AM12/1/21
to
On 01/12/2021 09:44, Ken wrote:
> On Wed, 1 Dec 2021 08:29:14 +0000, Martin Brown
> <'''newspam'''@nonad.co.uk> wrote:
>
>> On 30/11/2021 21:52, Graham J wrote:
>>> Mark wrote:
>>>
>>> [snip]
>>>
>>>>
>>>> I believe the pricier systems have more radios and more antennas so
>>>> that backhaul bandwidth within the mesh isn't shared with 'user
>>>> access' bandwidth, and the available bandwidth in any case should be
>>>> greater. However if the walls are very absorbing, its possible three
>>>> BT 'discs' may not be enough. You might need to be prepared to return it.
>>>>
>>>> If that turns out to be the case powerline might be a better way to go
>>>> - I have used five Netgear DD-WRT routers in repeater bridge mode in a
>>>> student house with concrete walls and coverage is barely good enough.
>>>> In a less severe environment the same setup was exceptionally good.
>>
>> Fixed point to fixed point powerline is OK provided you are not a radio ham.
>
> I don't exactly count as a radio ham. I had zero DAB reception and
> really poor FM in my house - until I removed my powerline stuff and
> replaced it with a mesh. Now both FM and DAB are just fine.
You must be in a very marginal signal area then. I have no problems at
> On Wed, 1 Dec 2021 08:29:14 +0000, Martin Brown
> <'''newspam'''@nonad.co.uk> wrote:
>
>> On 30/11/2021 21:52, Graham J wrote:
>>> Mark wrote:
>>>
>>> [snip]
>>>
>>>>
>>>> I believe the pricier systems have more radios and more antennas so
>>>> that backhaul bandwidth within the mesh isn't shared with 'user
>>>> access' bandwidth, and the available bandwidth in any case should be
>>>> greater. However if the walls are very absorbing, its possible three
>>>> BT 'discs' may not be enough. You might need to be prepared to return it.
>>>>
>>>> If that turns out to be the case powerline might be a better way to go
>>>> - I have used five Netgear DD-WRT routers in repeater bridge mode in a
>>>> student house with concrete walls and coverage is barely good enough.
>>>> In a less severe environment the same setup was exceptionally good.
>>
>> Fixed point to fixed point powerline is OK provided you are not a radio ham.
>
> I don't exactly count as a radio ham. I had zero DAB reception and
> really poor FM in my house - until I removed my powerline stuff and
> replaced it with a mesh. Now both FM and DAB are just fine.
all on FM and DAB has never really worked in North Yorkshire. DAB was
flawed by design and made worse by cutting the bitrate even further. The
only thing it does better than FM are the inter program gaps.
I have discovered a new problem with DAB radios in an emergency power
cut situation - they eat a set of 4xAA batteries every 8 hours runtime.
By comparison a classic FM radio will last about a week at 8hours/day.
Smartphones last a little longer but most are stone dead within 30
hours. Not much use when Powergrid can't get their act together and
their website and phone helpline shows incorrect status information :(
> By the time I discovered this I was already in the habit using getting
> all of my radio from streams but I don't miss the powerline.
>
> My setup is an Asus Lyra Mini with three APs. Since I bought it there
> have been quite a few negative reviews but it works fine for me, the
> RJ45 sockets on the APs supporting the things I have that don't
> support wifi.
>
> My latest usb dongle on my old Surface is connecting at 866Mbs.
these days at the cost of bonding loads of channels. It can get hairy in
terraces or blocks of flats with nearest neighbour interactions.
At the start of the pandemic a new work from home superwideband EE
device in a neighbours house blinded my systems to my own Wifi. The
signal level it gave in my house nearest them was higher than my own
Wifi. You should periodically check for competition on channels.
--
Regards,
Martin Brown
Chris Green
Dec 1, 2021, 5:33:04 AM12/1/21
to
Graham J <nob...@nowhere.co.uk> wrote:
> Mark wrote:
>
> [snip]
>
> >
> > I believe the pricier systems have more radios and more antennas so that
> > backhaul bandwidth within the mesh isn't shared with 'user access'
> > bandwidth, and the available bandwidth in any case should be greater.
> > However if the walls are very absorbing, its possible three BT 'discs'
> > may not be enough. You might need to be prepared to return it.
> >
> > If that turns out to be the case powerline might be a better way to go -
> > I have used five Netgear DD-WRT routers in repeater bridge mode in a
> > student house with concrete walls and coverage is barely good enough. In
> > a less severe environment the same setup was exceptionally good.
>
>
> Ultimately the only fully reliable arrangement is to have all the Access
> Points connected by Ethernet cable to their master controller and the
> internet router. That way the backhaul is much more reliable and
> potentially much faster - possibly up to 10 Gbit/sec to every AP. Then
> site as many APs as are needed to guarantee coverage - certainly one in
> each room - perhaps several. The controller then manages all the APs
> and instructs which one is to communicate with each client, changing the
> assignments as the client moves around.
>
This only works if the clients play ball though, the controller can't
> Mark wrote:
>
> [snip]
>
> >
> > I believe the pricier systems have more radios and more antennas so that
> > backhaul bandwidth within the mesh isn't shared with 'user access'
> > bandwidth, and the available bandwidth in any case should be greater.
> > However if the walls are very absorbing, its possible three BT 'discs'
> > may not be enough. You might need to be prepared to return it.
> >
> > If that turns out to be the case powerline might be a better way to go -
> > I have used five Netgear DD-WRT routers in repeater bridge mode in a
> > student house with concrete walls and coverage is barely good enough. In
> > a less severe environment the same setup was exceptionally good.
>
>
> Ultimately the only fully reliable arrangement is to have all the Access
> Points connected by Ethernet cable to their master controller and the
> internet router. That way the backhaul is much more reliable and
> potentially much faster - possibly up to 10 Gbit/sec to every AP. Then
> site as many APs as are needed to guarantee coverage - certainly one in
> each room - perhaps several. The controller then manages all the APs
> and instructs which one is to communicate with each client, changing the
> assignments as the client moves around.
>
*tell* a client to use another AP, all it can do is drop hints.
Unless it actually drops the existing connection of course but that's
hardly optimal.
I guess that modern mobile phones at least do know about the hints
they're given and will work well in a mesh setup. I doubt if many
laptops play ball properly.
> This is really only practical in an industrial setting where flood wired
> Ethernet would be the norm. But in a domestic environment, whenever you
> redecorate it is always worth putting in Ethernet cables, so that in
> time you have a sensible flood wiring arrangement.
>
the line of site WiFi I was trying to improve.
--
Chris Green
·
Martin Brown
Dec 1, 2021, 5:41:52 AM12/1/21
to
On 01/12/2021 10:30, Chris Green wrote:
exceptionally noise mains. Powerline works for me at claimed speed on
the same ring main and fast enough for a network printer to be usable
across different ringmains (extension vs original build). The latter is
certainly not guaranteed (in fact the makers warn it won't work).
--
Regards,
Martin Brown
> Graham J <nob...@nowhere.co.uk> wrote:
>>
>> This is really only practical in an industrial setting where flood wired
>> Ethernet would be the norm. But in a domestic environment, whenever you
>> redecorate it is always worth putting in Ethernet cables, so that in
>> time you have a sensible flood wiring arrangement.
>>
> Yes, all my APs are wired. I did try powerline but it was worse than
> the line of site WiFi I was trying to improve.
That is unusual unless you have very high end bonded channel Wifi or
>>
>> This is really only practical in an industrial setting where flood wired
>> Ethernet would be the norm. But in a domestic environment, whenever you
>> redecorate it is always worth putting in Ethernet cables, so that in
>> time you have a sensible flood wiring arrangement.
>>
> Yes, all my APs are wired. I did try powerline but it was worse than
> the line of site WiFi I was trying to improve.
exceptionally noise mains. Powerline works for me at claimed speed on
the same ring main and fast enough for a network printer to be usable
across different ringmains (extension vs original build). The latter is
certainly not guaranteed (in fact the makers warn it won't work).
--
Regards,
Martin Brown
notya...@gmail.com
Dec 1, 2021, 6:05:27 AM12/1/21
to
A mesh system will be more resiliant (redundant paths etc.), but costly - the BT system monthly rental is about as much as a single TP-Link repeater (£15). Before I wired to my TV my repeater gave 4k video on it without a problem, despite ~300mm of walls in the way.
A mesh system will also give better performance where not all the traffic is out to the internet (e.g. NAS drives, printers, CCTV cameras).
OTOH the TP -Link devices work very well, especially if you can wire them to the routers and / or set them up to get a reasonable signal at 5GHz.
Roderick Stewart
Dec 1, 2021, 6:27:10 AM12/1/21
to
On Tue, 30 Nov 2021 21:52:52 +0000, Graham J <nob...@nowhere.co.uk>
wrote:
>Ultimately the only fully reliable arrangement is to have all the Access
>Points connected by Ethernet cable to their master controller and the
>internet router.
This needs repeating, so I'm repeating it.
Rod.
wrote:
>Ultimately the only fully reliable arrangement is to have all the Access
>Points connected by Ethernet cable to their master controller and the
>internet router.
Rod.
Chris Green
Dec 1, 2021, 6:33:03 AM12/1/21
to
Martin Brown <'''newspam'''@nonad.co.uk> wrote:
circuits. Except on almost adjacent sockets (i.e. on the same ring in
adjacent rooms) I never got more than a few Mb/s from my trials of
Powerline adapters. I guess that would be OK for a network printer
but I was hoping to avoid adding some more UTP to give me 'proper'
fast ethernet or gigabit connections.
--
Chris Green
·
Graham J
Dec 1, 2021, 6:35:52 AM12/1/21
to
Martin Brown wrote:
[snip]
>
> You must be in a very marginal signal area then.
I think that is true for anywhere other than a large city.
> I have no problems at
> all on FM and DAB has never really worked in North Yorkshire. DAB was
> flawed by design and made worse by cutting the bitrate even further.
+1 !!!
> I have discovered a new problem with DAB radios in an emergency power
> cut situation - they eat a set of 4xAA batteries every 8 hours runtime.
> By comparison a classic FM radio will last about a week at 8hours/day.
Tends to be true of all modern portable radios. The early transistor
radios of my teenage years (1960s) would work for many months on a set
of batteries. The crossover distortion was horrible, of course.
I don't understand why DAB should be so power hungry. Surely the
receivers are all highly integrated chips ?
> Smartphones last a little longer but most are stone dead within 30
> hours.
That's clearly a marketing decision. Smartphones are really costume
jewellery and priced accordingly. Battery life was never an important
consideration. Early 2G & 3G voice-only mobiles were much less power
hungry, using 20-year old technology.
> Not much use when Powergrid can't get their act together and
> their website and phone helpline shows incorrect status information :(
That's inexcusable! Just one person updating their website hourly would
resolve that.
--
Graham J
[snip]
>
> You must be in a very marginal signal area then.
> I have no problems at
> all on FM and DAB has never really worked in North Yorkshire. DAB was
> flawed by design and made worse by cutting the bitrate even further.
> I have discovered a new problem with DAB radios in an emergency power
> cut situation - they eat a set of 4xAA batteries every 8 hours runtime.
> By comparison a classic FM radio will last about a week at 8hours/day.
radios of my teenage years (1960s) would work for many months on a set
of batteries. The crossover distortion was horrible, of course.
I don't understand why DAB should be so power hungry. Surely the
receivers are all highly integrated chips ?
> Smartphones last a little longer but most are stone dead within 30
> hours.
jewellery and priced accordingly. Battery life was never an important
consideration. Early 2G & 3G voice-only mobiles were much less power
hungry, using 20-year old technology.
> Not much use when Powergrid can't get their act together and
> their website and phone helpline shows incorrect status information :(
resolve that.
--
Graham J
Graham J
Dec 1, 2021, 6:39:54 AM12/1/21
to
Chris Green wrote:
[snip]
>> The controller then manages all the APs
>> and instructs which one is to communicate with each client, changing the
>> assignments as the client moves around.
>>
> This only works if the clients play ball though, the controller can't
> *tell* a client to use another AP, all it can do is drop hints.
> Unless it actually drops the existing connection of course but that's
> hardly optimal.
The controller knows which clients are connected to each AP and can
easily disable an AP to force a client to renegotiate if there are no
other clients using it.
> I guess that modern mobile phones at least do know about the hints
> they're given and will work well in a mesh setup. I doubt if many
> laptops play ball properly.
I think any laptop newer than about 5 years is probably OK.
--
Graham J
[snip]
>> The controller then manages all the APs
>> and instructs which one is to communicate with each client, changing the
>> assignments as the client moves around.
>>
> This only works if the clients play ball though, the controller can't
> *tell* a client to use another AP, all it can do is drop hints.
> Unless it actually drops the existing connection of course but that's
> hardly optimal.
easily disable an AP to force a client to renegotiate if there are no
other clients using it.
> I guess that modern mobile phones at least do know about the hints
> they're given and will work well in a mesh setup. I doubt if many
> laptops play ball properly.
--
Graham J
Tim+
Dec 1, 2021, 7:01:43 AM12/1/21
to
notya...@gmail.com <notya...@gmail.com> wrote:
> On Tuesday, 30 November 2021 at 20:00:47 UTC, Tim+ wrote:
>> We have a BT mesh system to provide seamless Wi-Fi throughout our house and
>> it works very well indeed. My daughter is having a lot of Wi-Fi issues
>> with the walls in her house so I was wondering about getting her one.
>>
>> When we bought it, there weren’t a lot of options on the market at the time
>> but a quick squiz at Amazon shows lots of new kids on the block from £49
>> for a three node system. The BT system it much pricier at £199.
>>
>> Anyone know if these cheaper ones are any good or have personal experience
>> of non-BT mesh systems? Must be fairly idiot proof to set up and decently
>> reliable. ;-)
>>
>> Tim
>>
>>
>> --
>> Please don't feed the trolls - to what?
>
> A mesh system will be more resiliant (redundant paths etc.), but costly -
> the BT system monthly rental
Rental? I would be buying a system.
> On Tuesday, 30 November 2021 at 20:00:47 UTC, Tim+ wrote:
>> We have a BT mesh system to provide seamless Wi-Fi throughout our house and
>> it works very well indeed. My daughter is having a lot of Wi-Fi issues
>> with the walls in her house so I was wondering about getting her one.
>>
>> When we bought it, there weren’t a lot of options on the market at the time
>> but a quick squiz at Amazon shows lots of new kids on the block from £49
>> for a three node system. The BT system it much pricier at £199.
>>
>> Anyone know if these cheaper ones are any good or have personal experience
>> of non-BT mesh systems? Must be fairly idiot proof to set up and decently
>> reliable. ;-)
>>
>> Tim
>>
>>
>> --
>> Please don't feed the trolls - to what?
>
> A mesh system will be more resiliant (redundant paths etc.), but costly -
> the BT system monthly rental
> is about as much as a single TP-Link repeater (£15).
access points and often need manual re-selection of the best point.
Tim+
Dec 1, 2021, 7:01:44 AM12/1/21
to
possibly work even better with them, but it works fine as it is.
Chris Green
Dec 1, 2021, 7:03:04 AM12/1/21
to
Graham J <nob...@nowhere.co.uk> wrote:
> Chris Green wrote:
>
> [snip]
>
> >> The controller then manages all the APs
> >> and instructs which one is to communicate with each client, changing the
> >> assignments as the client moves around.
> >>
> > This only works if the clients play ball though, the controller can't
> > *tell* a client to use another AP, all it can do is drop hints.
> > Unless it actually drops the existing connection of course but that's
> > hardly optimal.
>
> The controller knows which clients are connected to each AP and can
> easily disable an AP to force a client to renegotiate if there are no
> other clients using it.
>
I don't think there's ever an AP with only one client in our house.
> Chris Green wrote:
>
> [snip]
>
> >> The controller then manages all the APs
> >> and instructs which one is to communicate with each client, changing the
> >> assignments as the client moves around.
> >>
> > This only works if the clients play ball though, the controller can't
> > *tell* a client to use another AP, all it can do is drop hints.
> > Unless it actually drops the existing connection of course but that's
> > hardly optimal.
>
> The controller knows which clients are connected to each AP and can
> easily disable an AP to force a client to renegotiate if there are no
> other clients using it.
>
... and as I said if the AP *forces* a client to renegotiate that's
effectively a dropped connection which will be noticeable if using it
as a phone (i.e.VoWiFi) or streaming something.
> > I guess that modern mobile phones at least do know about the hints
> > they're given and will work well in a mesh setup. I doubt if many
> > laptops play ball properly.
>
> I think any laptop newer than about 5 years is probably OK.
>
--
Chris Green
·
Tweed
Dec 1, 2021, 12:03:40 PM12/1/21
to
without Ethernet backhaul in the domestic environment.
NY
Dec 1, 2021, 12:12:17 PM12/1/21
to
"Tim+" <tim.d...@gmail.com> wrote in message
news:2086218988.659994992.175196.tim.downie-gmail.com@news.individual.net...
We have a Linksys Velop system which took a fair amount of setting up to get
the nodes placed optimally: close enough that the nodes can see each other
(or at least see one central "parent" node) on 5 GHz used for the backhaul
node-to-node comms, and yet as far away from each other as possible so they
don't throw their toys out of the pram because they can't find
non-overlapping 2.4 GHz wifi channels.
We live in an L-shaped house with fairly thick walls (some of the house
dates from the 1800s) and the best location for the parent node is at the
end of one of the arms of the L - this makes for easy cable runs to my study
with Ethernet connected PC and to the Plex client (a Roku box) in the lounge
for watching recorded TV from my PC "server".
That needs 6 nodes to cover various rooms of the house, as well as the
workshop and part of the garden.
It works beautifully, with seamless handover as you walk round the house
with a mobile/tablet/laptop. The only problem is if there is a power cut and
all the nodes turn back on simultaneously. Some nodes fail to boot because
they are fighting for 2.4 GHz channels. The solution is to start them in a
specific order (determined by trial and error) which is a bit of a pain.
If I could turn off 2.4 GHz then the problem would go away, but some older
devices need it because they don't talk 5 GHz, and also 2.4 carries further
in the garden. Unfortunately 2.4 is all-or-nothing: you can't turn it on for
just a few nodes and off for all the rest.
news:2086218988.659994992.175196.tim.downie-gmail.com@news.individual.net...
the nodes placed optimally: close enough that the nodes can see each other
(or at least see one central "parent" node) on 5 GHz used for the backhaul
node-to-node comms, and yet as far away from each other as possible so they
don't throw their toys out of the pram because they can't find
non-overlapping 2.4 GHz wifi channels.
We live in an L-shaped house with fairly thick walls (some of the house
dates from the 1800s) and the best location for the parent node is at the
end of one of the arms of the L - this makes for easy cable runs to my study
with Ethernet connected PC and to the Plex client (a Roku box) in the lounge
for watching recorded TV from my PC "server".
That needs 6 nodes to cover various rooms of the house, as well as the
workshop and part of the garden.
It works beautifully, with seamless handover as you walk round the house
with a mobile/tablet/laptop. The only problem is if there is a power cut and
all the nodes turn back on simultaneously. Some nodes fail to boot because
they are fighting for 2.4 GHz channels. The solution is to start them in a
specific order (determined by trial and error) which is a bit of a pain.
If I could turn off 2.4 GHz then the problem would go away, but some older
devices need it because they don't talk 5 GHz, and also 2.4 carries further
in the garden. Unfortunately 2.4 is all-or-nothing: you can't turn it on for
just a few nodes and off for all the rest.
Roderick Stewart
Dec 1, 2021, 12:13:35 PM12/1/21
to
On Wed, 1 Dec 2021 11:35:21 +0000, Graham J <nob...@nowhere.co.uk>
wrote:
>I don't understand why DAB should be so power hungry. Surely the
>receivers are all highly integrated chips ?
A digital radio needs all the analogue circuitry that an AM or FM
radio would need just to receive the signal, amplify it, detect it and
then amplify the audio for speaker or headphones *plus* all the extra
circuitry to do the digital decoding. A typical analogue radio can
have half a dozen transistors, while the digital circuitry probably
uses millions. No matter how efficient that digital circuitry can be
made, its power consumption cannot be reduced to zero. Therefore
digital receivers will always be inherently more power hungry than
analogue ones.
Rod.
wrote:
>I don't understand why DAB should be so power hungry. Surely the
>receivers are all highly integrated chips ?
radio would need just to receive the signal, amplify it, detect it and
then amplify the audio for speaker or headphones *plus* all the extra
circuitry to do the digital decoding. A typical analogue radio can
have half a dozen transistors, while the digital circuitry probably
uses millions. No matter how efficient that digital circuitry can be
made, its power consumption cannot be reduced to zero. Therefore
digital receivers will always be inherently more power hungry than
analogue ones.
Rod.
Roderick Stewart
Dec 1, 2021, 12:18:33 PM12/1/21
to
general a cable connection is better than a wireless one. For what
it's worth, my home setup is partly cabled, partly wireless, like a
lot of them I expect, and it works as well as I need it to.
Rod.
NY
Dec 1, 2021, 12:25:34 PM12/1/21
to
"Roderick Stewart" <rj...@escapetime.myzen.co.uk> wrote in message
news:cvmeqglkucvu20nmr...@4ax.com...
If I was designing a new house, I would always install Cat 7 cable and
Ethernet sockets in all rooms, feeding back to a location with phone/fibre.
Individual rooms/parts of the house could then have wifi access points for
portable devices.
But for an existing house you have the problem of retro-installing Cat 7
without disturbing the decorations. The main problems are:
- if you run cables in the loft, you have to get them through the ceiling of
the room below, down the wall to a skirting-board-level socket
- if you run cables around the edge of rooms, tucked down the side of
carpets, you have to get the cable through doorways, maybe *across* a
doorway - and you can't tuck cables down the edge of hardwood floors ;-)
news:cvmeqglkucvu20nmr...@4ax.com...
Ethernet sockets in all rooms, feeding back to a location with phone/fibre.
Individual rooms/parts of the house could then have wifi access points for
portable devices.
But for an existing house you have the problem of retro-installing Cat 7
without disturbing the decorations. The main problems are:
- if you run cables in the loft, you have to get them through the ceiling of
the room below, down the wall to a skirting-board-level socket
- if you run cables around the edge of rooms, tucked down the side of
carpets, you have to get the cable through doorways, maybe *across* a
doorway - and you can't tuck cables down the edge of hardwood floors ;-)
NY
Dec 1, 2021, 12:25:34 PM12/1/21
to
"Martin Brown" <'''newspam'''@nonad.co.uk> wrote in message
news:so7jhf$l78$1...@gioia.aioe.org...
Our house has two consumer units ("fuse boxes") but even on the same ring
main (and therefore obviously the same fuse box) I found I got spectacularly
low data rates (*) between two powerline devices that were only one room
apart. I toyed with using powerline/wifi devices to provide coverage around
the house, before abandoning powerline as useless in our house. Mesh network
works fine for us.
(*) A few Mbps, instead of the claimed 400 Mbps, with frequent dropouts.
Chocolate teapot territory!
news:so7jhf$l78$1...@gioia.aioe.org...
main (and therefore obviously the same fuse box) I found I got spectacularly
low data rates (*) between two powerline devices that were only one room
apart. I toyed with using powerline/wifi devices to provide coverage around
the house, before abandoning powerline as useless in our house. Mesh network
works fine for us.
(*) A few Mbps, instead of the claimed 400 Mbps, with frequent dropouts.
Chocolate teapot territory!
Tweed
Dec 1, 2021, 12:41:23 PM12/1/21
to
be installing a Wi-Fi access point. There’s no reason why a second link, ie
to the AP in a mesh system shouldn’t work equally well. The key to it all
is doing it properly, both in selecting hardware that has sufficient radios
to avoid sharing the backhaul link with the clients, and properly located
such that the backhaul link is reliable. Granted, if you can’t establish a
reliable backhaul link then cable is the only alternative.
Graham J
Dec 1, 2021, 1:55:20 PM12/1/21
to
milliamp. So no excuse, really.
--
Graham J
Andy Burns
Dec 1, 2021, 2:26:01 PM12/1/21
to
Roderick Stewart wrote:
> Graham J wrote:
>
>> Ultimately the only fully reliable arrangement is to have all the Access
>> Points connected by Ethernet cable to their master controller and the
>> internet router.
>
> This needs repeating, so I'm repeating it.
I think you can get quite close to wired backhaul, without actually having wired
backhaul, by using tri-band access points, so that e.g. 2.4GHz and 5GHz radios
are for clients, with a 6GHz radio (or an additional 5GHz radio) is used for
wireless backhaul, which may even exceed 1Gbe wired backhaul.
> Graham J wrote:
>
>> Ultimately the only fully reliable arrangement is to have all the Access
>> Points connected by Ethernet cable to their master controller and the
>> internet router.
>
> This needs repeating, so I'm repeating it.
backhaul, by using tri-band access points, so that e.g. 2.4GHz and 5GHz radios
are for clients, with a 6GHz radio (or an additional 5GHz radio) is used for
wireless backhaul, which may even exceed 1Gbe wired backhaul.
Martin Brown
Dec 1, 2021, 3:09:59 PM12/1/21
to
thin wires or circuit breakers? I'm at a loss to explain why our
experience is so different. Mine are nothing special - cheapest
reputable ones I could find at the time about 6 years ago maybe more.
My fuse boxes are both modern circuit breaker type except in the garage
where the old house fuse box of prehistoric wire fuses still lives.
> (*) A few Mbps, instead of the claimed 400 Mbps, with frequent dropouts.
> Chocolate teapot territory!
power cycle is enough to make it go again. Our mains fails often enough
that such resets are usually not needed.
Last weekend and storm Arwen was very entertaining.
Northern Powergrid customer support leaves a lot to be desired. They
managed to tag our village as being "On Supply" despite having 33kV
lines on the ground, circuit breakers off and nothing we could do would
persuade them otherwise. Calls to 105 were dropped on the floor :(
What customer support?
I felt sorry for the engineers who eventually turned up to find that the
information they had was completely wrong. Individual houses off supply.
Because of that they had to take out consumer main fuses and check the
cable to the house before they could actually begin to even think about
doing anything useful. What a waste of time and effort!
Tick box mentality writ large and no scope for the engineers to take
villagers word for it that it wasn't just *them* it was the entire
village. A situation made more confusing because many farms and some
businesses have their own generators because supply here is so dodgy.
--
Regards,
Martin Brown
NY
Dec 1, 2021, 4:00:37 PM12/1/21
to
"Martin Brown" <'''newspam'''@nonad.co.uk> wrote in message
news:so8kqi$1kgu$1...@gioia.aioe.org...
>> Our house has two consumer units ("fuse boxes") but even on the same ring
>> main (and therefore obviously the same fuse box) I found I got
>> spectacularly low data rates (*) between two powerline devices that were
>> only one room apart. I toyed with using powerline/wifi devices to provide
>> coverage around the house, before abandoning powerline as useless in our
>> house. Mesh network works fine for us.
>
> Are they really fuse boxes as in old style chunky pottery things with thin
> wires or circuit breakers? I'm at a loss to explain why our experience is
> so different. Mine are nothing special - cheapest reputable ones I could
> find at the time about 6 years ago maybe more.
They use circuit breakers. I put "fuse box" in quotes only because it is the
>> main (and therefore obviously the same fuse box) I found I got
>> spectacularly low data rates (*) between two powerline devices that were
>> only one room apart. I toyed with using powerline/wifi devices to provide
>> coverage around the house, before abandoning powerline as useless in our
>> house. Mesh network works fine for us.
>
> Are they really fuse boxes as in old style chunky pottery things with thin
> wires or circuit breakers? I'm at a loss to explain why our experience is
> so different. Mine are nothing special - cheapest reputable ones I could
> find at the time about 6 years ago maybe more.
more commonly-used term for what is strictly called a consumer unit. Each CU
also has a single RCD such that an earth-leakage fault anywhere on the
lighting or ring main circuits on that CU will trip its RCD without
affecting the RCD of the other CU. Shame that each circuit doesn't have its
own RDC to limit how many circuits trip, but this would add to cost.
I imagine there are two CUs because the house was extended about 20 years
ago and the existing CU probably didn't have enough spare circuit breaker
slots to accommodate the extra lighting and ring main circuits. It was
probably cheaper to add an extra CU in the new part of the house than to
remove the old CU, replace it with a larger one (rewiring all the circuits)
and additionally cable all the new circuits back to the same point.
Both CUs go through a single electricity meter. I understand that it is
often the meter which blocks the powerline signals so *in theory* both CUs
should be able to talk to each other. Certainly two ring mains on the *same*
CU can talk fine: at our old house, we had one powerline on the upstairs
ring and another on the other ring, to get Ethernet between the router
(upstairs) and the Sky box and Roku box by the TV downstairs. We achieved
about 150 Mbps with that configuration.
But in our present house, two powerlines in adjacent sockets of a
double-gang manage about 100 and one of the powerlines in another socket a
few metres away drops to 80, and then slightly further away it's 50. All on
the same ring main, as proved by turning off a circuit breaker and seeing
which sockets lose power.
These tests were for the older CU and wiring, because that's where the
router is. I've not tested between sockets on the same ring of the newer CU
(dating from about 2005, I believe) because any realistic internet comms
would be between a computer and the router. Maybe I will investigate one
day...
NY
Dec 1, 2021, 4:08:05 PM12/1/21
to
"Andy Burns" <use...@andyburns.uk> wrote in message
news:j0q0i8...@mid.individual.net...
I've tested a Windows 10 laptop copying multi-GB files to/from a Windows 7
PC. Win 10 is connected by 5 GHz wifi, probably with one 5 GHz hop: laptop
to node A, node A to node B, node B by Ethernet to Win7 PC. And I achieved
transfer rates of around 0.3 - 0.5 Gbps.
Win 10 connected by Ethernet (instead of wifi) to same switch as Win 7
achieved rates of about 0.9 Mbps.
What was different was the transfer-rate graphs: over Ethernet the rate was
fairly constant with occasional blips where data is being read from or
written to HDD. With one leg using wifi, the peak rate was about the same
but there were a lot more fluctuations in speed.
news:j0q0i8...@mid.individual.net...
PC. Win 10 is connected by 5 GHz wifi, probably with one 5 GHz hop: laptop
to node A, node A to node B, node B by Ethernet to Win7 PC. And I achieved
transfer rates of around 0.3 - 0.5 Gbps.
Win 10 connected by Ethernet (instead of wifi) to same switch as Win 7
achieved rates of about 0.9 Mbps.
What was different was the transfer-rate graphs: over Ethernet the rate was
fairly constant with occasional blips where data is being read from or
written to HDD. With one leg using wifi, the peak rate was about the same
but there were a lot more fluctuations in speed.
Martin Brown
Dec 2, 2021, 4:52:10 AM12/2/21
to
On 01/12/2021 20:59, NY wrote:
> "Martin Brown" <'''newspam'''@nonad.co.uk> wrote in message
> news:so8kqi$1kgu$1...@gioia.aioe.org...
>>> Our house has two consumer units ("fuse boxes") but even on the same
>>> ring main (and therefore obviously the same fuse box) I found I got
>>> spectacularly low data rates (*) between two powerline devices that
>>> were only one room apart. I toyed with using powerline/wifi devices
>>> to provide coverage around the house, before abandoning powerline as
>>> useless in our house. Mesh network works fine for us.
>>
>> Are they really fuse boxes as in old style chunky pottery things with
>> thin wires or circuit breakers? I'm at a loss to explain why our
>> experience is so different. Mine are nothing special - cheapest
>> reputable ones I could find at the time about 6 years ago maybe more.
>
> They use circuit breakers. I put "fuse box" in quotes only because it is
> the more commonly-used term for what is strictly called a consumer unit.
> Each CU also has a single RCD such that an earth-leakage fault anywhere
> on the lighting or ring main circuits on that CU will trip its RCD
> without affecting the RCD of the other CU. Shame that each circuit
> doesn't have its own RDC to limit how many circuits trip, but this would
> add to cost.
That may be an important topological difference. My configuration has a
> "Martin Brown" <'''newspam'''@nonad.co.uk> wrote in message
> news:so8kqi$1kgu$1...@gioia.aioe.org...
>>> Our house has two consumer units ("fuse boxes") but even on the same
>>> ring main (and therefore obviously the same fuse box) I found I got
>>> spectacularly low data rates (*) between two powerline devices that
>>> were only one room apart. I toyed with using powerline/wifi devices
>>> to provide coverage around the house, before abandoning powerline as
>>> useless in our house. Mesh network works fine for us.
>>
>> Are they really fuse boxes as in old style chunky pottery things with
>> thin wires or circuit breakers? I'm at a loss to explain why our
>> experience is so different. Mine are nothing special - cheapest
>> reputable ones I could find at the time about 6 years ago maybe more.
>
> They use circuit breakers. I put "fuse box" in quotes only because it is
> the more commonly-used term for what is strictly called a consumer unit.
> Each CU also has a single RCD such that an earth-leakage fault anywhere
> on the lighting or ring main circuits on that CU will trip its RCD
> without affecting the RCD of the other CU. Shame that each circuit
> doesn't have its own RDC to limit how many circuits trip, but this would
> add to cost.
single RCD on the supply side of all of the consumer units. Any earth
leakage on any circuit and the whole lot goes off.
> I imagine there are two CUs because the house was extended about 20
> years ago and the existing CU probably didn't have enough spare circuit
> breaker slots to accommodate the extra lighting and ring main circuits.
> It was probably cheaper to add an extra CU in the new part of the house
> than to remove the old CU, replace it with a larger one (rewiring all
> the circuits) and additionally cable all the new circuits back to the
> same point.
house was rewired at the same time so no very old oxidised wiring.
> Both CUs go through a single electricity meter. I understand that it is
> often the meter which blocks the powerline signals so *in theory* both
> CUs should be able to talk to each other. Certainly two ring mains on
> the *same* CU can talk fine: at our old house, we had one powerline on
> the upstairs ring and another on the other ring, to get Ethernet between
> the router (upstairs) and the Sky box and Roku box by the TV downstairs.
> We achieved about 150 Mbps with that configuration.
> But in our present house, two powerlines in adjacent sockets of a
> double-gang manage about 100 and one of the powerlines in another socket
> a few metres away drops to 80, and then slightly further away it's 50.
> All on the same ring main, as proved by turning off a circuit breaker
> and seeing which sockets lose power.
rectifying some of the signal but that is just a wild guess. I have seen
sockets with crystals of copper sulphate growing inside them in damp
premises!
> These tests were for the older CU and wiring, because that's where the
> router is. I've not tested between sockets on the same ring of the newer
> CU (dating from about 2005, I believe) because any realistic internet
> comms would be between a computer and the router. Maybe I will
> investigate one day...
don't need to chase the last bit of performance anything more than 5M
will allow the TV and radio to stream (until *very* recently my wet
string connection to the outside world was the limiting factor).
Even less bandwidth is needed to support the network printer.
--
Regards,
Martin Brown
Roderick Stewart
Dec 2, 2021, 5:30:58 AM12/2/21
to
On Wed, 1 Dec 2021 18:54:49 +0000, Graham J <nob...@nowhere.co.uk>
If an analogue radio uses six transistors, and a digital chip uses a
million, then a digital radio must use six plus a million transistors.
If there really is a way to make six plus a million transistors
consume less power than just six, I wonder why nobody's using it?
Rod.
million, then a digital radio must use six plus a million transistors.
If there really is a way to make six plus a million transistors
consume less power than just six, I wonder why nobody's using it?
Rod.
Martin Brown
Dec 2, 2021, 6:07:42 AM12/2/21
to
On 02/12/2021 10:30, Roderick Stewart wrote:
> On Wed, 1 Dec 2021 18:54:49 +0000, Graham J <nob...@nowhere.co.uk>
> wrote:
>
>> Roderick Stewart wrote:
>>> On Wed, 1 Dec 2021 11:35:21 +0000, Graham J <nob...@nowhere.co.uk>
>>> wrote:
>>>
>>>> I don't understand why DAB should be so power hungry. Surely the
>>>> receivers are all highly integrated chips ?
>>>
>>> A digital radio needs all the analogue circuitry that an AM or FM
>>> radio would need just to receive the signal, amplify it, detect it and
>>> then amplify the audio for speaker or headphones *plus* all the extra
>>> circuitry to do the digital decoding. A typical analogue radio can
>>> have half a dozen transistors, while the digital circuitry probably
>>> uses millions. No matter how efficient that digital circuitry can be
>>> made, its power consumption cannot be reduced to zero. Therefore
>>> digital receivers will always be inherently more power hungry than
>>> analogue ones.
This was the most frugal reflex radio design ever Sinclair Radionics:
> On Wed, 1 Dec 2021 18:54:49 +0000, Graham J <nob...@nowhere.co.uk>
> wrote:
>
>> Roderick Stewart wrote:
>>> On Wed, 1 Dec 2021 11:35:21 +0000, Graham J <nob...@nowhere.co.uk>
>>> wrote:
>>>
>>>> I don't understand why DAB should be so power hungry. Surely the
>>>> receivers are all highly integrated chips ?
>>>
>>> A digital radio needs all the analogue circuitry that an AM or FM
>>> radio would need just to receive the signal, amplify it, detect it and
>>> then amplify the audio for speaker or headphones *plus* all the extra
>>> circuitry to do the digital decoding. A typical analogue radio can
>>> have half a dozen transistors, while the digital circuitry probably
>>> uses millions. No matter how efficient that digital circuitry can be
>>> made, its power consumption cannot be reduced to zero. Therefore
>>> digital receivers will always be inherently more power hungry than
>>> analogue ones.
https://www.petervis.com/Radios/sinclair-micromatic-pocket-radio/sinclair-micromatic-pocket-radio.html
2 transistor reflex design one RF front end and one audio to drive an
earpiece. A loudspeaker minimal design needs 3 to be properly audible
and 4 to be sufficiently frugal with its power consumption.
>> I've met sophisiticated digital systems on a chip that take less than a
>> milliamp. So no excuse, really.
clock crystal at 10uA but there is a limit to how much work it can do.
(at that current draw battery life is marginally longer than shelf life)
> If an analogue radio uses six transistors, and a digital chip uses a
> million, then a digital radio must use six plus a million transistors.
>
> If there really is a way to make six plus a million transistors
> consume less power than just six, I wonder why nobody's using it?
clock speed that you have to run them at to do software defined radio
and DAB decoding. Modern DSP chips are typically in the 50k-100k gates
range somewhat under 10^6 transistors.
It is the clock speed needed that really hurts power consumption ~250mA.
By comparison a classic analogue transistor radio will work OK at around
50mA (provided that you don't have the volume ghetto blaster loud) and
perhaps as low as 10mA to drive an earpiece.
Current consumption scales with how loud you have it. My battery powered
DAB radio seems profligate with batteries - it expects mains power.The
I expect battery operation was added as an afterthought by marketing.
It sure as hell wasn't engineered!
--
Regards,
Martin Brown
notya...@gmail.com
Dec 2, 2021, 6:08:27 AM12/2/21
to
On Thursday, 2 December 2021 at 10:30:58 UTC, Roderick Stewart wrote:
> On Wed, 1 Dec 2021 18:54:49 +0000, Graham J <nob...@nowhere.co.uk>
> wrote:
>
> >Roderick Stewart wrote:
> >> On Wed, 1 Dec 2021 11:35:21 +0000, Graham J <nob...@nowhere.co.uk>
> >> wrote:
> >>
> >>> I don't understand why DAB should be so power hungry. Surely the
> >>> receivers are all highly integrated chips ?
> >>
> >> A digital radio needs all the analogue circuitry that an AM or FM
> >> radio would need just to receive the signal, amplify it, detect it and
> >> then amplify the audio for speaker or headphones *plus* all the extra
> >> circuitry to do the digital decoding. A typical analogue radio can
> >> have half a dozen transistors, while the digital circuitry probably
> >> uses millions. No matter how efficient that digital circuitry can be
> >> made, its power consumption cannot be reduced to zero. Therefore
> >> digital receivers will always be inherently more power hungry than
> >> analogue ones.
> >
> >I've met sophisiticated digital systems on a chip that take less than a
> >milliamp. So no excuse, really.
> If an analogue radio uses six transistors, and a digital chip uses a
> million, then a digital radio must use six plus a million transistors.
Not at all a 1974 Intel 8080A only had 4,500 and a Motorola 68000 (32-bit MPU 1979) had eponymously 68,000.
> On Wed, 1 Dec 2021 18:54:49 +0000, Graham J <nob...@nowhere.co.uk>
> wrote:
>
> >Roderick Stewart wrote:
> >> On Wed, 1 Dec 2021 11:35:21 +0000, Graham J <nob...@nowhere.co.uk>
> >> wrote:
> >>
> >>> I don't understand why DAB should be so power hungry. Surely the
> >>> receivers are all highly integrated chips ?
> >>
> >> A digital radio needs all the analogue circuitry that an AM or FM
> >> radio would need just to receive the signal, amplify it, detect it and
> >> then amplify the audio for speaker or headphones *plus* all the extra
> >> circuitry to do the digital decoding. A typical analogue radio can
> >> have half a dozen transistors, while the digital circuitry probably
> >> uses millions. No matter how efficient that digital circuitry can be
> >> made, its power consumption cannot be reduced to zero. Therefore
> >> digital receivers will always be inherently more power hungry than
> >> analogue ones.
> >
> >I've met sophisiticated digital systems on a chip that take less than a
> >milliamp. So no excuse, really.
> If an analogue radio uses six transistors, and a digital chip uses a
> million, then a digital radio must use six plus a million transistors.
>
> If there really is a way to make six plus a million transistors
> consume less power than just six, I wonder why nobody's using it?
Yes there is and yes they are: -
> If there really is a way to make six plus a million transistors
> consume less power than just six, I wonder why nobody's using it?
Smaller transistors - far less current.
FET's, especially CMOS - very low switching currents.
Lower voltage - less resistive heating.
In any event the majority of the energy used in any modern radio is driving the speaker. A crystal set using an earpiece, doesn't need a battery, nor even a manufactured semiconductor device. One I built in ~1968 worked well with a contact diode, but still gave passable performance with a galena crystal I found whilst walking in the Ochil Hills.
>
> Rod.
Martin Brown
Dec 2, 2021, 6:29:19 AM12/2/21
to
On 02/12/2021 11:08, notya...@gmail.com wrote:
>
> In any event the majority of the energy used in any modern radio is
> driving the speaker. A crystal set using an earpiece, doesn't need a
Not any more it isn't.
>
> In any event the majority of the energy used in any modern radio is
> driving the speaker. A crystal set using an earpiece, doesn't need a
My DAB radio uses ~250mA when operating irrespective of volume and
enough in standby that you have to take the batteries out overnight or
install new ones on a daily basis. It kills a set every 8h of runtime.
I discovered this to my cost during our recent 2 day power outage.
The ancient analogue Sony (now deceased) it replaced would last more
than a week on new batteries and used just 3 rather than 4 AA cells.
> battery, nor even a manufactured semiconductor device. One I built
> in ~1968 worked well with a contact diode, but still gave passable
> performance with a galena crystal I found whilst walking in the Ochil
> Hills.
--
Regards,
Martin Brown
Roderick Stewart
Dec 2, 2021, 7:56:46 AM12/2/21
to
On Thu, 2 Dec 2021 11:07:38 +0000, Martin Brown
<'''newspam'''@nonad.co.uk> wrote:
[..]
In an analogue radio the clock speed is effectively zero because the
digital circuitry is absent. Whatever clock speed a digital radio runs
at it must be greater than zero. Something is greater than nothing.
Everything else about the radio - the audio amplifier(s) and
speaker(s) etc will be the same for the same type of radio (e.g.
pocket, bedside, or hi-fi tuner) because it's required regardless. The
only difference is the presence or absence of the digital circuitry,
which however efficient it is, cannot be made to run on nothing.
Rod.
<'''newspam'''@nonad.co.uk> wrote:
[..]
>>> I've met sophisiticated digital systems on a chip that take less than a
>>> milliamp. So no excuse, really.
>
>But at what clock speed? I can get a PIC16877 to tick over on a 32kHz
>clock crystal at 10uA but there is a limit to how much work it can do.
>(at that current draw battery life is marginally longer than shelf life)
>
>> If an analogue radio uses six transistors, and a digital chip uses a
>> million, then a digital radio must use six plus a million transistors.
>>
>> If there really is a way to make six plus a million transistors
>> consume less power than just six, I wonder why nobody's using it?
>
>It isn't really the number of transistors that hurts so much as the
>clock speed that you have to run them at to do software defined radio
>and DAB decoding. Modern DSP chips are typically in the 50k-100k gates
>range somewhat under 10^6 transistors.
[...]
>>> milliamp. So no excuse, really.
>
>But at what clock speed? I can get a PIC16877 to tick over on a 32kHz
>clock crystal at 10uA but there is a limit to how much work it can do.
>(at that current draw battery life is marginally longer than shelf life)
>
>> If an analogue radio uses six transistors, and a digital chip uses a
>> million, then a digital radio must use six plus a million transistors.
>>
>> If there really is a way to make six plus a million transistors
>> consume less power than just six, I wonder why nobody's using it?
>
>It isn't really the number of transistors that hurts so much as the
>clock speed that you have to run them at to do software defined radio
>and DAB decoding. Modern DSP chips are typically in the 50k-100k gates
>range somewhat under 10^6 transistors.
In an analogue radio the clock speed is effectively zero because the
digital circuitry is absent. Whatever clock speed a digital radio runs
at it must be greater than zero. Something is greater than nothing.
Everything else about the radio - the audio amplifier(s) and
speaker(s) etc will be the same for the same type of radio (e.g.
pocket, bedside, or hi-fi tuner) because it's required regardless. The
only difference is the presence or absence of the digital circuitry,
which however efficient it is, cannot be made to run on nothing.
Rod.
Martin Brown
Dec 2, 2021, 3:23:20 PM12/2/21
to
On 02/12/2021 12:56, Roderick Stewart wrote:
> On Thu, 2 Dec 2021 11:07:38 +0000, Martin Brown
> <'''newspam'''@nonad.co.uk> wrote:
> [..]
>>>> I've met sophisiticated digital systems on a chip that take less than a
>>>> milliamp. So no excuse, really.
>>
>> But at what clock speed? I can get a PIC16877 to tick over on a 32kHz
>> clock crystal at 10uA but there is a limit to how much work it can do.
>> (at that current draw battery life is marginally longer than shelf life)
>>
>>> If an analogue radio uses six transistors, and a digital chip uses a
>>> million, then a digital radio must use six plus a million transistors.
>>>
>>> If there really is a way to make six plus a million transistors
>>> consume less power than just six, I wonder why nobody's using it?
>>
>> It isn't really the number of transistors that hurts so much as the
>> clock speed that you have to run them at to do software defined radio
>> and DAB decoding. Modern DSP chips are typically in the 50k-100k gates
>> range somewhat under 10^6 transistors.
> [...]
>
> In an analogue radio the clock speed is effectively zero because the
> digital circuitry is absent. Whatever clock speed a digital radio runs
> at it must be greater than zero. Something is greater than nothing.
That would be a very silent radio.
> On Thu, 2 Dec 2021 11:07:38 +0000, Martin Brown
> <'''newspam'''@nonad.co.uk> wrote:
> [..]
>>>> I've met sophisiticated digital systems on a chip that take less than a
>>>> milliamp. So no excuse, really.
>>
>> But at what clock speed? I can get a PIC16877 to tick over on a 32kHz
>> clock crystal at 10uA but there is a limit to how much work it can do.
>> (at that current draw battery life is marginally longer than shelf life)
>>
>>> If an analogue radio uses six transistors, and a digital chip uses a
>>> million, then a digital radio must use six plus a million transistors.
>>>
>>> If there really is a way to make six plus a million transistors
>>> consume less power than just six, I wonder why nobody's using it?
>>
>> It isn't really the number of transistors that hurts so much as the
>> clock speed that you have to run them at to do software defined radio
>> and DAB decoding. Modern DSP chips are typically in the 50k-100k gates
>> range somewhat under 10^6 transistors.
> [...]
>
> In an analogue radio the clock speed is effectively zero because the
> digital circuitry is absent. Whatever clock speed a digital radio runs
> at it must be greater than zero. Something is greater than nothing.
Even the poxiest ones do 10kHz audio and hifi is good to around 20kHz.
Human hearing frequency response tails off quite rapidly after 15kHz.
The bulk of the power consumption goes to frequencies under 1000Hz.
> Everything else about the radio - the audio amplifier(s) and
> speaker(s) etc will be the same for the same type of radio (e.g.
> pocket, bedside, or hi-fi tuner) because it's required regardless. The
> only difference is the presence or absence of the digital circuitry,
> which however efficient it is, cannot be made to run on nothing.
mine draws is to run the DAB tuner/decoder chip. The power for the
speaker output is just a tiny fraction of the total current draw.
DAB needs roughly 5x as much power as classical AM/FM radio. The DAB
decoder chips were never designed with portability in mind.
--
Regards,
Martin Brown
notya...@gmail.com
Dec 4, 2021, 7:54:29 AM12/4/21
to
To get Radio Peking on MW* (logged) I used a six transistor Wien radio and about ten metres of external aerial.
* just after sunset, so two+ ionosphere bounces to reach Scotland - "now we do our morning exercises" - quite good reception for few minutes.
Roderick Stewart
Dec 4, 2021, 8:29:56 AM12/4/21
to
transmitter, even if it's not one you want to listen to, is to build a
receiver with two tuned inputs (or call it two receivers if you like)
where the one that is tuned to the powerful transmitter is a passive
receiver that rectifies the carrier and smooths out the modulation to
provide power for the other one, which can include some amplification.
This wouldn't suit every situation of course, but I think some people
have actually done it.
Rod.
0 new messages