The CM9 does have issues with some pure B clients of older types, this is normally a issue with long/short preamble. And should be solved by getting a updated firmware for the 802.11b card. I do know that some cards have proven dificult in the past, but I don't use any pure B cards regularly anymore. I do have a notebook with a proxim goldcard, but I haven't used it for more then a year now, I might try it with pfsense if I have some spare time in the future. Your best bet would be other cards or updated firmwares.

I just bought a used Linksys WRT54G for $15, reflashed it with DD-WRT firmware and use it as my wireless bridge.  If you're considering this just make sure it's a version 1-4 model, v5 won't work.

It works well enough to stream pretty much any media off my server.

Great, I'll make the order :).

Unless you have reason to believe this is really a DNS tunnel, my bet is that they've just configured OpenVPN to use port 53.  Blocking DNS to servers not under your control should fix the problem.

–Bill

ok, i will try it.

thank you very much !!!! or like one says in my country : Merci énormement ! ;)

The "magic" happens in the /sbin/athctrl.sh script.

Thank you very much, it's ok now !

PcEngines is shipping cases with pre-drilled holes (2port has 2 and 3port has 1) since the end of 2005.

I see.  When a laptop that is only capable of 11b in ad-hoc connects, does it also cause all other ad-hoc 11g 54mbps connections to be throttled down to 11b?  Even with odfm=cts/rts?  I have the 11g only checkbox enabled in the pfSense GUI but it still let the laptop connect, presumably because it is technically capable of 11g in AP Client mode.

Thanks very much guys..

I already ordered one.

Some of these vendors sell wireless stuff and all have donated something to pfSense http://www.pfsense.org/index.php?id=40

–Bill

This is not implemented in GUI, but you could use this parameter on command line, and possibly also hardcode it into /etc/inc/interfaces.inc

bssid address
    Specify the MAC address of the access point to use when operating
    as a station in a BSS network.  This overrides any automatic
    selection done by the system.  To disable a previously selected
    access point, supply any, none, or - for the address.  This
    option is useful when more than one access points have the same
    SSID.  Another name for the bssid parameter is ap.

This will tell the client to only associate to the AP with whatever mac address you specified.

Any chance you use the wrong antenna port ?
Also, you might want to try to lower the output power. If the card has a poor rf output it might be generating too much noise.
I have not encountered any behaviour like this with pfsense/freebsd, so this is just a guess.
You can also try to play with the following sysctls from a shell.

dev.ath.0.txantenna: 0,1,2 (antenna port 0 or 1, both=2)
dev.ath.0.rxantenna: 0,1,2 (antenna port 0 or 1, both=2)
dev.ath.0.diversity: 0,1 (0=disable 1=enable)
dev.ath.0.tpscale: 0,1,2,3,4 (size of increment that TPC will use to up/down the power, normally 1 is the best choice, atleast that is my experience)
dev.ath.0.tpc: 0,1 (0=disable 1=enable)
dev.ath.0.tpack: 63 (ack power)
dev.ath.0.tpcts: 63 (cts power)

A little more description:

dev.ath.0.txantenna: The antenna used to transmit (normally the same as RX antenna)
dev.ath.0.rxantenna: The antenna used to receive (normally the same as TX antenna)

(A different RX antenna can be used if your regulatory restrictions makes you want to use a higher RX gain antenna to overcome the output limits of your regdomain. An example would be; you can only output 20db E.I.R.P on a long distance link, and to overcome the signal loss you use a higher gain antenna for RX gain, since you are not transmitting at more then 20dbi you can still receive with 36dbi or more if you like too, as the regdomain restriction is normally only applied to transmitting power and not receive gain. This is perfectly leagal in all of the regdomains as far as I know)

dev.ath.0.diversity: Switch off and on the use of antenna diversity (normally only used if you have two antennas connected)
dev.ath.0.tpc: Switch on or off Transmission Power Control (can be tricky in point to multipoint applications)
dev.ath.0.tpscale: Size of the increment that TPC will use to up/down the power, normally 1 is the best choice, atleast that is my experience. A higher scale value will most likely make the link drop if the signal is close to what it needs to be and the TPC is trottled down.)
dev.ath.0.tpack: controll the ack power seperatly (normally the same as tpcts)
dev.ath.0.tpcts: controll the cts power seperatly (normally the same as tpack)

All this will be controlable from GUI in 1.1 (1.0 is frozen so no new features there)

EDIT: this info is added to the wiki http://wiki.pfsense.com/wikka.php?wakka=Wireless
EDIT2: this info is also added to the doc site: http://doc.pfsense.org/index.php/Wireless_networking_with_pfSense

These have been in FreeBSD's perforce for a long time but nobody has stepped up to the plate to convert all the drivers and thus get it into the tree.

I wouldn't hold my breath on this one.

Not having a clue what I'm trying to do is probably my problem! I'm not very good with 'paint', but have tried. It really isn't supposed to like like a mutant smiley face  :-.

What I am trying to do is have Wireless clients authenticated through the captive portal (I have set up IAS on my domain controller, and pointed the captive portal at it, with shared secrets). When authenticated, I want them to use WAN as default gateway and access resources on LAN.

Does any of that even make sense?? I hope so.
I would like to use WPA as well, but the wl-330g does not support its use with a RADIUS server, and as it's the access point, not pfSense, I guess I can't use pfSense's WPA support?

Network.JPG
Network.JPG_thumb
Network.JPG_thumb

It all depends on how you use it, like if you run in "pure-g" mode there is no need for this.
In general one would like CTS/RTS if you run in mixed mode (b/g) in most other scenarios you would have it "off".
What protection mode does is prefixing each OFDM data frame with a request to send/clear to send (RTS/CTS) complimentary code keying (CCK) frame sequence. The duration fields of the RTS and CTS frames should allow the IEEE 802.11b node to correctly set its network allocation vector and avoid collisions with the subsequent OFDM frames.
CTS to self wil lhave the effect of a 802.11g node telling a 802.11b node that it is going to transmit.
Basically this is used as a collision avoidence, since the G node can talk to a b node in DSSS but a B node can not understand OFDM.

Also note that Protection mode has a performance penalty because of overhead.

That is correct, the soldering for highpower cards was a older mod iirc.

I tried with TKIP and the i works. Tanks

Ok , excellent

thanks.