[CUWiN-Dev] Fwd: Re: Mesh Networking Project

Quantum Scientific Info at Quantum-Sci.com
Wed Jan 26 14:49:20 CST 2005


A tip of the hat to Pascal Dornier of pcengines.ch for a very inexpensive source for WiFi radios in Taiwan!  He says their CM9 (Atheros AR5004X chipset, mPCI) is  sub $30.


----------  Forwarded Message  ----------
> >Agree with mPCI too, but the sense of the project seems to be heading 
> >toward modifying memory of an existing router or buying a reference design 
> >and adding memory, etc.  I'm giving alternate options though.
> 
> If you can get your code to work with an existing AP chipset, it will be 
> the cheapest possible solution. The problems with that approach:
> 
> - limited expansion (second Ethernet or second radio are usually not possible)
> - these products are rather short lived, you will be at the mercy of the 
> vendor's product life cycle.

Right, this is significant.  I'll pass this along.

 
> >I would like to examine the CM9 WiFi card, which has the Atheros AR5004 
> >chipset, for the possibility of increasing Tx power.  Can you refer me to 
> >a few of these Taiwanese vendors?
> 
> www.wneweb.com.tw - I just received a load of them. There is a startup in 
> California playing with high power output / optimized RF sections, but I 
> suspect those things will not come cheap.

Thanks!

 
> > I'm also lobbying for 700 or 900MHz for better range, penetration, and 
> > quiet, but the cost of the radio may preclude this.
> 
> Those are licensed frequencies, aren't they ? And of course you would need 
> a custom radio. I suspect it will be more cost effective to use generic 2.4 GHz 
> technology with a highly directional antenna.

700MHz is not licensed, and I think 900 isn't either.  Motorola's newest Canopy uses 900MHz, and I think it makes alot of sense since pine needles are about the length of a 1/4 wave antenna for 2.4GHz.  Fir trees act like a giant, resonating, microwave-absorbing pillow.

The hope is there are reference designs or modules for these freqs, but I have to know whether there's any actual interest before investigaing further.  Devs are quite busy right now.


More information about the CU-Wireless-Dev mailing list