Sangoma U100 failing every Monday – USB port problem or Wanrouter issue?

Home » Asterisk Users » Sangoma U100 failing every Monday – USB port problem or Wanrouter issue?
Asterisk Users 9 Comments

Hi Everyone,

We are using two Sangoma U100 (USB FXO) units connected to an Acer Aspire
Revo (little PC running on Atom). The units work beautifully except for
Monday 🙂

It maybe a conincedence or maybe the fact that Saturday/Sunday is off and
something happens where one of these U100 modules goes into sleep and that’s
when all the 4 Dahdi channels are lost.

So, I have been getting Monday morning calls (Tuesday today but yesterday
was a day off anyhow) for the past 3 weeks complaining no inbound/outbound
calls can be made through PSTN.

When I log in to the server if I do a “wanrouter hwprove” I only see one of
the units. If I ask them to disconnect and reconnect the U100 that is
flashing (which is the one that failed) and do “wanrouter hwprobe” now it
shows again. So, I can do a wancfg_dahdi and it brings it up. Sometimes, I
have to recompile the whole wanpipe. But more or less I think it’s the USB
that fails or it could also be the wanrouter itself that fails.

Sangoma is closed today but maybe others from here can shed some light as it
also may pretain to my dahdi/asterisk settings. I know that for sure the
whole wanrouter/dahdi setup goes bad if the computer is abruptly shut-down
which is really stupid (always been the case with Sangoma Wanpipe) and hence
there is a safe shutdown script that I put for Wanrouter which does the job
if the computer is shutdown properly. I don’t believe the UPS failed over
the last few off days but I am looking into the logs anyhow.

I appreciate your feedback and let me know what info I can post here that
may help resolve the issue (such as output from dmesg or lspci?).


9 thoughts on - Sangoma U100 failing every Monday – USB port problem or Wanrouter issue?

  • Hi Bruce,

    The following would be useful for starters:

    1. cat /etc/wanpipe/*.conf

    2. ifconfig -a (from a working and non-working situation)

    3. lspci -v and lsusb -v (from a working and non-working situation)

    4. wanrouter hwprobe verbose (from a working and non-working situation)

    5. /var/log/messages (near the date the problem happened)

    Moises Silva
    Senior Software Engineer
    Sangoma Technologies Inc. | 100 Renfrew Drive, Suite 100, Markham ON L3R 9R6
    t. 1 905 474 1990 x128 | e.

  • Thanks for the input. I can not replicate the situation as it happens
    randomely or maybe over the weekend. However I have sent you all the
    requested command and logs in a separate e-mail for your analyzes. The only
    thing that stood out at me was the output of “lsusb -v” at the very end
    where it timed out.

    Since all lines didn’t work I am to assume that both module went down but
    per my diagnoses with “hwprobe” I could see one unit connected and the other
    was not when the problem happened. Simply connecting/disconnecting that unit
    or connecting it to another port solved the problem and it showed up in

    This is an Acer Aspire Revo mini PC. I am wondering if the U100s draw too
    much power? The only other USB connected device is the thumb size wireless
    connector for the keyboard.

    Acer computer:

    Looking forward to your analysis.



  • Hi Bruce,

    Don’t know if this will help – but I will butt in with what I have 🙂

    I’ve been using a Sangoma U100 adapter for about 2 years now. It is
    connected to a Compaq V2120 laptop (Celeron M 1.4GHz processor) which
    serves as my home server. It is actually the main reason I went for the
    U100 – as I couldn’t add a PCI or PCIe card to a laptop to get the FXO
    ports I needed. I have to say I really like the U100 – I believe it is
    the only low(ish) cost USB based FXO interface on the market.

    I have had occasional problems with it. I remember it used to just stop
    working – and the lights would start flashing. If I remember correctly –
    I went to Sangoma’s website and downloaded the latest wanpipe drivers,
    compiled and installed them – and everything was ok after that. At the
    moment I’m running Asterisk and wanpipe 3.5.11. I can’t remember
    what version of wanpipe was giving problems, I’m afraid.

    I also found that mine doesn’t really like to be hot-plugged – it just
    freezes the system with strange characters on the screen. But that was a
    while ago. Since I’ve learned it’s foibles – it must be at least one
    year since I had to look at it.



  • Thanks for the input. I have the latest drivers but it seems that there is
    some serious incompatibility issue with the kernel as when the FLASHING
    happens even if the system is restarted it’s still not detected. One has to
    re-plug it in and then it shows in “wanrouter hwprobe”.

    It could also be that the atom board is not compatible with the driver in my
    case. It your /var/log/messages upon restart do you any line that might
    match this:
    *cd /var/log/*
    *grep -o “device not accepting address” **
    *grep -o “USB device is disconnected” **
    *dmesg | egrep “device not accepting address” *
    *dmesg | egrep “USB device is disconnected” *

    If not then you are not experiencing the same issue. If you have then it’s a
    universal issue and not hardware specific.

    I would really appreciate it if you look into your logs and let me know.



  • Hi,

    If you are asking me to look into the logs when the U100 is not
    recognised and it is flashing – I’m afraid I can’t help much – as this
    problem went away and haven’t had issues any more for at least 6-8
    months now. Considering the U100 persist in giving problems even after
    computer restarts somehow sounds to me like a hardware issue. I wouldn’t
    be surprised if the Acer Aspire motherboard is not coping very well with
    both of them being plugged in – although, obviously, it should.

    This linux-usb FAQ seems to also suggest that the problem is hardware
    (based on the “device is not accepting address” error message):

    Have you tried disabling in bios everything you don’t use – maybe sound
    card, wifi etc. to free some IRQ’s?

    I run kernel – just in case it helps.

    Let me know if you need any other info. If you find a final solution – I
    would be curious to know – just in case I will stumble over this problem.

    It wouldn’t be a bad idea to build a script which grep’s lsusb output –
    and run it through cron maybe every 10 minutes – to try and determine
    more precisely when does the U100 stop being recognised. Then you can
    look through the logs to see if there is anything else happening in the
    system around the same time.



  • Thanks for the input. The errors I pointed out should be available in the
    /var/log/messages or messages.1 .2 .3 etc…or in dmesg each time you reboot
    the machine. At least that is the case with mine. If you don’t have it then
    it’s I guess very specific to this atom board. I know someone else added
    something along the lines “irq=fixconflict” to get rid of the issue with a
    wireless USB mouse showing same symptoms and CentOS. IRQ conflicts is also
    one of my thoughts as well but I can’t play at BIOS level at this time as
    the server is thousands miles away from me.



  • Hi Bruce,

    I have just rebooted the server and can confirm that neither dmesg,
    /var/log/messages or /var/log/syslog contain “device not accepting
    address” or “USB device is disconnected” error messages.


    At least that is the case with mine. If you

  • I’m a bit late here, but I can say that Sangoma support has always been
    extremely helpful when I call them. If you haven’t already, definitely
    give them a call.