best wireless architecture for microcontroller application

I have been searching the internet for advice on what sort of wireless
to use for a microcontroller application, but have only been finding
info on wireless with respect to PCs.

Does anyone have any advice on what to use?

I will have about 100 modules that will control LED arrays. I really
want to find something where I can create a network in which they all
communicate with each other rather a master/slave scenario.

Thanks in advance for any responses.

Cheers,
DHD

On 26 Jan 2007 14:24:11 -0800, "Dr. Honeydew" <(E-Mail Removed)>
wrote in <(E-Mail Removed) .com>:

>I have been searching the internet for advice on what sort of wireless
>to use for a microcontroller application, but have only been finding
>info on wireless with respect to PCs.
>
>Does anyone have any advice on what to use?
>
>I will have about 100 modules that will control LED arrays. I really
>want to find something where I can create a network in which they all
>communicate with each other rather a master/slave scenario.
>
>Thanks in advance for any responses.

Lots more info would be helpful; e.g.,
Maximum range needed?
Transmission speed needed?
Latency issues?
Error checking needed?
Duty cycle (% of time on the air)?
Power available?
Physical environment?
Sources of radio interference?

--
Best regards, FAQ for Wireless Internet: <http://Wireless.wikia.com>
John Navas FAQ for Wi-Fi: <http://wireless.wikia.com/wiki/Wi-Fi>
Wi-Fi How To: <http://wireless.wikia.com/wiki/Wi-Fi_HowTo>
Fixes to Wi-Fi Problems: <http://wireless.wikia.com/wiki/Wi-Fi_Fixes>

Zigbee from Jennic looks interesting, although it is based on an open
standards, using their technology is proprietary I think, like Echelon
Lonworks (who incidently promise their own wireless system in their
Pyxsos line) the upshot is that you have to pay a lot.

In terms of using a standard technology, Bluetooth might be an option,
as long as the modules are not physically too far apart.

On 26 Jan, 22:24, "Dr. Honeydew" <wjgoodr...@gmail.com> wrote:
> I have been searching the internet for advice on what sort of wireless
> to use for a microcontroller application, but have only been finding
> info on wireless with respect to PCs.
>
> Does anyone have any advice on what to use?
>
> I will have about 100 modules that will control LED arrays. I really
> want to find something where I can create a network in which they all
> communicate with each other rather a master/slave scenario.
>
> Thanks in advance for any responses.
>
> Cheers,
> DHD

Thanks John...
Yea, good call...
Here is some more info:
> Maximum range needed?
The units will be laid out in an array about 120ft in diameter, so no
one unit will be more than that distance form another.
> Transmission speed needed?
Hmm, didn't think too much about this one. The project is a sort of
interactive light show where people will trigger lights in a unit
nearest them and this will set off a kind of domino effect of other
units lighting up. Does that make sense?
> Latency issues?
Sorry, I'm kind of new at this stuff. What do you mean?
> Error checking needed?
Presumably that would be important. Again, I'm not sure how to
factor that in, I'm sorry.
> Duty cycle (% of time on the air)?
Hopefully the above description helps answer some of these questions,
'cause I'm at a loss here! ha.
> Power available?
Battery powered, probably 5 volts. There will be enough to power 60
LEDs per unit, so presumably plenty for components.
> Physical environment?
The desert. Literally. Modules will be protected with enclosures,
but temps will be rather high and enclosures will be in direct
sunlight.
> Sources of radio interference?
I wouldn't imagine much. I suppose if I use 802.11 I am at higher
risk for that?

Thanks so much.

Cheers,
Bill Goodrich



On Jan 26, 4:01 pm, John Navas <spamfilt...@navasgroup.com> wrote:
> On 26 Jan 2007 14:24:11 -0800, "Dr. Honeydew" <wjgoodr...@gmail.com>
> wrote in <1169850250.058680.219...@v33g2000cwv.googlegroups .com>:
>
> >I have been searching the internet for advice on what sort of wireless
> >to use for a microcontroller application, but have only been finding
> >info on wireless with respect to PCs.
>
> >Does anyone have any advice on what to use?
>
> >I will have about 100 modules that will control LED arrays. I really
> >want to find something where I can create a network in which they all
> >communicate with each other rather a master/slave scenario.
>
> >Thanks in advance for any responses.Lots more info would be helpful; e.g.,
> Maximum range needed?
> Transmission speed needed?
> Latency issues?
> Error checking needed?
> Duty cycle (% of time on the air)?
> Power available?
> Physical environment?
> Sources of radio interference?
>
> --
> Best regards, FAQ for Wireless Internet: <http://Wireless.wikia.com>
> John Navas FAQ for Wi-Fi: <http://wireless.wikia.com/wiki/Wi-Fi>
> Wi-Fi How To: <http://wireless.wikia.com/wiki/Wi-Fi_HowTo>
> Fixes to Wi-Fi Problems: <http://wireless.wikia.com/wiki/Wi-Fi_Fixes>

Thanks Lee.

Yea, I've had some experience with maxstream's OEM zigbee modules. I
had the same thought for this application. Haven't looked at Jennic's
yet.

Thanks.

Cheers,
DHD

On Jan 26, 4:10 pm, "lee" <leehan...@gmail.com> wrote:
> Zigbee from Jennic looks interesting, although it is based on an open
> standards, using their technology is proprietary I think, like Echelon
> Lonworks (who incidently promise their own wireless system in their
> Pyxsos line) the upshot is that you have to pay a lot.
>
> In terms of using a standard technology, Bluetooth might be an option,
> as long as the modules are not physically too far apart.
>
> On 26 Jan, 22:24, "Dr. Honeydew" <wjgoodr...@gmail.com> wrote:
>
> > I have been searching the internet for advice on what sort of wireless
> > to use for a microcontroller application, but have only been finding
> > info on wireless with respect to PCs.
>
> > Does anyone have any advice on what to use?
>
> > I will have about 100 modules that will control LED arrays. I really
> > want to find something where I can create a network in which they all
> > communicate with each other rather a master/slave scenario.
>
> > Thanks in advance for any responses.
>
> > Cheers,
> > DHD

On 26 Jan 2007 15:41:21 -0800, "Dr. Honeydew" <(E-Mail Removed)>
wrote in <(E-Mail Removed) .com>:

>> Maximum range needed?
> The units will be laid out in an array about 120ft in diameter, so no
>one unit will be more than that distance form another.

Bluetooth Class 1 (100 m range) should suffice. Likewise ZigBee.

>> Transmission speed needed?
> Hmm, didn't think too much about this one. The project is a sort of
>interactive light show where people will trigger lights in a unit
>nearest them and this will set off a kind of domino effect of other
>units lighting up. Does that make sense?

Sure, and although you ducked my question, sounds like speed isn't much
of an issue.

>> Latency issues?
> Sorry, I'm kind of new at this stuff. What do you mean?

How much typical and worst-case transmission delay can the application
tolerate? 1 ms? 10 ms? 100 ms? 1,000 ms? etc.

>> Error checking needed?
> Presumably that would be important. Again, I'm not sure how to
>factor that in, I'm sorry.

Will your application assume responsibility for retransmitting lost
packets, or must the radio system do that?

>> Duty cycle (% of time on the air)?
> Hopefully the above description helps answer some of these questions,
>'cause I'm at a loss here! ha.

Sounds pretty low, so 100 nodes may be practical. 0.1% is a typical
ZigBee duty cycle.

>> Physical environment?
> The desert. Literally. Modules will be protected with enclosures,
>but temps will be rather high and enclosures will be in direct
>sunlight.

Enclosures need to be transparent to radio, will need venting, and
you'll probably want sunshades to avoid overheating.

Question I forgot: What kind of security do you need? Protection
against hackers? Robust encryption? Nothing?

Anyway, a Bluetooth PAN (Personal Area Network) or piconet could be the
way to go if only 8 devices will be "active" at any one time; all other
devices would have to be in "parked" mode. (Latency could be an issue
if more than 8 devices try to become active at any one time.) Bluetooth
advantages are cost (low), standardization, and wide availability of
parts.

Another possible way to go is ZigBee, which has a simpler implementation
than Bluetooth, more flexibility (up to 255 devices on the network), and
even lower cost. That's where I'd personally start.

--
Best regards, FAQ for Wireless Internet: <http://Wireless.wikia.com>
John Navas FAQ for Wi-Fi: <http://wireless.wikia.com/wiki/Wi-Fi>
Wi-Fi How To: <http://wireless.wikia.com/wiki/Wi-Fi_HowTo>
Fixes to Wi-Fi Problems: <http://wireless.wikia.com/wiki/Wi-Fi_Fixes>