i now see that zadig did infact... not do it's job. it's running now, as admin. waiting on completion.

thanks for the help. i feel like an idiot.

Yes windows seems to sometimes get an error that on linux means to drop the device.
On windows it may just mean I need to reinitialise it.
It's at the top of the todo list now.

Anyway once it does drop the device (ZOMBIE), hotplug will pick it up again within 5 seconds.
You can see in the API ... the number of devices will grow by 1 each time this happens (and also each time, a new ZOMBIE will appear with nodev = true)
The ZOMBIE name comes from the fact that I will (soon?) come up with a way to map the hotplug devices back onto the old matching ZOMBIE device ... and revive it ... it's not completely dead yet

Anyway, this is all display related, the total hashing power should still be the same with a small loss per device (of between 0 and 5 seconds) each time one goes ZOMBIE and then hotplugs.
The 'going ZOMBIE' process is immediate - no delay at all, as soon as a short list of errors occurs, it a slam dunk on the head, ZOMBIE straight away.
If I work out a tidy way to revive it before giving up on it, the only gain will be the time from ZOMBIE to hotplug that is currently lost.
It will simply be the underlying usbutils code that makes the ZOMBIE flag, intercepting the ZOMBIE happening and trying to revive it before giving up on it.

... and yes this is all Windows ... just in case anyone else was wondering.

Linux should hopefully be rock solid.

Thanks for trying to help, anyone else?

I used this, two years ago.

https://en.bitcoin.it/wiki/Mining_hardware_comparison

Strange things happened here on my rpi (Raspbian wheezy). 3.2.0 detected my Lancelot (BLT), plug and pray is working fine. But no luck with any of my Block Erupter USBs, not detected, no plug and play at all.

All devices connected using the same powered usb 2.0 hub. These are working fine with 3.1.1 using -S options.

Whenever I quit 3.2.0 and try running 3.1.1 again, it can't access the USB devices anymore, have to reboot the rpi...

Any hints are welcome.

I am having exactly the same problem on the pi. My ZTEXs Quads are recognised without problems but the Eruptor Gizmodos won't show up.
libusb-1.0.0 is installed.

Perhaps it is working with Kano because he is using the Fedora dist on the Raspberry instead of Wheezy?

pastebin (don't paste here) the output of the end of 'dmesg' when you plug the AMU in
also 2 cgminer outputs:
1) cgminer -n
2) cgminer -D -T -verbose ... your options ... 2> debug.log
and let that run for about 30 seconds - and pastebin it or upload the log somewhere

I don't have a powered hub, I guess I better get one to see what's happening myself
(tried to get one last week and oddly none of the shops I went to had one ... except one that was crap)

One other thing to try is to leave it unplugged for about a minute then see if that fixes it
I had mine go into a weird state not being recognised after using different OSs and being left unplugged for a while fixed it for me
The error I got when this happened in the debug log was an error 4 setting the baud
I might try see if that is reproduceable - the error and the unplug fix, and if so, when it happens print a message to do that.

Odd.  Get this on the benchmark (immediately )

 [2013-05-31 15:00:09] Started at
 [2013-05-31 15:00:09] Pool: Benchmark
 [2013-05-31 15:00:09] Runtime: 380564 hrs : 0 mins : 9 secs
 [2013-05-31 15:00:09] Average hashrate: 0.0 Kilohash/s
 [2013-05-31 15:00:09] Solved blocks: 0
 [2013-05-31 15:00:09] Best share difficulty: 0
 [2013-05-31 15:00:09] Share submissions: 0
 [2013-05-31 15:00:09] Accepted shares: 0
 [2013-05-31 15:00:09] Rejected shares: 0
 [2013-05-31 15:00:09] Accepted difficulty shares: 0
 [2013-05-31 15:00:09] Rejected difficulty shares: 0
 [2013-05-31 15:00:09] Hardware errors: 0
 [2013-05-31 15:00:09] Utility (accepted shares min): 0.00/min               
 [2013-05-31 15:00:09] Work Utility (diff1 shares solved min): 0.00/min

 [2013-05-31 15:00:09] Stale submissions discarded due to new blocks: 0         
 [2013-05-31 15:00:09] Unable to get work from server occasions: 0             
 [2013-05-31 15:00:09] Work items generated locally: 0
 [2013-05-31 15:00:09] Submitting work remotely delay occasions: 0             
 [2013-05-31 15:00:09] New blocks detected on network: 0

 [2013-05-31 15:00:09] Summary of per device statistics:

Never mind

# ./cgminer -D -T --verbose
 [2013-05-31 15:02:15] Started cgminer 3.2.0
 [2013-05-31 15:02:15] All devices disabled, cannot mine!
#