Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.misc > #87088 > unrolled thread

Recent Experience With RF "Modem-ish" Data Links ?

Back to article view | Back to comp.os.linux.misc

Contents

  Recent Experience With RF "Modem-ish" Data Links ? c186282 <c186282@nnada.net> - 2026-05-25 01:46 -0400
    Re: Recent Experience With RF "Modem-ish" Data Links ? rbowman <bowman@montana.com> - 2026-05-25 17:56 +0000
      Re: Recent Experience With RF "Modem-ish" Data Links ? c186282 <c186282@nnada.net> - 2026-05-26 01:27 -0400
        Re: Recent Experience With RF "Modem-ish" Data Links ? TheLastSysop <thelastsysop@dev.null> - 2026-05-26 22:11 +0000
          Re: Recent Experience With RF "Modem-ish" Data Links ? c186282 <c186282@nnada.net> - 2026-05-26 23:04 -0400
            Re: Recent Experience With RF "Modem-ish" Data Links ? rbowman <bowman@montana.com> - 2026-05-27 03:21 +0000
              Re: Recent Experience With RF "Modem-ish" Data Links ? c186282 <c186282@nnada.net> - 2026-05-27 00:18 -0400
                Re: Recent Experience With RF "Modem-ish" Data Links ? rbowman <bowman@montana.com> - 2026-05-27 07:09 +0000
              Re: Recent Experience With RF "Modem-ish" Data Links ? The Natural Philosopher <tnp@invalid.invalid> - 2026-05-27 14:14 +0100
        WiFi range ... Lars Poulsen <lars@beagle-ears.com> - 2026-05-31 21:19 -0700
          Re: WiFi range ... TheLastSysop <thelastsysop@dev.null> - 2026-06-01 04:26 +0000
        Re: Recent Experience With RF "Modem-ish" Data Links ? Andy Burns <usenet@andyburns.uk> - 2026-06-01 08:10 +0100
          Re: Recent Experience With RF "Modem-ish" Data Links ? TheLastSysop <thelastsysop@dev.null> - 2026-06-01 07:33 +0000

#87088 — Recent Experience With RF "Modem-ish" Data Links ?

There WERE some long-time players. Can't find them
anymore.

I'm talking about 400/900 Mhz band "radio modems".
Many can deliver (slow) data over a good mile.

I've seen a few that are USB, but most attach to
whatever you designate as the serial ports on
your PI/Ard/Whatever. They can do RS-232, RS-485
preferably, over a relatively low speed radio link.

Some of the older players, they seem to be gone,
but there ARE more modern providers that offer
small/cheap transceivers.

400 MHz seems best, most range.

Such links are ideal for 'data devices' - where
you don't need live video or anything - just
periodic numbers. Micro-controller like devices,
including most Ards, can work OK on solar power
and only come alive for a moment on timer input.
Built them, chips and solder up, know.

Always DID look at the 'radio modems', but never
had a particular, needed, USE for them.

Note : solar-powered, use the "Lipo-Rider Pro"
power/charge device from Seeed ... most of the
others do NOT reg low-draw voltage very well -
which can get rather high depending on what
yer solar cells receive. Tested many. Amazon
re-sells these. DO pick yer batteries well
however, had ONE explode when I touched it,
months after it had been recharged. Giant
crimson flame and big mass of nasty smoke.
When I retired one employee noted how I just
calmly watched it burn ... well ... nothing
to DO about it until it's done ....  :-)

Put all such batteries into a metal coffee can
marked "HAZARD" when I retired and put it on
an outside shelf. HOPE they found a place to
dump them ..... maybe a six-foot hole in the
ground might be OK .......

Anyway ... NOW I may have a use for RF "modems",
a link to a kind of far-off property site. Do
not need video, MAYbe the odd still image. The
low speed isn't relevant, just the range. WiFi
isn't gonna cut it at all.

[toc] | [next] | [standalone]

#87115

On Mon, 25 May 2026 01:46:15 -0400, c186282 wrote:

> Anyway ... NOW I may have a use for RF "modems",
> a link to a kind of far-off property site. Do not need video, MAYbe the
> odd still image. The low speed isn't relevant, just the range. WiFi
> isn't gonna cut it at all.

How far off?  The nrf24L01 modules have been around for a long time and 
can go out to 100m. They can be finicky. There are a number of LoRa 
modules with greater range.

https://community.element14.com/technologies/internet-of-things/b/blog/
posts/a-comprehensive-guide-to-lora-modules

[toc] | [prev] | [next] | [standalone]

#87130

On 5/25/26 13:56, rbowman wrote:
> On Mon, 25 May 2026 01:46:15 -0400, c186282 wrote:
> 
>> Anyway ... NOW I may have a use for RF "modems",
>> a link to a kind of far-off property site. Do not need video, MAYbe the
>> odd still image. The low speed isn't relevant, just the range. WiFi
>> isn't gonna cut it at all.
> 
> How far off? 


   About 1000 meters.

   NO power utilities there.

   COULD, in theory, get a 5G router - but would
   have to have a much bigger solar power setup.
   The ner-do-wells might SEE it, tear it up.

   This is an eternal modern problem. Did a number
   of field environmental units awhile back, solar,
   5W panels (enough for ARDs that 'sleep' in between
   recordings). But - had to DISGUISE them ... as
   fake veggies. Some mil-OD canvas strips, a little
   brown spray paint, works pretty well. You have to
   get really close to notice them.

   Am interested in 'weather/environ' info and
   maybe still frames from a low-rez THERMAL
   camera. A 'live mic' might be interesting
   even at rather low bandwidth.


> The nrf24L01 modules have been around for a long time and
> can go out to 100m. They can be finicky. There are a number of LoRa
> modules with greater range.
> 
> https://community.element14.com/technologies/internet-of-things/b/blog/
> posts/a-comprehensive-guide-to-lora-modules

   Have no experience with "Lo-Ra" ... though it looks
   pretty good on paper. I think it's popular for
   'smart farm' applications especially.

   The 400mhz band 'modems' can indeed deliver out to
   1000 meters or beyond. As said, not too much SPEED ...
   but I don't need that much speed.

   Radio RS-485 ... 'antique' in some respects, but
   STILL HAS A PLACE. From research it still seems to
   be used in lots of 'industrial' facilities, replacing
   old hard wires. You want decent range and noise immunity.
   485 was 'multi-drop' from the get-go, so you can network
   a bunch of them across your facilities. The lower freq
   penetrates/diffuses-thru buildings much better than WiFi.

   Oh well, at least no 'ground potential' issues like
   with the old hard wires  :-)

   ANYway ... I am interested in the current commercial
   units - there is a selection. Some are serial, some
   plug into USB, think I saw an I2C version. Not sure
   what's "best" for the $$$ now. The old/good brands
   from the distant days seem gone.

   Does anyone have experiences with these newer units ?

   The new IT guy ... very smart with M$ packages but
   hardly any programming/hardware knowledge. That's what
   the new director wanted. Sure my environmental stations
   have rotted away by now. They'd better figure something
   out, eventually the lithiums WILL burst into flame !  :-)

   Note : Switching everything over to M$ (or Apple) cloud
   shit DOES have a big political advantage - if anything
   goes wrong it's NOT OUR FAULT. In short, accept expensive
   CRAP in return for deflecting any blame. This IS what
   it's come to.

   Just wait until Vlad/Xi/Kim declare total cyberwar ...
   those big concerns are gonna go DOWN - hard and
   forever. All your data, all your day-2-day - ZAP.

[toc] | [prev] | [next] | [standalone]

#87166

> About 1000 meters.
> NO power utilities there.
> Am interested in 'weather/environ' info and
> maybe still frames from a low-rez THERMAL
> camera. A 'live mic' might be interesting
> even at rather low bandwidth.

For 1 km and a small solar/battery box, split the requirements first.
Weather/environment telemetry is easy. Still frames are possible but slow. A
live microphone is the thing most likely to push you out of the cheap low-power
radio class.

LoRa is a good fit for small packets: sensor readings, battery voltage, status,
alarms, maybe a tiny thumbnail if you are patient. Do not expect it to behave
like slow WiFi. Depending on spreading factor and legal duty cycle limits, a few
kbit/s or less is a more realistic planning number than the headline figures.

The current cheap modules worth looking at are usually UART/SPI LoRa boards
based on Semtech SX127x/SX126x parts. Ebyte E32/E22/E220 style UART modules are
common if you want "serial cable replacement" behavior. Digi XBee/XBee-PRO
900HP, RFD900x, Microhard, FreeWave, etc. are more expensive but more finished
products. Check the band and power limits for your country; 433/868/915 MHz
modules are not interchangeable legally.

A few practical points:

* Use a real outdoor antenna, placed high and dry, before throwing power at the
problem. Fresnel clearance matters even at only 1 km. * Use packets with
sequence numbers and CRCs, not a bare byte stream, even if the radio sells
itself as transparent serial. * If you need multi-drop, handle
addressing/retries in your protocol. "Radio RS-485" boxes often only make a
transparent half-duplex serial link; collision behavior can be ugly. * For
battery life, let the remote node sleep and wake on a schedule. A Pi can work,
but a small MCU plus radio will be much easier to power and hide. * Send
summaries often and images rarely. A tiny thermal frame may be acceptable every
few minutes; audio probably wants a different link.

If you have line of sight, a pair of small directional 2.4/5 GHz outdoor CPE
units can do 1 km easily, but they draw watts continuously and are more visible.
For disguised, low-duty-cycle environmental telemetry, LoRa or a sub-GHz serial
modem is the more natural starting point.

-- 
TheLastSysop <thelastsysop@dev.null>
"I survived the great rm -rf / rehearsal and all I got was this .signature."

[toc] | [prev] | [next] | [standalone]

#87171

On 5/26/26 18:11, TheLastSysop wrote:
>> About 1000 meters.
>> NO power utilities there.
>> Am interested in 'weather/environ' info and
>> maybe still frames from a low-rez THERMAL
>> camera. A 'live mic' might be interesting
>> even at rather low bandwidth.
> 
> For 1 km and a small solar/battery box, split the requirements first.
> Weather/environment telemetry is easy. Still frames are possible but slow. A
> live microphone is the thing most likely to push you out of the cheap low-power
> radio class.

   Agree. Periodic "telemetry" can be made very compact, even if
   you just send in ASCII. Less if binary, and there are a few
   cheapo ways to compress further.

   A low-rez video frame ... delivered ON DEMAND while holding
   any other stuff in a buffer if needed ... is do-able.

   Sound ... even at a low bit rate ... is a LOT of info.

   But it'd be cool.

> LoRa is a good fit for small packets: sensor readings, battery voltage, status,
> alarms, maybe a tiny thumbnail if you are patient. Do not expect it to behave
> like slow WiFi. Depending on spreading factor and legal duty cycle limits, a few
> kbit/s or less is a more realistic planning number than the headline figures.
> 
> The current cheap modules worth looking at are usually UART/SPI LoRa boards
> based on Semtech SX127x/SX126x parts. Ebyte E32/E22/E220 style UART modules are
> common if you want "serial cable replacement" behavior. Digi XBee/XBee-PRO
> 900HP, RFD900x, Microhard, FreeWave, etc. are more expensive but more finished
> products. Check the band and power limits for your country; 433/868/915 MHz
> modules are not interchangeable legally.
> 
> A few practical points:
> 
> * Use a real outdoor antenna, placed high and dry, before throwing power at the
> problem. Fresnel clearance matters even at only 1 km. * Use packets with
> sequence numbers and CRCs, not a bare byte stream, even if the radio sells
> itself as transparent serial. * If you need multi-drop, handle
> addressing/retries in your protocol. "Radio RS-485" boxes often only make a
> transparent half-duplex serial link; collision behavior can be ugly. * For
> battery life, let the remote node sleep and wake on a schedule. A Pi can work,
> but a small MCU plus radio will be much easier to power and hide. * Send
> summaries often and images rarely. A tiny thermal frame may be acceptable every
> few minutes; audio probably wants a different link.
> 
> If you have line of sight, a pair of small directional 2.4/5 GHz outdoor CPE
> units can do 1 km easily, but they draw watts continuously and are more visible.
> For disguised, low-duty-cycle environmental telemetry, LoRa or a sub-GHz serial
> modem is the more natural starting point.
> 

   "LoRa" may be the better way to go these days. I'm an old
   fart and thus more familiar with the old serial protocols,
   but that doesn't mean LoRa is inherently evil.

   Now, can I do LoRa with like a higher-end Ard ? PIs are
   nice but they're power hogs, can't 'sleep'. Haven't
   looked into the BeagleBones recently ... DO have a BBB
   in The Heap somewhere though.

   Anyway, property#2 is not physically contiguous with
   property#1 ... so I can't run wires without getting
   sued, or arrested.

[toc] | [prev] | [next] | [standalone]

#87172

On Tue, 26 May 2026 23:04:18 -0400, c186282 wrote:

>    Now, can I do LoRa with like a higher-end Ard ? PIs are nice but
>    they're power hogs, can't 'sleep'. Haven't looked into the
>    BeagleBones recently ... DO have a BBB in The Heap somewhere though.

https://store.arduino.cc/products/arduino-mkr-wan-1310

[toc] | [prev] | [next] | [standalone]

#87176

On 5/26/26 23:21, rbowman wrote:
> On Tue, 26 May 2026 23:04:18 -0400, c186282 wrote:
> 
>>     Now, can I do LoRa with like a higher-end Ard ? PIs are nice but
>>     they're power hogs, can't 'sleep'. Haven't looked into the
>>     BeagleBones recently ... DO have a BBB in The Heap somewhere though.
> 
> https://store.arduino.cc/products/arduino-mkr-wan-1310


   Ah HA !!!

   Thanks !

   MicroCONTROLLERS generally offer detailed power
   state control - microPROCESSORS, not.

   Have considerable experience using Ards, and the
   low-power library, to collect field data with
   only a very small solar panel.

   The Mega-2560 is the way to go - lots more mem
   and general capability over the Uno. May have
   to lightly tweak a few libs to compensate for
   I/O pin placement.

   Hideous environment - someone still sells the
   'RuggedDurino' 2560 clone - all pins are
   protected against voltage surges.

[toc] | [prev] | [next] | [standalone]

#87181

On Wed, 27 May 2026 00:18:23 -0400, c186282 wrote:

>    Have considerable experience using Ards, and the low-power library,
>    to collect field data with only a very small solar panel.

Adafruit has a couple of similar products with the SAM D21 Cortex-M

https://www.adafruit.com/product/3178

The old Unos were great for their day but they have their limits even the 
R4. The nice thing about the Cortex-M designs is most of them have enough 
storage that you can use MicroPython, CircuitPython, or the traditional 
Arduino libraries. Even the Picos with the Arduino Core and Phil 
Hightower's RP2040 library can look like an Arduino.

SparkFun has one too.

https://www.sparkfun.com/sparkfun-lora-thing-plus-explorable.html

SparkFun and AdaFruit are in a pissing contest and I tend to favor 
Adafruit. They have a lot of available resources and have done quite a bit 
to promote MCUs. 

There's a MPLAD extension for VS Code that replaces the older Microchip 
IDE if you're a real glutton. Like the C SDK for the Picos I think that's 
a real break glass in case of emergency deal. I'm quite happy with the 
uniformity of the Pythons or Arduino across the various MCUs I want to 
play with. 

[toc] | [prev] | [next] | [standalone]

#87196

On 27/05/2026 04:21, rbowman wrote:
> On Tue, 26 May 2026 23:04:18 -0400, c186282 wrote:
> 
>>     Now, can I do LoRa with like a higher-end Ard ? PIs are nice but
>>     they're power hogs, can't 'sleep'. Haven't looked into the
>>     BeagleBones recently ... DO have a BBB in The Heap somewhere though.
> 
> https://store.arduino.cc/products/arduino-mkr-wan-1310
> 
Pico Pis can me made to sleep with the addition of a TPL5110 nanotimer.

I am looking at over a year battery life on 3xAA  lithium primaries.


-- 
“The urge to save humanity is almost always only a false face for the 
urge to rule it.”
– H. L. Mencken

[toc] | [prev] | [next] | [standalone]

#87319 — WiFi range ...

On 2026-05-25 22:27, c186282 wrote:
> On 5/25/26 13:56, rbowman wrote:
>> On Mon, 25 May 2026 01:46:15 -0400, c186282 wrote:
>>
>>> Anyway ... NOW I may have a use for RF "modems",
>>> a link to a kind of far-off property site. Do not need video, MAYbe the
>>> odd still image. The low speed isn't relevant, just the range. WiFi
>>> isn't gonna cut it at all.
>>
>> How far off? 
> 
> 
>    About 1000 meters.

My company's high reliability radios use essentially 20 year old WiFi 
technology, and we have customers than routinely run them over distances 
of 5-10 km. It is not the technology per se that limits distance; it is 
the cheap off-the-shelf gear that is crap.

We use an 802.11b modulator, in 900 MHz or 2400 MHz bands, at speeds 
from 100 kbps effective data transfer rates to 8Mbps. We do NOT us the 
802.11 MAC layers, because we WANT to be guaranteed to be incompatible 
with WiFi. We pair this modem chipset with a very good front-end (low 
noise receive amplifier and very clean, very linear power amplifier. And 
then we put good directional antennas on them. We have one customer who 
runs a 100km radio link with a 27dBi antenna at 500mW (27 dBm) power 
output. For years.

I bet many commercial WiFi radios if paired with a decent front-end and 
a good antenna, even an 18dBi panel antenna, would give you a reliable 
link over 5km/3miles, so long as you have a clean line of sight. Its not 
really that hard.

-- 
Lars Poulsen - an old geek in Santa Barbara, California

[toc] | [prev] | [next] | [standalone]

#87320 — Re: WiFi range ...

>On Sun, 31 May 2026 21:19:08 -0700, Lars Poulsen <lars@beagle-ears.com> wrote:
>On 2026-05-25 22:27, c186282 wrote:
>> On 5/25/26 13:56, rbowman wrote:
>>> On Mon, 25 May 2026 01:46:15 -0400, c186282 wrote:
>>>
>>>> Anyway ... NOW I may have a use for RF "modems",
>>>> a link to a kind of far-off property site. Do not need video, MAYbe the
>>>> odd still image. The low speed isn't relevant, just the range. WiFi
>>>> isn't gonna cut it at all.
>>>
>>> How far off?
>>
>>
>>    About 1000 meters.
>
>My company's high reliability radios use essentially 20 year old WiFi
>technology, and we have customers than routinely run them over distances
>of 5-10 km. It is not the technology per se that limits distance; it is
>the cheap off-the-shelf gear that is crap.
>
>We use an 802.11b modulator, in 900 MHz or 2400 MHz bands, at speeds
>from 100 kbps effective data transfer rates to 8Mbps. We do NOT us the
>802.11 MAC layers, because we WANT to be guaranteed to be incompatible
>with WiFi. We pair this modem chipset with a very good front-end (low
>noise receive amplifier and very clean, very linear power amplifier. And
>then we put good directional antennas on them. We have one customer who
>runs a 100km radio link with a 27dBi antenna at 500mW (27 dBm) power
>output. For years.
>
>I bet many commercial WiFi radios if paired with a decent front-end and
>a good antenna, even an 18dBi panel antenna, would give you a reliable
>link over 5km/3miles, so long as you have a clean line of sight. Its not
>really that hard.

For a 1 km link, I would start with the boring purpose-built gear rather than
trying to make ordinary indoor WiFi stretch that far.  A pair of outdoor
CPE/bridge units with directional panels will usually be far less fiddly than
routers plus random antennas.

The practical checks are mostly RF geometry, not Linux-specific:

* clear line of sight, including some Fresnel-zone clearance;
* fixed mounting and careful alignment at both ends;
* outdoor-rated cable/enclosures and proper grounding/lightning protection;
* enough margin for rain, leaves, and future clutter in the path.

If only low-rate data and the occasional still image are needed, backing the
link down to a slower modulation/rate can buy a lot of reliability.  Once the
bridge is up, Linux just sees an Ethernet path, which is the nicest possible
outcome.

-- 
TheLastSysop <thelastsysop@dev.null>
"I survived the great rm -rf / rehearsal and all I got was this .signature."

[toc] | [prev] | [next] | [standalone]

#87325

c186282 wrote:

> rbowman wrote:
> >> c186282 wrote:
>>
>>> I may have a use for RF "modems",
>>> a link to a kind of far-off property site. Do not need video, MAYbe the
>>> odd still image. The low speed isn't relevant, just the range. WiFi
>>> isn't gonna cut it at all.
>>
>> How far off? 
> 
>    About 1000 meters.
>    NO power utilities there.
Maybe look at specialised WiFi?  802.11ah HaLow

[toc] | [prev] | [next] | [standalone]

#87326

>On Mon, 1 Jun 2026 08:10:41 +0100, Andy Burns <usenet@andyburns.uk> wrote:
>c186282 wrote:
>
>> rbowman wrote:
>> >> c186282 wrote:
>>>
>>>> I may have a use for RF "modems",
>>>> a link to a kind of far-off property site. Do not need video, MAYbe the
>>>> odd still image. The low speed isn't relevant, just the range. WiFi
>>>> isn't gonna cut it at all.
>>>
>>> How far off?
>>
>>    About 1000 meters.
>>    NO power utilities there.
>Maybe look at specialised WiFi?  802.11ah HaLow

For 1000 m and no mains, I would split the question into "link budget" and
"power budget" before buying anything.

If there is clean line of sight, a pair of outdoor point-to-point CPE radios
(Ubiquiti/MikroTik/TP-Link CPE type kit) is usually simpler than old-style RF
modems and can be throttled down to very little traffic.  The catch is that the
remote end still needs a small panel, battery, charge controller and a
weatherproof enclosure, so idle power matters more than peak throughput.

If it is not line-of-sight, or if the remote end is just sending occasional
sensor data/status and maybe not images, then LoRa/LoRaWAN-type gear is worth a
look.  It is much easier to run from a small battery/solar setup, but it is not
a substitute for an IP-ish link with images.

HaLow sits between those worlds.  I would check local regulatory limits and the
actual antenna options before assuming it beats a boring directional 2.4/5 GHz
bridge at only 1 km.

-- 
TheLastSysop <thelastsysop@dev.null>
"I survived the great rm -rf / rehearsal and all I got was this .signature."

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.misc

csiph-web