This morning I had to reboot snark (my desktop box), with a Unicomp UB404LA (the trackball variant) attached. Ubuntu failed to see it on boot, but picked it up properly when it was unplugged and replugged.
This is the keyboard that has been flaky before on a machine where I have never seen problems with Unicomps. It's actually my wife's keyboard; we swapped because it was frequently dropping communication with the USB hub on minx, her desktop.
On the other hand, Cathy reports no issues at all with my keyboard on her machine (it's a UB40PGA, the nipple-mouse variant).
The attached link was the beginning of this troubleshooting saga. I think the next step is to try running the UB404LA from a powered USB hub.
This is the keyboard that has been flaky before on a machine where I have never seen problems with Unicomps. It's actually my wife's keyboard; we swapped because it was frequently dropping communication with the USB hub on minx, her desktop.
On the other hand, Cathy reports no issues at all with my keyboard on her machine (it's a UB40PGA, the nipple-mouse variant).
The attached link was the beginning of this troubleshooting saga. I think the next step is to try running the UB404LA from a powered USB hub.
i also have the trackball variant (what a waste, next one is gonna be just a plain @#$@$# keyboard), and it has the same behaviour. intel 6 series EHCI chipset/rev05. lsusb shows the device has a 100ma max draw, which is not much considering my phone draws 500ma and the USB2.0 spec guarantees 500ma per host port. same behaviour occurs if the keyboard is the only usb device on the machine.
# lsusb -v | egrep '(Bus 0|MaxPower)'
Bus 004 Device 010: ID 17f6:1109 Unicomp, Inc
MaxPower 100mA
Bus 004 Device 012: ID 22b8:4362 Motorola PCS
MaxPower 500mA
im not about to bounce this box during the workday, but IIRC, the device never initializes or enumerates from warm/cold boot. if it was power on current draw, one would think the usb controller would be properly flipping out due to HW resets.
dont think its a power issue, but interested to see if the external hub fixes it.
[edit] might not be an amperage thing, but according to some accounts, the USB controller in the unicomp isnt exactly USB compatible; undervoltage within USB spec will cause it to not init properly. i'll dig out my DI-148 and get a scope on it during boot sometime...Jun 25, 2013
Eric RaymondOwner+Alex Lustenberg
I confirm that lsusb shows 100mA draw on my device as well. My hub says 0mA MaxPower which I'm assuming is spurious.
Your post is most interesting, especially the last paragraph. Please investigate further and keep us posted. If we can verify that tis as a model- or part-specific problem perhaps we can bring some pressure on UniComp to fix it.Jun 25, 2013- yup, the internal hubs will show 0ma, as they are passive. even then, the upstream USB2 controller ports should have 500ma available per spec, plenty to drive 100ma the unicomp exposes as its requirement.Jun 25, 2013
