Loren has it right - you loose 15 dB in a link budget and that's the difference in working or not for many installs. I try to keep 20 dB margin when I'm done but that's not always possible.
One very very tiny detail remains though - cisco under-rates their hardware (its laboratory performance is significantly superior to the glossy spec sheet). Most other carriers just use the numbers the chipset vendors provide (which is why they all look the same) - but few if any know much about the proper design of a sensitive, selective RF front end that can handle 10's of MHz of bandwidth. That's what separates the taiwain & china copycat cards from the cisco's of the world. And I've been to Suzhou, china - I can guarantee there is NO anechoic chamber setup for testing this stuff, just trucks and boats waiting to load up on shore and head for the USA. I think Cicso inherited their specs from aironet when they bought them out, and aironet wanted to be conservative about it. They improved the receiver specsheet from the 340 to 350 series in the post aironet-era. Also, nobody publishes real world tests - those conducted in harsh environments other than a 50 ohm cable (with stuff like broadband impulse noise mixers tossed in for good measure) - then compare the before/after data rates. It's during this test that the "home" 802.11 cards suddenly break their link, while better cards just suffer some speed / performance degradation. I crank up the "offending" signal until each card breaks by 0.1 dB increments and record the difference between manufacturers using 2 tests - in band and out of band interferance. The real world has all sorts of noise - a 50 ohm cable and attenuator test is way too friendly for this type of hardware. A last note - if you put some of these "home" grade APs in an attic or hoffman box outside on a day like today, you will exceed their published thermal spec (at least here in the bay area). I have also tested vendors cards inside environmental chambers and once again, you get "good" designs and not-so-good designs that won't take an attic or dropped ceiling for that matter. One company had to pull out their whole network of cheap APs from the T-bar dropped ceiling tiles one summer when they called me to consult. (Their IT department had bragged about how much $ they saved) I stuck a temperature probe in the ceiling, replaced their AP with a cisco 340 (best at the time) and poof- there was a small spot of working wireless. A week later there was a big pile of (unnamed) APs on the ground and a bunch of empty cisco boxes & cds laying there and the whole thing was working better than before. I was just glad they chose to pay by the job instead of by the hour for that one! Everett > > In short, WAP11's are deaf and mute compared to Cisco 350's. > > If memory serves, WAP11 transmit power is 35 milliwatt verses Cisco's 100. WAP11 > recieve sensitivity is -82 dBm verses Cisco's -90 dBm. > > You have lost around 15 dB from your link budget. > > A quick calc suggests the link should work with WAP11's and 24 dBi antennas. What > type of antennas are you using? > > Loren Zemenick > > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] Behalf Of Joe Haggard > Sent: Monday, September 22, 2003 9:51 AM > To: [EMAIL PROTECTED] > Subject: [BAWUG] Linksys Wap 11 Ver 2.6 to ver 2.8 wireless bridge > > > I am trying to setup a wireless bridge between my house and a friends house. This > distance is 11.2 miles, clear line of sight, across a valley here in the Mojave > Desert. We had the link working great using some barrowed Cisco 350s, 78% signal > strength and very low packet loss and retries. We had to return then 350s and are > now trying to get the link back up with Linksys Wap11s. He has a ver 2.6 and I have > the 2.8 both are running current firmware. We set the linksys' up in bridge mode > arcross my living room to verify that it was working. We then hooked up each AP on > the high gain antenna's and its not working. > > We would like any advice on things to try or reasons why it is not working. We are > also not sure how to tell which antenna is the right one and which one is the left > one. > > Thanks in advance, > > Joe Haggard > ?zm?? k > {.n+? X (j)fj b?? *?, > > -- > general wireless list, a bawug thing <http://www.bawug.org/> > [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless > -- general wireless list, a bawug thing <http://www.bawug.org/> [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
