Prev: Low Rds(on) tempco
Next: Termination of AC coupled 100MHz clok to generate low jitter 50MHz clock?
From: Tim Watts on 25 Jan 2010 05:22
Hi,

New to the group... But did google around the archives first, so please
excuse if this has been done to death - I did try(!)

I play with AVRs (Tiny/Mega) as a hobby (linux sysadmin, perl/C
programmer by trade).

I fancy having a play with some radio datalinks (with a view to home
automation) - but there seem to be a million modules and a fair number of
frequencies and standards. I'm just after some pointers based off some
simple requirements:

Important:

a) Range - 20m through masonry (11" brick wall and 4" wall sort of thing,
no rebar), 40m free air (could accept 10/20m)

b) Usable data rate - 10's to 100-ish kbit/sec

c) Cheap - 20 pounds sterling give or take, say 40 US dollars for a small
easy to mount module (wire pads/header pins - no funky surface mount
modules, my soldering isn't that advanced)

Pie in the sky wish list

d) Ideally simple framing built in - ie I clock a bunch of data in and
hit "send" so to speak, and it transmits. Receive buffers frame and
wiggles an interrupt. "Frame" could mean 8 bit word, or entire long
packet (100's words).

But ultimately I *could* live with wiggling some pins on a simple RF
module where the 2 pins send different signals over the carrier to
indicate 1 and 0.

d) would be nice as it simplifies programming by miles, but not if it
impacts on c)

Not too bothered whether 433MHz, 868MHz, or 2.4GHz (this is the UK BTW),
though I suspect 868MHz would be slightly better being less crowded.

Zigbee looked interesting but I don't really need a whole protocol stack
(I can do that) and it's not cheap.


Any pointers as to any technogolgies or modules that would be worth
looking at would be most gratefully received. I've looked - wood and
trees syndrome - don't know how to filter the choices down...

Ta muchly :)

Tim

--
Tim Watts

Managers, politicians and environmentalists: Nature's carbon buffer.
From: John Walliker on 25 Jan 2010 06:17
On 25 Jan, 10:22, Tim Watts <t...(a)dionic.net> wrote:
I fancy having a play with some radio datalinks (with a view to home
automation) - but there seem to be a million modules and a fair number
of
frequencies and standards. I'm just after some pointers based off some
simple requirements:

How bothered are you about power consumption? Can you use mains
signalling or just lay cable?

If you decide to go for basic modules without complex modulation
methods and software stacks, then there are three main types to
consider:

1) On-off (amplitude) keying

2) Frequency shift keying - wideband

3) Frequency shift keying - narrowband

The range and cost increase in the order 1 -> 3.

For the same transmitter power, receiver sensitivity and receiver
bandwidth lower frequencies will generally give longer range than
higher, mostly because the lower frequency receiver antenna (of
equivalent directivity) will capture more of the incident power .
However, higher frequencies will make it easier to use directional
antennae as they will be smaller. High frequencies diffract round
corners better, while low frequencies go through walls better.

I have successfully used the GT1 25mW transmitter and GR1 superhet
receiver modules available from several UK suppliers over a range of
about 2km in open air with a well-matched dipole antenna at each end.
I picked a frequency of 434.225 MHz because it is relatively quiet in
my area.

Another approach to getting long range is to take advantage of the UK
0.5W band at around 868 MHz, but the modules are more expensive than
the lower power ones.

There are some very easy-to-use single-chip Manchester coder/decoders
available for use with FSK modules - RF600 and RF800.

Have a look at Radiometrix, Low Power Radio Solutions, MK Consultants
web sites. And of course the Ofcom website for frequency allocations
and other rules such as permitted power levels and duty cycle limits.
In particular, look for frequencies which are allowed in the UK but
not in the rest of Europe. These will generally be much quieter than
those which can be used in any European country.

If you do decide to go for Zigbee, Farnell sell reasonably priced
Zigbee modules, but the development tools are quite expensive.

John
From: Tim Watts on 25 Jan 2010 07:52
On Mon, 25 Jan 2010 03:17:50 -0800, John Walliker <jrwalliker(a)gmail.com>
wibbled:

> On 25 Jan, 10:22, Tim Watts <t...(a)dionic.net> wrote: I fancy having a
>> play with some radio datalinks (with a view to home automation) - but
>> there seem to be a million modules and a fair number of
>> frequencies and standards. I'm just after some pointers based off some
>> simple requirements:
>
> How bothered are you about power consumption? Can you use mains
> signalling or just lay cable?

Hi John,

Thanks for the reply.

Cabling: Well, I will have a wired DC SELV supply to avoid falling foul
of the IEE Wiring Regulations if I want to deploy any of these in
"Special Locations" like bathrooms. So current consumption will not an
issue there, but obviously, this rules out mains signalling.

I had considered dropping cables round and using a bastardised CAN bus or
RS485 drops or something, but radio would be cool for a number of
reasons, including the education factor. If it's cheap enough it may be
cost comparable with a cabled method (bearing in mind cables,
terminations, junction boxes etc). Also means I can extend the system to
outside use.

> If you decide to go for basic modules without complex modulation methods
> and software stacks, then there are three main types to consider:
>
> 1) On-off (amplitude) keying
>
> 2) Frequency shift keying - wideband
>
> 3) Frequency shift keying - narrowband
>
> The range and cost increase in the order 1 -> 3.

I've never been any good with analogue stuff, but I do understand those
terms.

> For the same transmitter power, receiver sensitivity and receiver
> bandwidth lower frequencies will generally give longer range than
> higher, mostly because the lower frequency receiver antenna (of
> equivalent directivity) will capture more of the incident power .
> However, higher frequencies will make it easier to use directional
> antennae as they will be smaller. High frequencies diffract round
> corners better, while low frequencies go through walls better.

OK.

> I have successfully used the GT1 25mW transmitter and GR1 superhet
> receiver modules available from several UK suppliers over a range of
> about 2km in open air with a well-matched dipole antenna at each end. I
> picked a frequency of 434.225 MHz because it is relatively quiet in my
> area.

That seems very impressive for the power (thinking how many milliwatts my
WIFI runs at)

> Another approach to getting long range is to take advantage of the UK
> 0.5W band at around 868 MHz, but the modules are more expensive than the
> lower power ones.

Ah - thanks - I didn't know powers like 0.5W were permissable.

> There are some very easy-to-use single-chip Manchester coder/decoders
> available for use with FSK modules - RF600 and RF800.

Very interesting - just scanned the datasheet.

> Have a look at Radiometrix, Low Power Radio Solutions, MK Consultants
> web sites.

Just for starters, I really like the look of this one:

http://www.lprs.co.uk/product_info.php?cPath=35&products_id=56

(easy-Radio 433-4MHz FM transceiver)

Probably about the level I'm looking for.

> And of course the Ofcom website for frequency allocations
> and other rules such as permitted power levels and duty cycle limits. In
> particular, look for frequencies which are allowed in the UK but not in
> the rest of Europe. These will generally be much quieter than those
> which can be used in any European country.

Good idea. Most manufacturers will be going for common frequencies in a
common market - makes sense to try for a quieter band.

> If you do decide to go for Zigbee, Farnell sell reasonably priced Zigbee
> modules, but the development tools are quite expensive.

I'll have a look there next.

Many thanks John, lots of excellent pointers :)

Cheers

Tim




--
Tim Watts

Managers, politicians and environmentalists: Nature's carbon buffer.
From: Tim Watts on 25 Jan 2010 07:58
On Mon, 25 Jan 2010 12:52:11 +0000, Tim Watts <tw(a)dionic.net> wibbled:


> Just for starters, I really like the look of this one:
>
> http://www.lprs.co.uk/product_info.php?cPath=35&products_id=56
>
> (easy-Radio 433-4MHz FM transceiver)

Incidently, please excuse my ignorance: what would be a sufficient
antenna for one of these? Bit of straight open ended wire, 1/4 wavelength
(so 17cm or so for 433MHZ)? Or is it better to go for 1/2 wavelength if
possible, or some sort of loop? Does it need to be precise for resonance
purposes?

--
Tim Watts

Managers, politicians and environmentalists: Nature's carbon buffer.
From: John Walliker on 25 Jan 2010 08:28
On 25 Jan, 12:58, Tim Watts <t...(a)dionic.net> wrote:
> On Mon, 25 Jan 2010 12:52:11 +0000, Tim Watts <t...(a)dionic.net> wibbled:
>
> > Just for starters, I really like the look of this one:
>
> >http://www.lprs.co.uk/product_info.php?cPath=35&products_id=56
>
> > (easy-Radio 433-4MHz FM transceiver)

Looks nice and easy to use, but less transmitter power than some of
the others.

>
> Incidently, please excuse my ignorance: what would be a sufficient
> antenna for one of these? Bit of straight open ended wire, 1/4 wavelength
> (so 17cm or so for 433MHZ)? Or is it better to go for 1/2 wavelength if
> possible, or some sort of loop? Does it need to be precise for resonance
> purposes?
>

The antennae on the page you posted a link to would be a good starting
point. Remember that they need a ground plane or counterpoise of at
least similar dimensions.

You could just use a 1/4-wave length of wire. Tuning is not terribly
critical for a 1/4-wave monopole, but you could monitor the RSSI
output of the receiver while adjusting the wire length by about +/-
20% in an uncluttered environment (such as outdoors). Loops need to
be tuned quite carefully. Helical antennae are shorter, but tend to
be less efficient and need more accurate tuning.

For the long range link I was experimenting with, I used home-made
resonant sleeve dipoles at each end, made by applying adhesive copper
tape to plastic conduit with the coax down the middle. The tuning was
done with the help of a vector network analyzer and the antenna was
about 3m above ground at one end and about 2m above ground at the
other. I'll check my GPS data later to see what the maximum range was.

As a general rule, unless you need to make a directional antenna, a
1/4-wave monopole or a properly balanced 1/2-wave dipole will give the
best and most consistent results compared with most other structures.
If you need smaller size there is always a compromise between
efficiency, size and bandwidth (the narrower the bandwidth the harder
it is to keep the antenna optimally tuned in a varying environment).

John
 |  Next  |  Last
Pages: 1 2 3
Prev: Low Rds(on) tempco
Next: Termination of AC coupled 100MHz clok to generate low jitter 50MHz clock?