From: Maximilian Engelhardt <maxi@daemonizer.de>
To: brcm80211-dev-list@broadcom.com
Cc: linux-wireless@vger.kernel.org
Subject: Re: Poor wireless reception on BCM4313
Date: Fri, 12 Jul 2013 14:55:33 +0200 [thread overview]
Message-ID: <30534146.LWDmOHhHCG@eisbaer> (raw)
In-Reply-To: <1624172.s6pnfBcJL7@fuchs>
[-- Attachment #1: Type: text/plain, Size: 4955 bytes --]
On Friday 21 June 2013 20:26:25 Maximilian Engelhardt wrote:
> Hello developers,
>
> I'm having problems with the reception performance of my BCM4313 wireless
> card. The card is basically working fine, but the reception is quite bad
> compared to other wifi equipment. I first suspected a hardware problem, but
> after I tried Windows 7 with the same hardware and having no problem there
> I'm quite sure this is a software issue.
>
>
> Here is a more detailed explanation of a test setup I performed to show the
> problem:
>
> I have an AP in about 2m distance and I connected to it by my BCM4313 card
> and by a USB dongle (TP-Link TL-WN821N, using ath9k_htc). Both connections
> work and I can ping the AP using both cards.
>
> iw dev $DEV link shows a signal of about -68 dBm for the broadcom card and
> about -51 dBm for the atheros dongle. I have no idea in how far one can
> trust these values, but the difference between these two cards mostly seems
> to be 20-30 dB, also in my further experiments (greater distance from the
> AP).
>
> Now with both dongles being connected I move my laptop away from the AP and
> of course the signal is getting weaker. However at some point the broadcom
> card looses the connection to the AP while the USB dongle still has good
> connection (at this point the signal strength of the broadcom card is about
> -90dBm and -60dBm for the USB dongle). In my test with the Windows driver
> at that point the connection was still good and maximum distance till I
> lost the connection was much larger. So I don't think is a hardware
> problem.
>
> In general I can say my BCM4313 card is only usable when I'm in vicinity of
> the AP. If I move further away the connection becomes unusable slow or is
> lost, while it is still working well using other cards/drivers.
>
>
> I'm currently using a 3.10-rc5 kernel with Debian jessie (kernel is from
> experimental, but the same behavior has been there with kernel 3.9, 3.8 (and
> older)). The driver used for the broadcom card is brcmsmac.
>
>
> I did a search non the Internet regarding this issue and found some other
> people having the same problem but no real solution to this. There is a
> workaround some Ubuntu related sites, which consists of blacklisting the
> brcmsmac driver and installing the wl driver. With the wl driver this
> problem seems to be fixed according to the web (I didn't try it myself).
>
> As I have no idea where this problem could be and how I can debug it, so I'm
> writing this mail now. If you need any addition information I'm happy to
> provide it.
> From the numbers shown above it seems like there is an attenuator of 20-30dB
> active on the receiver of the broadcom card.
>
[...]
Hello,
Is there any progress on this issue?
Here are the results of some additional test I performed.
I set up a AP (11g) and connected two client to it. The clients were my
bcm4313 wireless card and the TP-Link TL-WN821N USB dongle which was connected
to the laptop with the broadcom card. I connected both cards to the AP. The
distance between the cards and the AP was about 3m.
Then I set up another wireless card (TP-WN722N) in monitor mode about 3m away
from the AP and about 3m away from the laptop with the two wireless cards.
This card was connected to a different PC and I captured the data of the
clients and the AP. I then had a look at the reported SSI. You can see my
measured results below:
SSI for packets from the AP: -27 to -23dBm
SSI for packets from client1 (broadcom): -20 to -14dBm
SSI for packets from client2 (atheros): -40 to -38dBm
The received signal from the broadcom card were quite strong. The signal from
the AP was a bit weaker. Interestingly the signal from the second card was
noticeably weaker. I didn't investigate any further as I wanted to know how
the broadcom cards performs in transmitting. So for me transmitting power
seems to be fine with the broadcom card. My guess for the weaker signal from
the 2nd client is ether some active power saving or different propagation
conditions.
Here are the reported signal strength from the clients (while being connected
to the AP):
client1 (broadcom): -74 to -75dBm
client2 (atheros): -44 to -42dBm
So for me this test shows that the card only seems to have problems on
receiving packets, transmitting seems to work fine.
Additionally I performed a test with the wl driver. I booted an ubuntu live CD
and installed the wl driver there. Then I connected to my AP using this driver
and tested reception. Everything looked like it was supposed to work. After
connecting to the AP I got a reported signal strength of about -38dBm. I then
moved away from the AP and was still able to get good connection where before
I wasn't able to get any connection at all using the brcmsmac driver.
Please let me know if I should perform any additional tests or if you need any
additional information.
Greetings,
Maxi
[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]
next prev parent reply other threads:[~2013-07-12 12:55 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-21 18:26 Poor wireless reception on BCM4313 Maximilian Engelhardt
2013-06-21 18:48 ` Arend van Spriel
2013-06-21 19:27 ` Maximilian Engelhardt
2013-07-12 12:55 ` Maximilian Engelhardt [this message]
2013-07-12 13:47 ` Piotr Haber
2013-07-12 15:27 ` Maximilian Engelhardt
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=30534146.LWDmOHhHCG@eisbaer \
--to=maxi@daemonizer.de \
--cc=brcm80211-dev-list@broadcom.com \
--cc=linux-wireless@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).