[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Grant]

I've installed my Ubiquity SR71-E with a PCIe adapter but I get:

wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 1/3)
wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 2/3)
wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 3/3)
wlan0: direct probe to xx.xx.xx.xx.xx.xx timed out

I'm connecting to a WPA2 AP and I'm using the same settings that work
with an ath5k card.

- Grant

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Mohammed Shafi]

On Fri, Jul 29, 2011 at 12:01 AM, Grant <emailgrant@gmail.com> wrote:
> I've installed my Ubiquity SR71-E with a PCIe adapter but I get:
>
> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 1/3)
> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 2/3)
> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 3/3)
> wlan0: direct probe to xx.xx.xx.xx.xx.xx timed out
>
> I'm connecting to a WPA2 AP and I'm using the same settings that work
> with an ath5k card.

please provide the following details:

*lspci
*which version of compat wireless or wireless testing you are using
*is  it a noisy environment and how far is the distance between AP and Station.

if you can please provide
*wpa-supplicant version
* does the same thing happens with no security?

>
> - Grant
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>



-- 
shafi

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Mohammed Shafi]

On Fri, Jul 29, 2011 at 12:04 PM, Mohammed Shafi
<shafi.wireless@gmail.com> wrote:
> On Fri, Jul 29, 2011 at 12:01 AM, Grant <emailgrant@gmail.com> wrote:
>> I've installed my Ubiquity SR71-E with a PCIe adapter but I get:
>>
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 1/3)
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 2/3)
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 3/3)
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx timed out
>>
>> I'm connecting to a WPA2 AP and I'm using the same settings that work
>> with an ath5k card.
>
> please provide the following details:
>
> *lspci
> *which version of compat wireless or wireless testing you are using
> *is ?it a noisy environment and how far is the distance between AP and Station.
>
> if you can please provide
> *wpa-supplicant version
> * does the same thing happens with no security?

we can identify if some thing is wrong, (like 4-way handshake timeout)
by looking at /var/log/daemon.log
and for this please enable the debugging messages in
usr/share/dbus-1$ vi system-services/fi.epitest.hostap.WPASupplicant.service

[D-BUS Service]
Name=fi.epitest.hostap.WPASupplicant
Exec=/sbin/wpa_supplicant -u -s -dddddddddKt
User=root
~
~
~
~


>
>>
>> - Grant
>> _______________________________________________
>> ath9k-devel mailing list
>> ath9k-devel at lists.ath9k.org
>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>>
>
>
>
> --
> shafi
>



-- 
shafi

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Grant]

>> I've installed my Ubiquity SR71-E with a PCIe adapter but I get:
>>
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 1/3)
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 2/3)
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 3/3)
>> wlan0: direct probe to xx.xx.xx.xx.xx.xx timed out
>>
>> I'm connecting to a WPA2 AP and I'm using the same settings that work
>> with an ath5k card.
>
> please provide the following details:
>
> *lspci

02:00.0 Network controller: Atheros Communications Inc. AR928X
Wireless Network Adapter (PCI-Express) (rev 01)

> *which version of compat wireless or wireless testing you are using

I've tried wireless-tools-29 and 30_pre9.

> *is ?it a noisy environment and how far is the distance between AP and Station.

I'm in an apartment so there are some other networks around but it
isn't very dense.  The station and 802.11g AP are 15 feet apart in the
same room with a nearly direct line-of-sight.

> if you can please provide
> *wpa-supplicant version

wpa_supplicant-0.7.3-r2

> * does the same thing happens with no security?

Yes.

- Grant

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Grant]

>>> I've installed my Ubiquity SR71-E with a PCIe adapter but I get:
>>>
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 1/3)
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 2/3)
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 3/3)
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx timed out
>>>
>>> I'm connecting to a WPA2 AP and I'm using the same settings that work
>>> with an ath5k card.
>>
>> please provide the following details:
>>
>> *lspci
>> *which version of compat wireless or wireless testing you are using
>> *is ?it a noisy environment and how far is the distance between AP and Station.
>>
>> if you can please provide
>> *wpa-supplicant version
>> * does the same thing happens with no security?
>
> we can identify if some thing is wrong, (like 4-way handshake timeout)
> by looking at /var/log/daemon.log
> and for this please enable the debugging messages in
> usr/share/dbus-1$ vi system-services/fi.epitest.hostap.WPASupplicant.service
>
> [D-BUS Service]
> Name=fi.epitest.hostap.WPASupplicant
> Exec=/sbin/wpa_supplicant -u -s -dddddddddKt
> User=root

I followed all of these instructions and rebooted but I don't have
/var/log/daemon.log.  I use metalog and there is nothing of interest
in /var/log/everything/current.  Please let me know if there is
anything I can do to get more debug output.

- Grant

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Grant]

>>> I've installed my Ubiquity SR71-E with a PCIe adapter but I get:
>>>
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 1/3)
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 2/3)
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx (try 3/3)
>>> wlan0: direct probe to xx.xx.xx.xx.xx.xx timed out
>>>
>>> I'm connecting to a WPA2 AP and I'm using the same settings that work
>>> with an ath5k card.
>>
>> please provide the following details:
>>
>> *lspci
>> *which version of compat wireless or wireless testing you are using
>> *is ?it a noisy environment and how far is the distance between AP and Station.
>>
>> if you can please provide
>> *wpa-supplicant version
>> * does the same thing happens with no security?
>
> we can identify if some thing is wrong, (like 4-way handshake timeout)
> by looking at /var/log/daemon.log
> and for this please enable the debugging messages in
> usr/share/dbus-1$ vi system-services/fi.epitest.hostap.WPASupplicant.service
>
> [D-BUS Service]
> Name=fi.epitest.hostap.WPASupplicant
> Exec=/sbin/wpa_supplicant -u -s -dddddddddKt
> User=root

It's fixed!  The problem was actually interference from an ath5k
802.11g card I had in the same system.  I've noticed that any wireless
card installed in the system will pump out enough RF to interfere with
any other card in the system even if the interfering card's drivers
are compiled out of the kernel.  I guess they transmit as soon as
they're powered up?

- Grant

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Mihai Moldovan]

* On 31.07.2011 01:41 AM, Grant wrote:
> It's fixed!  The problem was actually interference from an ath5k
> 802.11g card I had in the same system.  I've noticed that any wireless
> card installed in the system will pump out enough RF to interfere with
> any other card in the system even if the interfering card's drivers
> are compiled out of the kernel.

I've had similar problems before, sorry. More specifically, I've been
using 2 ath9k operated cards in one system and my STAs were able to
connect for a short time, but were as well disconnected after at most 5
minutes. Another fact aggravating this problem was using the same SSID
for both AP mode cards.

I suspect using two cards on the same base frequency (i.e. 2GHz) in the
same system (or even in the same range) is a bad thing to do, even if
the channels themselves are not overlapping.

On the other hand, my current setup involves 2 ath9k driven cards, each
with separate SSIDs and on different base frequencies (2GHz and 5GHz)
which doesn't show any of these problems and indeed is running fine,
from what I can tell.

Every now and then my laptop will "disconnect" from the network and any
connection time out until I manually tell the system to reconnect to the
network or cycle the WLAN card's power (thus forcing a reconnect), but
I'm not entirely sure what's causing it, as I'm not getting any errors
on either my STA (which, granted, is an OS X system and I don't even
know where Apple would log wireless errors but via syslog or in the
Kernel log ring buffer, no one showing any error traces though) or my
AP. Weirdly it does only seem to happen when streaming a video file over
NFS, so I'm reluctant to blame it on the 2 cards setup. :)

Best regards,


Mihai

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4369 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20110731/f687e723/attachment.bin 

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Adrian Chadd]

I bet there's some RF leaking out, either from the TX, or maybe the PLL/clock.

Unfortunately it's the kind of thing you can only diagnose with a
sensitive spectrum analyser.

I had that problem with a box w/ two (high powered) NICs in 2ghz mode,
and I only discovered this issue with a spectrum analyser. I found
that there was enough RF leaking in/out of the u.fl connectors and
(cheap) cable that it started counting as interference.

Silly, but true. :)


Adrian

On 31 July 2011 09:04, Mihai Moldovan <ionic@ionic.de> wrote:
> * On 31.07.2011 01:41 AM, Grant wrote:
>> It's fixed! ?The problem was actually interference from an ath5k
>> 802.11g card I had in the same system. ?I've noticed that any wireless
>> card installed in the system will pump out enough RF to interfere with
>> any other card in the system even if the interfering card's drivers
>> are compiled out of the kernel.
>
> I've had similar problems before, sorry. More specifically, I've been
> using 2 ath9k operated cards in one system and my STAs were able to
> connect for a short time, but were as well disconnected after at most 5
> minutes. Another fact aggravating this problem was using the same SSID
> for both AP mode cards.
>
> I suspect using two cards on the same base frequency (i.e. 2GHz) in the
> same system (or even in the same range) is a bad thing to do, even if
> the channels themselves are not overlapping.
>
> On the other hand, my current setup involves 2 ath9k driven cards, each
> with separate SSIDs and on different base frequencies (2GHz and 5GHz)
> which doesn't show any of these problems and indeed is running fine,
> from what I can tell.
>
> Every now and then my laptop will "disconnect" from the network and any
> connection time out until I manually tell the system to reconnect to the
> network or cycle the WLAN card's power (thus forcing a reconnect), but
> I'm not entirely sure what's causing it, as I'm not getting any errors
> on either my STA (which, granted, is an OS X system and I don't even
> know where Apple would log wireless errors but via syslog or in the
> Kernel log ring buffer, no one showing any error traces though) or my
> AP. Weirdly it does only seem to happen when streaming a video file over
> NFS, so I'm reluctant to blame it on the 2 cards setup. :)
>
> Best regards,
>
>
> Mihai
>
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>
>

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Grant]

> I bet there's some RF leaking out, either from the TX, or maybe the PLL/clock.

I wonder why the two antennas on a single card don't interfere with each other?

> Unfortunately it's the kind of thing you can only diagnose with a
> sensitive spectrum analyser.
>
> I had that problem with a box w/ two (high powered) NICs in 2ghz mode,
> and I only discovered this issue with a spectrum analyser. I found
> that there was enough RF leaking in/out of the u.fl connectors and
> (cheap) cable that it started counting as interference.

Which spectrum analyzer do you use and are you happy with it?

A quick question.  My small, cheap, freebie wifi antennas work on
2.4Ghz and 5Ghz.  Why are more expensive antennas stated to work only
with one frequency or the other?

- Grant


> Silly, but true. :)
>
>
> Adrian
>
> On 31 July 2011 09:04, Mihai Moldovan <ionic@ionic.de> wrote:
>> * On 31.07.2011 01:41 AM, Grant wrote:
>>> It's fixed! ?The problem was actually interference from an ath5k
>>> 802.11g card I had in the same system. ?I've noticed that any wireless
>>> card installed in the system will pump out enough RF to interfere with
>>> any other card in the system even if the interfering card's drivers
>>> are compiled out of the kernel.
>>
>> I've had similar problems before, sorry. More specifically, I've been
>> using 2 ath9k operated cards in one system and my STAs were able to
>> connect for a short time, but were as well disconnected after at most 5
>> minutes. Another fact aggravating this problem was using the same SSID
>> for both AP mode cards.
>>
>> I suspect using two cards on the same base frequency (i.e. 2GHz) in the
>> same system (or even in the same range) is a bad thing to do, even if
>> the channels themselves are not overlapping.
>>
>> On the other hand, my current setup involves 2 ath9k driven cards, each
>> with separate SSIDs and on different base frequencies (2GHz and 5GHz)
>> which doesn't show any of these problems and indeed is running fine,
>> from what I can tell.
>>
>> Every now and then my laptop will "disconnect" from the network and any
>> connection time out until I manually tell the system to reconnect to the
>> network or cycle the WLAN card's power (thus forcing a reconnect), but
>> I'm not entirely sure what's causing it, as I'm not getting any errors
>> on either my STA (which, granted, is an OS X system and I don't even
>> know where Apple would log wireless errors but via syslog or in the
>> Kernel log ring buffer, no one showing any error traces though) or my
>> AP. Weirdly it does only seem to happen when streaming a video file over
>> NFS, so I'm reluctant to blame it on the 2 cards setup. :)
>>
>> Best regards,
>>
>>
>> Mihai

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Adrian Chadd]

On 1 August 2011 03:27, Grant <emailgrant@gmail.com> wrote:
>> I bet there's some RF leaking out, either from the TX, or maybe the PLL/clock.
>
> I wonder why the two antennas on a single card don't interfere with each other?

That's a question for the RF/baseband engineers, who likely know more
about it than I do.

But having antenans too close to each other can cause interference. I
remember reading an article or two on the near field effects of
antennas that are too closely spaced; some digging may pull up further
clue.

>> Unfortunately it's the kind of thing you can only diagnose with a
>> sensitive spectrum analyser.
>>
>> I had that problem with a box w/ two (high powered) NICs in 2ghz mode,
>> and I only discovered this issue with a spectrum analyser. I found
>> that there was enough RF leaking in/out of the u.fl connectors and
>> (cheap) cable that it started counting as interference.
>
> Which spectrum analyzer do you use and are you happy with it?

When I had access to one, the Agilent N9912A served me pretty well.
They're just all very expensive and not really in the realm of "toys"
for free software guys doing this (mostly) for free.

On the flip side, you can get some atheros-based NICs that output
basic spectral analyser type information - check out
http://www.metageek.net/products/wi-spy/ for an example.

> A quick question. ?My small, cheap, freebie wifi antennas work on
> 2.4Ghz and 5Ghz. ?Why are more expensive antennas stated to work only
> with one frequency or the other?

Again an RF engineer would know better, but I bet it's because cheap
ones are like 0 or 1 dBi gain and the antenna element is small enough
to provide ok gain and SWR at both wavelengths.



Adrian

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Alex Hacker]

Really boys, it's unbelievable story! We works with Atheros based cards for
years, our RF engenners have spectrum analyzers connected to these cards
continuously. I'd just ask him - nobody never see any out of band interference
from these cards except the pci-e carrier at 2.5 and 5.0 GHz. So I'm shure that
the issue lies somewhere else, not in RF plane.
About antennas. Of coarse any object in the anntenna near field causes a
distortion of gain/directivity and SWR so that the system performance can
degrade. Anyway passive antenna system can not be source of out of band
interference.

With best regrads,
Alex.

On Mon, Aug 01, 2011 at 10:22:53AM +0800, Adrian Chadd wrote:
> On 1 August 2011 03:27, Grant <emailgrant@gmail.com> wrote:
> >> I bet there's some RF leaking out, either from the TX, or maybe the PLL/clock.
> >
> > I wonder why the two antennas on a single card don't interfere with each other?
> 
> That's a question for the RF/baseband engineers, who likely know more
> about it than I do.
> 
> But having antenans too close to each other can cause interference. I
> remember reading an article or two on the near field effects of
> antennas that are too closely spaced; some digging may pull up further
> clue.
> 
> >> Unfortunately it's the kind of thing you can only diagnose with a
> >> sensitive spectrum analyser.
> >>
> >> I had that problem with a box w/ two (high powered) NICs in 2ghz mode,
> >> and I only discovered this issue with a spectrum analyser. I found
> >> that there was enough RF leaking in/out of the u.fl connectors and
> >> (cheap) cable that it started counting as interference.
> >
> > Which spectrum analyzer do you use and are you happy with it?
> 
> When I had access to one, the Agilent N9912A served me pretty well.
> They're just all very expensive and not really in the realm of "toys"
> for free software guys doing this (mostly) for free.
> 
> On the flip side, you can get some atheros-based NICs that output
> basic spectral analyser type information - check out
> http://www.metageek.net/products/wi-spy/ for an example.
> 
> > A quick question. ?My small, cheap, freebie wifi antennas work on
> > 2.4Ghz and 5Ghz. ?Why are more expensive antennas stated to work only
> > with one frequency or the other?
> 
> Again an RF engineer would know better, but I bet it's because cheap
> ones are like 0 or 1 dBi gain and the antenna element is small enough
> to provide ok gain and SWR at both wavelengths.
> 
> 
> 
> Adrian

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Alex Hacker]

Important exeption is the channel sidebands and baseband filters fall-off. You
sholdn't use adjacnt channels for different networks. At least 40MHz gap
should be left between H20 channels, and as much as 80MHz between HT40+ and
HT40- channels.

On Mon, Aug 01, 2011 at 11:19:03AM +0600, Alex Hacker wrote:
> Really boys, it's unbelievable story! We works with Atheros based cards for
> years, our RF engenners have spectrum analyzers connected to these cards
> continuously. I'd just ask him - nobody never see any out of band interference
> from these cards except the pci-e carrier at 2.5 and 5.0 GHz. So I'm shure that
> the issue lies somewhere else, not in RF plane.
> About antennas. Of coarse any object in the anntenna near field causes a
> distortion of gain/directivity and SWR so that the system performance can
> degrade. Anyway passive antenna system can not be source of out of band
> interference.
> 
> With best regrads,
> Alex.
> 

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Adrian Chadd]

I've also noticed that there's sometimes quite noticable differences
in how "clean" the output is from various NICs.

Eg, when pushing the SR-2 or SR-71A at max power, versus the Unex DNMA
series high power NICs. The Unex ones are much, much cleaner.


Adrian


On 1 August 2011 16:31, Alex Hacker <hacker@epn.ru> wrote:
> Important exeption is the channel sidebands and baseband filters fall-off. You
> sholdn't use adjacnt channels for different networks. At least 40MHz gap
> should be left between H20 channels, and as much as 80MHz between HT40+ and
> HT40- channels.
>
> On Mon, Aug 01, 2011 at 11:19:03AM +0600, Alex Hacker wrote:
>> Really boys, it's unbelievable story! We works with Atheros based cards for
>> years, our RF engenners have spectrum analyzers connected to these cards
>> continuously. I'd just ask him - nobody never see any out of band interference
>> from these cards except the pci-e carrier at 2.5 and 5.0 GHz. So I'm shure that
>> the issue lies somewhere else, not in RF plane.
>> About antennas. Of coarse any object in the anntenna near field causes a
>> distortion of gain/directivity and SWR so that the system performance can
>> degrade. Anyway passive antenna system can not be source of out of band
>> interference.
>>
>> With best regrads,
>> Alex.
>>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>

[ath9k-devel] Ubiquity SR71-E: direct probe timed out

[Alex Hacker]

Hi, Adrian!
Yeah, the Ubiquity is young company trying to get his place on the market.
So that they like to overdrive Sirenza RF PAs to impermissible levels. This
leads to raising unwanted emission in adjacent channels, but I don't
believe that they have a fake FCC certificate. :) Another issue with Ubiquity
cards is the absence of output RF filters, so it can emit some 2nd (>4.8GHz)
and 3rd (>7.2 GHz) harmonics when work at high power.
BTW the SR71 specification claims 26dBm at MCS0 and only 19dBm at MCS7.
Anyway if you do not use adjacent channels it is should not be a big problem
(except for the neighbouring networks of course).

Best regards,
Alex.

On Mon, Aug 01, 2011 at 05:07:36PM +0800, Adrian Chadd wrote:
> I've also noticed that there's sometimes quite noticable differences
> in how "clean" the output is from various NICs.
> 
> Eg, when pushing the SR-2 or SR-71A at max power, versus the Unex DNMA
> series high power NICs. The Unex ones are much, much cleaner.
> 
> 
> Adrian
> 
> 
> On 1 August 2011 16:31, Alex Hacker <hacker@epn.ru> wrote:
> > Important exeption is the channel sidebands and baseband filters fall-off. You
> > sholdn't use adjacnt channels for different networks. At least 40MHz gap
> > should be left between H20 channels, and as much as 80MHz between HT40+ and
> > HT40- channels.
> >
> > On Mon, Aug 01, 2011 at 11:19:03AM +0600, Alex Hacker wrote:
> >> Really boys, it's unbelievable story! We works with Atheros based cards for
> >> years, our RF engenners have spectrum analyzers connected to these cards
> >> continuously. I'd just ask him - nobody never see any out of band interference
> >> from these cards except the pci-e carrier at 2.5 and 5.0 GHz. So I'm shure that
> >> the issue lies somewhere else, not in RF plane.
> >> About antennas. Of coarse any object in the anntenna near field causes a
> >> distortion of gain/directivity and SWR so that the system performance can
> >> degrade. Anyway passive antenna system can not be source of out of band
> >> interference.
> >>
> >> With best regrads,
> >> Alex.
> >>
> > _______________________________________________
> > ath9k-devel mailing list
> > ath9k-devel at lists.ath9k.org
> > https://lists.ath9k.org/mailman/listinfo/ath9k-devel
> >