As many of you know, my old laptop that I was using to test PCMCIA versions of 
the bcm43xx hardware
died a few weeks ago. Thanks to the generosity of the contributors to the 
project, I was able to
replace it with a modest computer purchased on EBay. It is an M-Tech eSlate 
400k, which has an AMD
450 MHz CPU and arrived with a 4 GB hard drive with Windows 98. Using parts 
salvaged from my old
computer, I was able to bring the memory up to 256 MB, the maximum supported by 
the motherboard, and
it now has a 100 GB hard drive. I loaded openSUSE 10.2 on the machine and 
downloaded the
wireless-2.6 git tree. The machine is slower than the CPU speed would indicate; 
however, it gets the
job done.

I now have 3 different versions of the bcm43xx hardware to test. The chip 
models and pertinent
revisions are as follows:

1. BCM4306 Rev 2 with PHY Analog 1 Rev 2 and 2050 Radio Rev 2. (PCMCIA)
2. BCM4318 Rev 2 with PHY Analog 3 Rev 7 and 2050 Radio Rev 8. (PCMCIA)
3. BCM4311 Rev 1 with PHY Analog 4 Rev 8 and 2050 Radio Rev 2. (mini-PCIe card)

Performance testing is accomplished using Iperf with the server connected to my 
Linksys WRT54G V5
AP/router via a 100 BaseTX wired connection. If I boot Windows XP and run Iperf 
from it, I get 19.5
Mbs reported. As the Iperf results show considerable jitter, I always run at 
least 5 trials and
report the maximum. The distance from my test computer to the AP is 
approximately 2m. All bcm43xx
tests were done with the latest patches applied.

Each interface was tested at rates set from 11 to 48 Mbs with the resulting 
measured transmission
rate show below:

Rate Set        11      18      24      36      48
----------------------------------------------------
4306            4.65    3.34    6.64    3.48    2.34
4318            6.12    9.86    11.8    14.6    16.2
4311            5.83    9.19    10.8    12.9    3.08

The improvement over a couple of weeks ago is remarkable. The latest tweaks to 
the specifications
have been very beneficial. Thanks again to Joe Jezak. Note that the 4318, which 
used to have the
worst performance, is now clipping along with performance more than 80% of the 
Windows driver
written with full access to the hardware specs. The old 4306 with a PHY Rev 1 
still needs the most work.

A second part of my test procedure is to use a Wi-Spy 2.4 GHz spectrum analyzer
(http://metageek.net) and their Chanalyzer 2.0 software to look at the output 
from the various
radios at a distance of about 1.8 m. The results show the following:

Card            Amplitude at 1.8 m with rate=24 Mbs
---------------------------------------------------
4306            -50 to -60 dBm
4318            -60 to -70 dBm
4311            -55 to -65 dBm

I'm surprised that amplitude of signal does not correlate with performance; 
however, the 4311
amplitude does diminish a lot at the 48 Mbs rate, which probably explains the 
falloff in performance
at that rate. In addition, the 4306 showed a much poorer separation of the 
individual subchannels in
the spectrum.

Larry
_______________________________________________
Bcm43xx-dev mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Reply via email to