Belkin Cardbus USB2 adaptors too hot.

Julian Stacey jhs at berklix.org
Thu Mar 19 12:31:24 PDT 2009


Problem solved :-)
All hangs I mentioned till now were coming from 2 identical enclosures
I have, they report as:
	vendor=0x058f product=0x6390 devclass=0x00 devsubclass=0x00
	release=0x0100 sernum="" intclass=0x08 intsubclass=0x06
2 other enclosures of different ID I have are OK.

Should we extend devd or BSD USB code to issue a brief warning
when devices match any in a lookup list of suspect rogues
? (Not the first flakey USB device I've encountered, Some Sony 2G
sticks were bad, all 4, ( "vendor" "0x10d6"; "product" "0x1100";
"devclass" "0x00"; "devsubclass" "0x00"; "release" "0x0100"; )

PS Thanks for you power etc advice Warner :-)

Further backround before conclusion above:
  {
  I moved same new 250G disc in same enclosure to another 7.1-RELEASE
  tower PC with & without USB hub, & with a current doubler cable (ie
  to suck up to a 2nd 0.5A from 2nd socket if disk wants)
  
  testblock -v /mnt/tmp/bla1
  stopped writing at 5G.
  
  /var/log/messages & console:
   umass0: BBB reset failed, TIMEOUT
   umass0: BBB bulk-in clear stall failed, IOERROR
   umass0: BBB bulk-out clear stall failed, IOERROR
   umass0: BBB reset failed, TIMEOUT
   umass0: BBB bulk-out clear stall failed, IOERROR
   umass0: BBB reset failed, TIMEOUT
   g_vfs_done():da0a[WRITE(offset=65536, length=2048)]error = 5
   g_vfs_done():da0a[WRITE(offset=6144000, length=16384)]error = 5
   g_vfs_done():da0a[WRITE(offset=6160384, length=6144)]error = 5
   g_vfs_done():da0a[WRITE(offset=198229819392, length=16384)]error = 5
   ....
  
  I swapped the disc to a different USB to IDE converter, omitted case,
  disc does not get hot, but USB to IDE chip painfully hot (on good converter)
  + another power doubler cable, straight into PC, no hub,
  it wrote then read all remaining FS free space of 217 G succesfully.
  
  I swapped my other previously failing 120 G disc out of bad enclosure,
  into a known good enclosure, NO problems now on that either, have written
  106 with no hang, & read verified 60 G as I type.
  }

Cheers,
Julian
-- 
Julian Stacey: BSDUnixLinux C Prog Admin SysEng Consult Munich www.berklix.com
  Mail plain ASCII text.  HTML & Base64 text are spam. www.asciiribbon.org


More information about the freebsd-usb mailing list