Need advice on wireless network

On Mon, 25 Feb 2008 20:06:21 -0800 (PST), I said, "Pick a card, any
card" and amerdsp <amerdsp@hotmail.com> instead replied:

>On Feb 25, 5:02 pm, donald <Don...@dontdoithere.com> wrote:
>> amerdsp wrote:
>> > On Feb 25, 2:54 pm, donald <Don...@dontdoithere.com> wrote:
>>
>> >> Did you design your hardware or just copied what was easy.
>>
>> >> You need to design your own hardware to get the distance you want.
>>
>> > I am planning to design my own but for now I just ran the demo that
>> > came with the CC2500 chip.  Thats when I ran into the distance/range
>> > problem.
>>
>> Your not understanding whats going on here.
>>
>> This is not a problem.
>>
>> The board you tested was not designed for your application.
>>
>> It was designed as a demo.
>
>The kit is marketed as a development kit not a demo board. Since this
>kit is meant to advertise the best in the CC2500 chip, I assumed they
>would put a decent antenna to stand out from the crowd.

Nonsense. It's made to test in lab conditions not in the field. You
want field performance, buy a working, off the shelf model of
something already designed.

>> Your application requires distance, this board was not designed for
>> distance.
>>
>> As has been suggested, an antenna will help.
>
>I agree the antenna is very important.  What I was asking is that
>perhaps the CC2500 is not meant to have long range transmission.  You
>can use the best antenna out there, but if you do not have the power
>to drive it nothing is going to happen.

There are high gain antennae available. I gave you links to places
where you can actually build your own. A small stub antenna may have
a gain of 2db to 5db while a dish or bowtie will exceed 16db.
That'll put you into the 100 meter range with just .5mw or less
depending on other conditions.

>> But YOU need to design it for the application,
>
>Yes, indeed.  Thats what I am planning to do but not before I know
>that the hardware I am using is able to get the ranges I want.
>
>As I said, RF is not my strength and I thank you for your input.

Then get someone who is good with RF to design your weakest link.
It's not that hard to learn. In another post, I gave you links to
lots of help. Go there and read. Those guys wrote for the amateur.

Why did you come here to ignore the only advice that will help you
out? You can ask further questions about antenna design. While most
here are digital guys there are a few with more well rounded
expertise. No harm in asking unless, like you did here, you choose
to dismiss it as too hard.

More power is not the answer. It may even be illegal where you live.
--
Ray

On Tue, 26 Feb 2008 13:21:26 +0900, Ray Haddad
<rhaddad@iexpress.net.au> wrote:

>On Mon, 25 Feb 2008 20:06:21 -0800 (PST), I said, "Pick a card, any
>card" and amerdsp <amerdsp@hotmail.com> instead replied:

>>The kit is marketed as a development kit not a demo board. Since this
>>kit is meant to advertise the best in the CC2500 chip, I assumed they
>>would put a decent antenna to stand out from the crowd.
>
>Nonsense. It's made to test in lab conditions not in the field. You
>want field performance, buy a working, off the shelf model of
>something already designed.
>
>>> Your application requires distance, this board was not designed for
>>> distance.
>>>
>>> As has been suggested, an antenna will help.
>>
>>I agree the antenna is very important.  What I was asking is that
>>perhaps the CC2500 is not meant to have long range transmission.  You
>>can use the best antenna out there, but if you do not have the power
>>to drive it nothing is going to happen.
>
>There are high gain antennae available. I gave you links to places
>where you can actually build your own. A small stub antenna may have
>a gain of 2db to 5db while a dish or bowtie will exceed 16db.
>That'll put you into the 100 meter range with just .5mw or less
>depending on other conditions.

Antennas with "gain" have also directivity so this is mainly usable
with stationary nodes in more or less line of sight conditions. In a
stationary system, the sensors could use directional antennas, but the
master would have to use an nondirectional antenna if the sensors are
not in the same general direction (or use switchable directional
antennas and switch in the correct antenna, when accessing a specific
sensor). 

However, if the nodes are stationary, why use wireless connections in
the first place, what prevents the use of wires ?

In applications, in which the communication path is through
reflections or in which the nodes might be movable or
reflecting/absorbing surfaces (such as humans) will move in and out of
the communication path, the directional antenna can do more harm. In
such situations, antennas with more than 3-6 dBi gain (illuminating
less than half of an hemisphere) should be avoided, in order to be
able to use the best reflection at a particular moment.

If the evaluation board simply relies that a suitable PCB track may
radiate something, but may be quite lossy, with gains in the order of
-6 to -20 dB, so even a simple 1/4 wave external antenna would improve
the distance considerably.

It should be noted that when using license free frequency bands, the
device may have sufficient range in an interference free test setup,
but in the actual world with various kind of interference from various
sources, the range could be drastically reduced. 

Paul

On Feb 25, 11:06 pm, amerdsp <amer...@hotmail.com> wrote:

> I agree the antenna is very important.  What I was asking is that
> perhaps the CC2500 is not meant to have long range transmission.  You
> can use the best antenna out there, but if you do not have the power

I haven't looked at the CC2500 but I assume it puts out at least 0dBm.
If you want more power, you need to add a PA.

If you don't have the expertise and equipment (got a 2.5GHz spectrum
analyzer?) it is very difficult - quasi-impossible - to experiment
with this stuff and you are really much better served by buying a
module with a pre-matched antenna. Bluetooth, maybe.