On Jun 26, 2008, at 11:56 AM, Michael Buesch wrote:
On Thursday 26 June 2008 03:00:58 Dale Walsh wrote:ssb-sprom -i minipci_sprom -o minipci_new_sprom --antgbg 0x88So you seriously think that increasing the antenna gain value in the sprom by about 34 dBi is doing any good?Who in their right mind would attempt a 34dbi increase?The quoted command above, which changes the antenna gain by about 33.5dBi, is a quote from your mail that explains the ssb-sprom bug.
That was an example valse I used so I could easily see what was modified, making a change of +0x01 or -0x01 means you have to look hard to see what changes to ensure it is the correct byte, at this time I'm working from a file and want to be sure that whatever is modified in the file is correct and currently it is not.
Most cards shouldn't need to be tweaked at all while others might require a minor increase or decrease to improve performance, this is all I am after.
An increase of this magnitude is not a wise increase considering that an increase of 10dbi is a signal gain of 1000 times of all the ambiants (including noise). An increase of 1-3dbi should make a difference and it might need a decrease rather than an increase, I'm talking about performance tuning, not going nuts with an increase so strong that the noise level becomes so overwhelming that there is no way the signal would not be drowned out.You should seriously go to a good bookstore and buy a book that explains what an "antenna gain" is. However, as an alternative you can simply increase the antenna gain value of one of your cards by 3dBi. You'll notice what happens then. You'll probably be surprised. :) (However, you'll probably explain the results of the test with your weird noise theories. hm...)
Weird noise theory, maybe reading a book on RF theory might help you to understand about noise and the S/N ratio and how the noise can affect superimposed modulation and frequency shifting which is the principal of wlan cards, wireless routers and AP's.
When you increase input sensitivity it does not just increase the sensitivity for the signal, it also increases sensitivity to everything which includes white noise.
You are implying that there is no white noise and this white noise does not affect the signal and that just isn't true.
The receiver portion of most lan card is PLL based with a clipping circuit to prevent the input signal from over saturating the the first stage of the receiver and causing damage, increasing input sensitivity beyond this level does not increase the signal past the clipped level but it does allow the level of the noise to reach this same level if sensitivity is turned up significantly.
A 1db or 2db increase in input sensitivity is probably more than enough to improve a cards performance if it lacks it and can compensate for design changes in poorly designed cards, a 10db increase is essentially an increase of 1000% which makes a 500mw signal look like a 5watt signal, this also increases the generated harmonics, the 3rd order harmonics being the most damaging since very little filtering is included in wlan cards.
We are now discussing RF fundamentals and we should be discussing the application which is what I originally was writing about, if you wish to continue a discussion about RF principals I would be more than happy to discuss this off list since it does not apply to fixing the ssb-sprom application.
-- Greetings Michael.
-- Dale
PGP.sig
Description: This is a digitally signed message part
_______________________________________________ Bcm43xx-dev mailing list [email protected] https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
