Fix for Windows 7 forgetting wireless security password

I’ve been going nuts on my new Dell notebook.  Every time I attempted to reconnect to many of my home access points, I would get re-prompted for the wireless security key (some WEP, some WPA).  Windows wasn’t remembering my wireless security key.  I couldn’t find any way to get them to save.  When I checked in the Wireless Center, only one connection was showing up (the one that saved the key information).  Googling wasn’t coming up with much, other than suggestions to turn off “IEEE authentication” (I think what was meant was the IEEE 802.11 service management — that is, the WLAN service) or run MSConfig and turn off the adapter service.

The last one gave me a clue.  The machine was installed (built by my IT support) with Intel PROSet/Wireless WiFi manager.  I’d always ignored it, and just used Windows wifi management.  Well … mistake.  It turns out that if you have Intel PROSet running, you need to use it to start and stop all connections — otherwise, even if you’re exclusively using the Windows manager to start/stop, it won’t save your connection information.  Switching to just using the Intel manager to start/stop solved my problem.

Samsung SyncMaster 204B died as well …

(Note on previous post … simple as pie to pop the WD drive out of the enclosure and directly mount it in my server … back with access.)

Also, my (2.5 year old) Samsung SyncMaster 204B has started to go — when first turned on, it’s dark, then gradually starts showing the picture, although as of today it remains dim and flickering.  Presumably the backlight inverter … apparently pretty common to fail, at least as discussed here (with pix and hints on fixing with capacitor replacements) and here.  You can buy at least some inverters from LCD Parts, and InventGeek has an article on how to swap them.  Luckily, I’m still under warranty (which includes the backlight) … unluckily, I bought it in the US, and am now in the Netherlands.  We’ll see how it goes …

Western Digital Elements external hard drive power supply dead

It’s certainly not been a great week for hardware for me.  Now the USB 2 Western Digital Elements (1Tb) external drive that I had hooked up to my Linux box quit on me.  Appears to be the power brick — disconnecting the enclosure and unplugging/plugging the brick causes the green light to light briefly, then fade away — although I haven’t actually tested the pins to prove that they’re dead and it’s not the enclosure itself.  The thing is only about 6 months old; what a crock.

Apparently, the failure for these drives is endemic.  See, for instance:

The consensus appears to be that WD can’t/won’t replace the power supplies without, at best, a hassle.  While Asian Power Devices appears to manufacture a compatible brick, I think I’m just going to go get a new SATA enclosure (or maybe just remount the drive internally) … which is too bad, as I liked the rubber feet on the WD enclosure.  Sigh.

Anyhow … given the similar problems others have had, and the minimal opportunities to rectify the problem, I’d stay away from these drives.

Chrome ate my Thinkpad

OK … not quite, but close enough.  I’ve now noticed that when running Chrome (in my typically abusive manner — with 15-30 tabs open for my queued up reading matter slash multi-tasking), my CPU temperature rises about 2-4° (C) due to the Chrome hammering of the CPU.  Is that an essential design trade-off (separate processes per page = constant processing of javascript, Flash, etc.), or is this just an (incredibly annoying) implementation defect, and will a future version of Chrome hibernate the processes/pages that aren’t currently active/visible (if I can’t see it, either turn it off, or nice it down so far that it won’t substantially impact the CPU).  As it is, if I leave Chrome running (again, with many tabs), I typically see CPU usage > 75%, and eventually approaching > 95%.  Luckily, Chrome has automatic saving of current tabs … so it’s a breeze to just start and stop it at will (hint, Firefox:  stop asking me if I want to save my current tabs … of course I do!).

But why am I now monitoring the Chrome effect on my CPU temperature?  Because earlier this week, in the midst of all sorts of irreproducible work (of course), my 3-year-old Lenovo T42P abruptly quit, and wouldn’t return to life.  BIOS diagnosis:  fan kaput — and to the point, the notebook would shut down immediately thereafter.

Took the old fan out, which was very easy, following the instructions in the clean T42 Manual.  For finding the right part number (a cutely named FRU = “Field Replaceable Unit” … guess I’m in the field now!), I used the Service Parts Diagram for the T42p.  Turns out that my part number – 13R2657– has been replaced by part number 41W5204 (confirmed in the Thinkpad forums, and also by IBM NL).  Since I’m in the Netherlands, I had to order from IBM NL … they were pleasant and efficient, and I received the fan overnight.  Putting the new fan in was a breeze.

Disturbingly, though, the Thinkpad booted up, but then would quit after 10-20 minutes … and by “quit”, I mean:  turn off abruptly, with no traces left in the event logs.  Not so cool.  So I installed “Emeritus'” Windows Thinkpad fan management app (if you’re running Linux on a Thinkpad, there’s a whole host of app support for you … Google it), as discussed in this thread, set it to be a bit more aggressive about kicking the fan up in response to temperature steps … and … so far, so good.  I’m holding at about 53-54° C when on battery, even with my Chrome abuse.