AT90S2313 | AT90S8515 | ATMega | ATmega128 | ECL | FETS | IAR | Keyboard | LCD | STK50 | TMOS | UART

Hie i'm new to the AVR environment. I'm an engineering student doing my final year project, my project is on a wireless transmiter that will receive 2 analogue inputs, store the data in memory and then transmit the data by GSM to a computer that will then plot the data on a graph. i don't know where to start..........can somebody help. Which micro-controller is best suited for this job?

------------------------------------

______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.


(You need to be a member of avrclub -- send a blank email to avrclub-subscribe@yahoogroups.com )

On Jul 16, 2009, at 4:45 PM, vmtetwa wrote:

> Hie i'm new to the AVR environment. I'm an engineering student doing
> my final year project, my project is on a wireless transmiter that
> will receive 2 analogue inputs, store the data in memory and then
> transmit the data by GSM to a computer that will then plot the data
> on a graph. i don't know where to start..........can somebody help.
> Which micro-controller is best suited for this job?

Well, at least you start off by admitting its a homework assignment.

Final year of engineering school and you haven't yet learned which end
of the problem to start with? The CPU is the last item you should
select. Can't select a CPU until you have defined everything else that
you know what the CPU has to do.

Lets assume your problem statement and suggested solution is sane, and
start whittling at the details. I would start with, "Why GSM?" Do you
have some sort of requirement for world wide delivery of your data?

Lets say the GSM specification sticks. Now you need to figure out what
GSM radio you will use and how you will pay for its network access.

Perhaps the school project doesn't really need world wide data
delivery but all you really need is to demonstrate your project
communicating to/through a device. I find the Aerocomm (now Laird
Technologies) radios to be awfully handy, small, long range, and
inexpensive. Starting on page 5 of the Mouser catalog: http://www.mouser.com/catalog/catalogusd/638/5.pdf
through page 7. The toughest question is whether you get a peer-to-
peer model or a client/server model. 900 MHz band may not be legal in
the UK and the 2.4 GHz radios are only client/server so that might be
your answer. When only two radios are involved I don't think it much
matters. Download the user manuals and decide for yourself. I do
recommend purchase of the bundled "Starter Kits" as there is enough
there to make everything work in a few minutes. Then you can take one
side apart to build into your project. The boxed radios do not come
apart.

The Laird/Aerocomm radios interface very similar to OEM GSM modules
via asynchronous serial at TTL/CMOS voltages. The data you send to the
radios will be different, but its the same idea.

Now you can start thinking about your CPU. How much processing do you
need for your data? You've studied the GSM manual and the Laird manual
and selected one or the other, so how much processing do you need to
support your radio?

Mention two analog inputs, how much precision and how often does it
have to be updated? Time to start downloading CPU data sheets to see
if you can get the precision and speed with an on-chip A/D.

Eventually, you'll be ready to make a CPU selection.

I would suggest in the AVR family that you look for something that is
supported by the inexpensive AVR Dragon development board. Sadly, last
time I looked the only solid documentation (and supported CPU list)
for the Dragon was in the built-in help files in AVR Studio. Must
install AVR Studio to access.

Probably would want a CPU with JTAG debug interface as its usually
faster than DebugWire. But make sure if you use an AVR that you select
one with DebugWire or JTAG. These days there is no sane reason not to.
Devices without are not less expensive.

If your project is one that should ship in the millions then there is
no point worrying about a nickel here and there on the CPU as we've
already thrown our budget all to heck in purchasing an off the shelf
OEM radio module.

Once Upon A Time about 5 years ago I was faced with the task of
selecting a CPU that was to be interfaced to a cellphone and installed
under electric power meters. There was a fair amount of pre-existing
conditions that had to be met so I couldn't change the radio
selection. There was a need for two serial interfaces, that or toggle
between radio and other. In the end I selected an ATmega64L as its
price was about $3 in our quantity. As far as I was concerned that was
close enough to free that I wasn't interested in optimizing the
selection. Figured if code bloat necessitated we could jump to the
ATmega128 by just changing the P/N. Later 256's were available and
probably would work in the footprint I used.

A year later when I left the product had been in volume production for
months and object code was just over 14k bytes (of the available 64k).
Last year I found they were pushing the 64k limit due to feature-itis.

--
David Kelly N4HHE, d...@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

------------------------------------

______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.


(You need to be a member of avrclub -- send a blank email to avrclub-subscribe@yahoogroups.com )

Do not get discouraged! I worked many years with equipment that did just th=
at and added frequency hopping to it!!=A0 Currently, I do not have a name f=
or a micro controller that will serve you.=A0 It only takes a good research=
to come up with the latest.=A0 I am on it.

Down here is late at night.=A0 I have to pull my shift and get off at early=
hours of the morning.=A0 I will help you make a small research for you.=A0=
It will be about 10 hours from now.....ok?=A0=A0 nice and interesting.....=
........................................
--- On Thu, 7/16/09, David Kelly wrote:

From: David Kelly
Subject: Re: [AVR club] New to Avr & need to design a wireless (gsm) applic=
ation
To: a...@yahoogroups.com
Date: Thursday, July 16, 2009, 7:30 PM

=A0

=20=20=20=20
=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20

On Jul 16, 2009, at 4:45 PM, vmtetwa wrote:

> Hie i'm new to the AVR environment. I'm an engineering student doing=20=20

> my final year project, my project is on a wireless transmiter that=20=20

> will receive 2 analogue inputs, store the data in memory and then=20=20

> transmit the data by GSM to a computer that will then plot the data=20=20

> on a graph. i don't know where to start....... ...can somebody help.=20=20

> Which micro-controller is best suited for this job?

Well, at least you start off by admitting its a homework assignment.

Final year of engineering school and you haven't yet learned which end=20=20

of the problem to start with? The CPU is the last item you should=20=20

select. Can't select a CPU until you have defined everything else that=20=20

you know what the CPU has to do.

Lets assume your problem statement and suggested solution is sane, and=20=20

start whittling at the details. I would start with, "Why GSM?" Do you=20=20

have some sort of requirement for world wide delivery of your data?

Lets say the GSM specification sticks. Now you need to figure out what=20=20

GSM radio you will use and how you will pay for its network access.

Perhaps the school project doesn't really need world wide data=20=20

delivery but all you really need is to demonstrate your project=20=20

communicating to/through a device. I find the Aerocomm (now Laird=20=20

Technologies) radios to be awfully handy, small, long range, and=20=20

inexpensive. Starting on page 5 of the Mouser catalog: http://www.mouser.. =
com/catalog/ catalogusd/ 638/5.pdf=20

through page 7. The toughest question is whether you get a peer-to-=20

peer model or a client/server model. 900 MHz band may not be legal in=20=20

the UK and the 2.4 GHz radios are only client/server so that might be=20=20

your answer. When only two radios are involved I don't think it much=20=20

matters. Download the user manuals and decide for yourself. I do=20=20

recommend purchase of the bundled "Starter Kits" as there is enough=20=20

there to make everything work in a few minutes. Then you can take one=20=20

side apart to build into your project. The boxed radios do not come=20=20

apart.

The Laird/Aerocomm radios interface very similar to OEM GSM modules=20=20

via asynchronous serial at TTL/CMOS voltages. The data you send to the=20=20

radios will be different, but its the same idea.

Now you can start thinking about your CPU. How much processing do you=20=20

need for your data? You've studied the GSM manual and the Laird manual=20=20

and selected one or the other, so how much processing do you need to=20=20

support your radio?

Mention two analog inputs, how much precision and how often does it=20=20

have to be updated? Time to start downloading CPU data sheets to see=20=20

if you can get the precision and speed with an on-chip A/D.

Eventually, you'll be ready to make a CPU selection.

I would suggest in the AVR family that you look for something that is=20=20

supported by the inexpensive AVR Dragon development board. Sadly, last=20=20

time I looked the only solid documentation (and supported CPU list)=20=20

for the Dragon was in the built-in help files in AVR Studio. Must=20=20

install AVR Studio to access.

Probably would want a CPU with JTAG debug interface as its usually=20=20

faster than DebugWire. But make sure if you use an AVR that you select=20=20

one with DebugWire or JTAG. These days there is no sane reason not to.=20=20

Devices without are not less expensive.

If your project is one that should ship in the millions then there is=20=20

no point worrying about a nickel here and there on the CPU as we've=20=20

already thrown our budget all to heck in purchasing an off the shelf=20=20

OEM radio module.

Once Upon A Time about 5 years ago I was faced with the task of=20=20

selecting a CPU that was to be interfaced to a cellphone and installed=20=20

under electric power meters. There was a fair amount of pre-existing=20=20

conditions that had to be met so I couldn't change the radio=20=20

selection. There was a need for two serial interfaces, that or toggle=20=20

between radio and other. In the end I selected an ATmega64L as its=20=20

price was about $3 in our quantity. As far as I was concerned that was=20=20

close enough to free that I wasn't interested in optimizing the=20=20

selection. Figured if code bloat necessitated we could jump to the=20=20

ATmega128 by just changing the P/N. Later 256's were available and=20=20

probably would work in the footprint I used.

A year later when I left the product had been in volume production for=20=20

months and object code was just over 14k bytes (of the available 64k).=20=20

Last year I found they were pushing the 64k limit due to feature-itis.

--

David Kelly N4HHE, dkelly@HiWAAY. net



(You need to be a member of avrclub -- send a blank email to avrclub-subscribe@yahoogroups.com )

i think that atmega64l,atmega128 are the best cases,they have 2 RX&TX pin.
64k,128k of memory,but i get into trouble by programming 64l.
also you may to do with 2 chip of atmega8 by spi.

--- On Fri, 7/17/09, vmtetwa wrote:

From: vmtetwa
Subject: [AVR club] New to Avr & need to design a wireless (gsm) applicatio=
n
To: a...@yahoogroups.com
Date: Friday, July 17, 2009, 2:15 AM

=A0

=20=20=20=20
Hie i'm new to the AVR environment. I'm an engineering st=
udent doing my final year project, my project is on a wireless transmiter t=
hat will receive 2 analogue inputs, store the data in memory and then trans=
mit the data by GSM to a computer that will then plot the data on a graph. =
i don't know where to start....... ...can somebody help. Which micro-contro=
ller is best suited for this job?

=20

=20=20=20=20=20=20

=20=20=20=20
=20=20=20=20
=09
=09=20
=09
=09

=09
=09
=09
=20=20=20=20=20=20

[Non-text portions of this message have been removed]

------------------------------------



(You need to be a member of avrclub -- send a blank email to avrclub-subscribe@yahoogroups.com )