|
 |
small PCB request for an open source project - RF Cafe Forums
|
ZigDCC
Post subject: small PCB request for an open source project
Unread postPosted: Thu Jun 02, 2005 6:27 am
Hello folks,
May one of you develop a small 2x3cm PCB design for an RF headend
including RF chip itself and optionnal space for SMT ceramic antenna?
Here is an idea of what this project would need to start
using RF:
Image
From the middle to the left you can
see Chipcon's cc2420, an IEEE 802.15.4 CoS working in the 2.4GHz ISM
band.
On the bottom of this design (outside of this PCB) an antenna
can be wired.
On the right side is a microprocessor and its cristal.
That one is not needed anymore.
The new PCB must be designed
as module which can be soldered on a motherboard which will be equipped
with its own processor (Atmel, Microchip or what ever else is needed
behind the head end).
The new PCB should also provide options
to either SMT ceramic antenna onboard or for external (in house) antenna.
The cc2420 documentation is here:
http://www.chipcon.com/files/CC2420_Data_Sheet_1_2.pdf
With something like 6x8cm its reference design is a way too big
to fit in small embedded applications and is also a way too expensive
($540 in Chipcon's webshop).
The cc2420 can be found for $4 and
an SMT antenna costs not much more.
Objectives would be to meet
$20 - maybe much less - for assembled cc2420 RF headend.
One of the issues I see (for costs) is that Chipcon's ref. design
is based on a quad layer PCB.
I would expect a more standard
dual layer PCB. Can this be acheived?
The chip itself
features up to 100m RF range.
The project requirements will be
to set up connexions from 4cm - only - to some 10th of meters.
From 4cm - only - to some 5 to 10 meters will be welcome if there
are issues with huger ranges.
I could make several attempts
by my own with Cadsoft's Eagle PCB design tool but this will probably
not be the best way for RF and power efficiency.
Any volunter
to take this over or does such a headend already exist and I missed
it?
Best regards
Top
ZigDCC
Post
subject:
Unread postPosted: Thu Jun 02, 2005 6:40 am
Here
is a closest view of the chip on its ref. design:
Image
Top
Guest
Post subject:
Unread postPosted:
Thu Jun 02, 2005 8:14 am
How much$$$$$$
Top
ZigDCC
Post subject:
Unread postPosted: Thu Jun 02, 2005
8:36 am
I have currently none, my last are gone in a cheap AVR
JTAG programmer :)
What should I think from this:
http://www.egr.msu.edu/classes/ece480/g ... ctures.htm
Does
it mean mimics of the ref design provide acceptable results?
Image
Regards
Top
ZigDCC
Post
subject:
Unread postPosted: Thu Jun 02, 2005 10:41 am
I can
add following regarding $$$
All ZigDCC related materials including
the hardware designs will be made available under GNU GPL license.
for more informations see project aim at Sourceforge:
https://sourceforge.net/docman/display_doc.php?docid=28698&group_id=140277
Regards[/i]
Top
ZigDCC
Post subject:
Unread postPosted: Thu Jun 02, 2005 1:22 pm
I will leave that
cc2420 aside because prototyping is too expensive due to quad layer
PCB.
cc2500 is much more interesting as its ref design is only
2 layer PCB based. cc2500 differs also in the way it doesn't provide
hardware encryption. A later cc2520 may feature this.
So the question is still open.
May one just mimic a ref.
design and acheive honorous performances?
Regards
cc2500 specs can be found here:
http://www.chipcon.com/index.cfm?kat_id
... dok_id=183
PCB design is much more simple as shown here,
on page 2:
http://www.chipcon.com/files/CC2500_221204_LoRes.pdf
Top
ZigDCC
Post subject:
Unread postPosted:
Fri Jun 03, 2005 9:54 am
ZigDCC wrote:
So the question is
still open.
May one just mimic a ref. design and acheive honorous
performances?
not sure... I'd like to use standard 1.6mm
FR4.
http://ieeexplore.ieee.org/iel4/5437/14 ... ber=667012
http://www.rogerscorporation.com/mwu/techtip4.htm
But all and every RF blocks are built inside the cc2420 so how could
the substrate interact with the chip internals? Are such chips designed
to be used exclusively on quad layers RO4000 substrates?
The loss throught strips to the antenna seems acceptable on RF4 knowing
a ceramic chip antenna could be placed some 1/3 inch away from the RF
chip itself:
Image
So why would one use expensive
quad layer RO4000 substrate where FR4 may approx do the job?
What is approx, will it be close or less than 1dB or should several
10dB loss be more resonnable expectations?
Some WUSB components
are qualified to work on standard FR4:
http://www.fractus.com/2-4.htm
Why should one use quad layer RO4000 where two layers FR4 may
do the job?
The components and layout between the chip
and the antenna may be an issue but everything can be wired really close
and even unmatched according to chipcon's datasheet:
Quote:
The antenna should be connected as close as possible to the IC. If the
antenna is located away from the RF pins the antenna should be matched
to the feeding transmission line (50 ?).
I think
I will give it a try in dual layer FR4 with a small ceramic chip antenna
close to the RF CoS.
Dear Guest, have you an opinion about my
chances? :wink:
Top
CK
Post subject:
Unread
postPosted: Thu Jun 09, 2005 5:10 pm
Why Chipcon?
I've already
seen early designs of this size on .8mm board material with F antenna
on double sided FR4 using the Freescale MC13192 chip and 9S08 family
processor with estimated costs lower than what you want.
Top
ck
Post subject: zigbee
Unread postPosted: Sat
Jun 25, 2005 3:11 pm
CK wrote:
Why Chipcon?
I've already
seen early designs of this size on .8mm board material with F antenna
on double sided FR4 using the Freescale MC13192 chip and 9S08 family
processor with estimated costs lower than what you want.
have
you experience on zigbee?
Top
ZigDCC
Post
subject:
Unread postPosted: Sat Jun 25, 2005 3:23 pm
ZigDCC
wrote:
ZigDCC wrote:
So the question is still open.
HAVE
YOU EXPERIENCE ON ZIGBEE?
May one just mimic a ref. design
and acheive honorous performances?
not sure... I'd like
to use standard 1.6mm FR4.
http://ieeexplore.ieee.org/iel4/5437/14
... ber=667012
http://www.rogerscorporation.com/mwu/techtip4.htm
But all and every RF blocks are built inside the cc2420
so how could the substrate interact with the chip internals? Are such
chips designed to be used exclusively on quad layers RO4000 substrates?
The loss throught strips to the antenna seems acceptable on
RF4 knowing a ceramic chip antenna could be placed some 1/3 inch away
from the RF chip itself:
Image
So why would one
use expensive quad layer RO4000 substrate where FR4 may approx do the
job?
What is approx, will it be close or less than 1dB or should
several 10dB loss be more resonnable expectations?
Some WUSB
components are qualified to work on standard FR4:
http://www.fractus.com/2-4.htm
Why should one use quad layer RO4000 where two layers FR4 may
do the job?
The components and layout between the chip
and the antenna may be an issue but everything can be wired really close
and even unmatched according to chipcon's datasheet:
Quote:
The antenna should be connected as close as possible to the IC. If the
antenna is located away from the RF pins the antenna should be matched
to the feeding transmission line (50 ?).
I think
I will give it a try in dual layer FR4 with a small ceramic chip antenna
close to the RF CoS.
Dear Guest, have you an opinion about my
chances? :wink:
Posted 11/12/2012
|
|
|
|
|
