Date: Sun, 6 Nov 2011 12:25:55 -0800 From: "Adrian Chadd" <adrian.chadd@gmail.com> To: "lev@freebsd.org" <lev@freebsd.org>, "Adrian Chadd" <adrian@freebsd.org> Cc: "freebsd-wireless@freebsd.org" <freebsd-wireless@freebsd.org> Subject: Re: update - new hal flag to force a full chip reset Message-ID: <4eb6ed5d.2401440a.7470.ffffa7aa@mx.google.com> In-Reply-To: <939228459.20111106231805@serebryakov.spb.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
Well I am still trying to establish what exactly is going on in noisy envir= onments. The ar9160 doesn't behave this way. A bit of background in case you feel like doing some digging.. The ANI patent doc explains what is going on. The ani code tweaks some para= meters for weak signal detection, as well as what the desired signal size= and level is. What I am not clear on yet is whether it's a signal sizing= issue (ie, whether the ambient or bursty noise is falling within the signa= l level/size ani programs in) or whether the radio setup/calibration code= is doing something wrong in 2.4ghz mode. So it may be worthwhile reading the patent doc and ar5416 ani code, add som= e debugging tweaks and do some experiments with the parameters. I've setup ath9k and it has similar behaviour. Sigh. :) Adrian Sent from my Palm Pre on AT&T On Nov 6, 2011 11:19 AM, Lev Serebryakov <lev@freebsd.org> wrote:=20 Hello, Adrian. You wrote 6 =D0=BD=D0=BE=D1=8F=D0=B1=D1=80=D1=8F 2011 =D0=B3., 22:15:19: >>> Good luck, and thanks to everyone who has been testing my code= out! >> Do you need any help? Maybe, running AP with additional debu= g output, >> or something like this? > Just dump the NF calibration results (sysctl dev.ath.X.hal.debug=3D0x8) > and crank it up to say, 2 to 5 seconds (sysctl hw.ath.longcal=3D5). > Let's see what the NF results are like. I've sent your these data, do you need more from exactly the same setup? :) I could make tests in the middle of working say, when environment isn't so noisy -- less APs are seen, and spee is better. > I have a nasty feeling that the AR9280 series NICs report some rather > high NF values. The values themselves aren't actually raw NF dBm, > they're just a dB value - but the fact that it's so high is likely a > symptom of the overall problem (which is likely that the signal levels > are all configured wrong when faced with interference.) BTW, is here other ATH PCI chipsets with 802.11n? I've checked all Mirkotik Mini-PCI cards, for example, all of them are based on 9820, with different setups of RF part, some of them are single-band (2.4Ghz only), some of them are dual-band, some of them are dual-band and have amplifier, which adds some dBs of output, but chipset is the same. And the same with TP-Link cards, which could be purchased here (in Russia). It seems, --=20 // Black Lion AKA Lev Serebryakov <lev@FreeBSD.org>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4eb6ed5d.2401440a.7470.ffffa7aa>