Wireless Development with PICs

Started by Ashwin Shankar August 16, 2003
Hi,
I'm curious to find out if anyone out there has/wants to use PICs as
controllers for RF communications. I'm currently working on a standalone
embedded board with integrated wireless capability. The current
processor revision is using a RT-Java Embedded. The package will come
standard with a mesh networking software library in Java
(Microchip-compatible C code currently in development) to implement p2p
communication out-of-the-box.

I'm trying to figure out if there is a need for developers to use PICS
as part of wireless systems. Before anyone mentions the rfPICS that
MicroChip has, the wireless chipset I'm using is in the 2.4 GHz
frequency, as well as supporting upto 1.5Mbps data streams. So it's not
really the same targeted userbase.

I'd appreciate a developer's point-of-view critique of the concept and
what I should be focusing on.

Thanks,
Ashwin




Maybe exactly what I am looking for, but there are too many other
specific requirements. I was considering doing it myself with Tiny OS.
Do you have any preliminary specs?

Chad

--- Ashwin Shankar <> wrote:
> Hi,
> I'm curious to find out if anyone out there has/wants to use PICs as
> controllers for RF communications. I'm currently working on a
> standalone
> embedded board with integrated wireless capability. The current
> processor revision is using a RT-Java Embedded. The package will come
> standard with a mesh networking software library in Java
> (Microchip-compatible C code currently in development) to implement
> p2p
> communication out-of-the-box.
>
> I'm trying to figure out if there is a need for developers to use
> PICS
> as part of wireless systems. Before anyone mentions the rfPICS that
> MicroChip has, the wireless chipset I'm using is in the 2.4 GHz
> frequency, as well as supporting upto 1.5Mbps data streams. So it's
> not
> really the same targeted userbase.
>
> I'd appreciate a developer's point-of-view critique of the concept
> and
> what I should be focusing on.
>
> Thanks,
> Ashwin
>


__________________________________




Chad,
Currently, we're working on processors by aJile systems. These
processors execute native Java byte-code, and can support upto 2 JVMs
natively. Clock speed would be about 100Mhz. Currently, we're trying
to determine if there is enough processing power for the host
controller to handle the baseband functionality as well as local
tasks. If not, we'll have to use a PIC as the baseband controller. We
are estimating 8MB of memory onboard.

The RFIC we're using handles upto 1.5Mbps data streams, range is
dependent on antenna design but can aggregate it out to about 100m
and
uses DSSS.

We're about 4 months away from production. In the mean time, as I
mentioned, I'm trying to figure out how to make this appealing to
other developers. If you could let me know what kind of requirements
you're looking for, it might help in pinpointing some key selling
points.

Thanks,
Ashwin
--- In , Chad Russel <chadrussel@y...> wrote:
> Maybe exactly what I am looking for, but there are too many other
> specific requirements. I was considering doing it myself with Tiny
OS.
> Do you have any preliminary specs?
>
> Chad
>
> --- Ashwin Shankar <ashwin@i...> wrote:
> > Hi,
> > I'm curious to find out if anyone out there has/wants to use PICs
as
> > controllers for RF communications. I'm currently working on a
> > standalone
> > embedded board with integrated wireless capability. The current
> > processor revision is using a RT-Java Embedded. The package will
come
> > standard with a mesh networking software library in Java
> > (Microchip-compatible C code currently in development) to
implement
> > p2p
> > communication out-of-the-box.
> >
> > I'm trying to figure out if there is a need for developers to use
> > PICS
> > as part of wireless systems. Before anyone mentions the rfPICS
that
> > MicroChip has, the wireless chipset I'm using is in the 2.4 GHz
> > frequency, as well as supporting upto 1.5Mbps data streams. So
it's
> > not
> > really the same targeted userbase.
> >
> > I'd appreciate a developer's point-of-view critique of the concept
> > and
> > what I should be focusing on.
> >
> > Thanks,
> > Ashwin
> > __________________________________
>




Ashwin,

I did not notice before, but the data rate is much higher than I need,
therefore cost, and power will also be too high. :-( The range is
correct about 100 meters, but the average power must be fairly low
,i.e. < 0.02 watts, and I need to network up to ~500 devices. I will
need to live with what ever speed can be done with these parameters.

Regards,
Chad

--- ashwiins <> wrote:
> Chad,
> Currently, we're working on processors by aJile systems. These
> processors execute native Java byte-code, and can support upto 2 JVMs
>
> natively. Clock speed would be about 100Mhz. Currently, we're trying
> to determine if there is enough processing power for the host
> controller to handle the baseband functionality as well as local
> tasks. If not, we'll have to use a PIC as the baseband controller. We
>
> are estimating 8MB of memory onboard.
>
> The RFIC we're using handles upto 1.5Mbps data streams, range is
> dependent on antenna design but can aggregate it out to about 100m
> and
> uses DSSS.
>
> We're about 4 months away from production. In the mean time, as I
> mentioned, I'm trying to figure out how to make this appealing to
> other developers. If you could let me know what kind of requirements
> you're looking for, it might help in pinpointing some key selling
> points.
>
> Thanks,
> Ashwin
> --- In , Chad Russel <chadrussel@y...> wrote:
> > Maybe exactly what I am looking for, but there are too many other
> > specific requirements. I was considering doing it myself with Tiny
>
> OS.
> > Do you have any preliminary specs?
> >
> > Chad
> >
> > --- Ashwin Shankar <ashwin@i...> wrote:
> > > Hi,
> > > I'm curious to find out if anyone out there has/wants to use PICs
>
****Snip**** __________________________________




Chad,
I would appreciate it if you would keep me informed of which radio
you
selected. I agree with data rate issues as well. The particular
reason
why we selected this one was to ensure that we can provide solutions
in the 900Mhz/2.4Ghz/5.8Ghz ranges as well as ensure we can control
transmit power based on signal strentgh received.

Another reason we went with this chip was that the greater bandwith
implies shorter burst times. So you're at peak output power for a
shorter period of time. And there's is less network congestion due to
shorter burst times. Also helps in handling more traffic if need be,
like a second backbone.

Are you planning to implement a TDMA model or a TDD model for
communications timing?

Cost-wise I'm not really sure if you're going to get something very
cheap unless you opt for a rfPIC. Most radio modules out there on the
market are in the region from $150-$300, course they have a bunch of
features like node id, CRC, and encryption.

I'd be interested in learning more about how you achieve the 100m
range yet keep the power levels so low.

Regards,
Ashwin
--- In , Chad Russel <chadrussel@y...> wrote:
> Ashwin,
>
> I did not notice before, but the data rate is much higher than I
need,
> therefore cost, and power will also be too high. :-( The range is
> correct about 100 meters, but the average power must be fairly low
> ,i.e. < 0.02 watts, and I need to network up to ~500 devices. I
will
> need to live with what ever speed can be done with these parameters.
>
> Regards,
> Chad





Ashwin,

I have been hoping for something off the shelf, but considering the
requirements, I assumed I would have to come up with my own design for
the rf and modem. The ~500 units is max for one installation, with
thousands of installations out there, so the development cost can be
recovered and the price kept low.

I only know some of the basics of networking, but am going to have to
come up to speed rapidly. I saw a system from Berkley called Tiny OS
which sounds perfect for my application. I just have to learn all
about it yet and am open to any other suggestions.

I was going to keep the average power down by using high power rf in a
short burst and an 'energy pump'.

Chad --- ashwiins <> wrote:
> Chad,
> I would appreciate it if you would keep me informed of which radio
> you
> selected. I agree with data rate issues as well. The particular
> reason
> why we selected this one was to ensure that we can provide solutions
> in the 900Mhz/2.4Ghz/5.8Ghz ranges as well as ensure we can control
> transmit power based on signal strentgh received.
>
> Another reason we went with this chip was that the greater bandwith
> implies shorter burst times. So you're at peak output power for a
> shorter period of time. And there's is less network congestion due to
>
> shorter burst times. Also helps in handling more traffic if need be,
> like a second backbone.
>
> Are you planning to implement a TDMA model or a TDD model for
> communications timing?
>
> Cost-wise I'm not really sure if you're going to get something very
> cheap unless you opt for a rfPIC. Most radio modules out there on the
>
> market are in the region from $150-$300, course they have a bunch of
> features like node id, CRC, and encryption.
>
> I'd be interested in learning more about how you achieve the 100m
> range yet keep the power levels so low.
>
> Regards,
> Ashwin
> --- In , Chad Russel <chadrussel@y...> wrote:
> > Ashwin,
> >
> > I did not notice before, but the data rate is much higher than I
> need,
> > therefore cost, and power will also be too high. :-( The range is
> > correct about 100 meters, but the average power must be fairly low
> > ,i.e. < 0.02 watts, and I need to network up to ~500 devices. I
> will
> > need to live with what ever speed can be done with these
> parameters.
> >
> > Regards,
> > Chad


__________________________________




Chad,
What exactly is an energy pump? I'm curious to understand what exactly
the benefits of using this method are.

If you're looking to deploy a sensor network, then something like
TinyOS might fit the bill although I don't think hardware is available
as yet.

If you're looking for a bit more of processing power, I'm not sure
they will help. This was a similar problem we faces and had to build
ourselves to solve.

-A

--- In , Chad Russel <chadrussel@y...> wrote:
> Ashwin,
>
> I have been hoping for something off the shelf, but considering the
> requirements, I assumed I would have to come up with my own design
for
> the rf and modem. The ~500 units is max for one installation, with
> thousands of installations out there, so the development cost can be
> recovered and the price kept low.
>
> I only know some of the basics of networking, but am going to have
to
> come up to speed rapidly. I saw a system from Berkley called Tiny OS
> which sounds perfect for my application. I just have to learn all
> about it yet and am open to any other suggestions.
>
> I was going to keep the average power down by using high power rf in
a
> short burst and an 'energy pump'.
>
> Chad



Energy pump is sort of one of my inventions. Like a charge pump with a
gas tank. These units steal power from other devices and store the
energy.

Except for what may be required for the networking, the PIC has enough
horsepower for my application.

Chad

--- ashwiins <> wrote:
> Chad,
> What exactly is an energy pump? I'm curious to understand what
> exactly
> the benefits of using this method are.
>
> If you're looking to deploy a sensor network, then something like
> TinyOS might fit the bill although I don't think hardware is
> available
> as yet.
>
> If you're looking for a bit more of processing power, I'm not sure
> they will help. This was a similar problem we faces and had to build
> ourselves to solve.
>
> -A
>
> --- In , Chad Russel <chadrussel@y...> wrote:
> > Ashwin,
> >
> > I have been hoping for something off the shelf, but considering the
> > requirements, I assumed I would have to come up with my own design
> for
> > the rf and modem. The ~500 units is max for one installation, with
> > thousands of installations out there, so the development cost can
> be
> > recovered and the price kept low.
> >
> > I only know some of the basics of networking, but am going to have
> to
> > come up to speed rapidly. I saw a system from Berkley called Tiny
> OS
> > which sounds perfect for my application. I just have to learn all
> > about it yet and am open to any other suggestions.
> >
> > I was going to keep the average power down by using high power rf
> in
> a
> > short burst and an 'energy pump'.
> >
> > Chad


__________________________________




Are you talking about a system of devices that "borrow" emitted power from whatever happens to be broadcasted in the area?  Like a solar cell that charges it's battery from the sun then run the <whatever> off the battery? You may want to make sure this isn't a commercial application and if it is, make sure you can prove your not borrowing 60Hz power floating around the area waves.. as stupid as it is, that's illegal to use emitted power from power lines. I am not 100% up on that law, but I know it's a no-no to use radiated power off the big lines..
 
Also, have you don't any medium/long range radio modem work?
 
Charles
-----Original Message-----
From: Chad Russel [mailto:c...@yahoo.com]
Sent: Wednesday, August 20, 2003 3:15 PM
To: p...@yahoogroups.com
Subject: Re: [piclist] Re: Wireless Development with PICs

Energy pump is sort of one of my inventions.  Like a charge pump with a
gas tank.  These units steal power from other devices and store the
energy. 

Except for what may be required for the networking, the PIC has enough
horsepower for my application.

Chad

--- ashwiins <a...@imagenar.com> wrote:
> Chad,
> What exactly is an energy pump? I'm curious to understand what
> exactly
> the benefits of using this method are.
>
> If you're looking to deploy a sensor network, then something like
> TinyOS might fit the bill although I don't think hardware is
> available
> as yet.
>
> If you're looking for a bit more of processing power, I'm not sure
> they will help. This was a similar problem we faces and had to build
> ourselves to solve.
>
> -A
>
> --- In p...@yahoogroups.com, Chad Russel <chadrussel@y...> wrote:
> > Ashwin,
> >
> > I have been hoping for something off the shelf, but considering the
> > requirements, I assumed I would have to come up with my own design
> for
> > the rf and modem.  The ~500 units is max for one installation, with
> > thousands of installations out there, so the development cost can
> be
> > recovered and the price kept low.
> >
> > I only know some of the basics of networking, but am going to have
> to
> > come up to speed rapidly. I saw a system from Berkley called Tiny
> OS
> > which sounds perfect for my application.  I just have to learn all
> > about it yet and am open to any other suggestions.
> >
> > I was going to keep the average power down by using high power rf
> in
> a
> > short burst and an 'energy pump'.
> >
> > Chad__________________________________to unsubscribe, go to http://www.yahoogroups.com and follow the instructions

">Yahoo! Terms of Service.



Now that is an interesting idea. :-D Where do "they" draw the line? A
friend uses a circuit that amplifies and rectifies the 60 hz pick up
from the human body to activate a device. Is that stealing?

The power I am stealing is from control circuis, hopefully without the
control circuits knowing it. :)

Medium long range? Does teletype FSK count? :-p Actually I have been
thinking about the HF possiblility, because line of sight may not
always be possible.

Chad

--- Charles Douvier <> wrote:
> Are you talking about a system of devices that "borrow" emitted power
> from
> whatever happens to be broadcasted in the area? Like a solar cell
> that
> charges it's battery from the sun then run the <whatever> off the
> battery?
> You may want to make sure this isn't a commercial application and if
> it is,
> make sure you can prove your not borrowing 60Hz power floating around
> the
> area waves.. as stupid as it is, that's illegal to use emitted power
> from
> power lines. I am not 100% up on that law, but I know it's a no-no to
> use
> radiated power off the big lines..
>
> Also, have you don't any medium/long range radio modem work?
>
> Charles
> -----Original Message-----
> From: Chad Russel [mailto:]
> Sent: Wednesday, August 20, 2003 3:15 PM
> To:
> Subject: Re: [piclist] Re: Wireless Development with PICs > Energy pump is sort of one of my inventions. Like a charge pump
> with a
> gas tank. These units steal power from other devices and store the
> energy.
>
****SNIP****

__________________________________