On Mon, 20 Apr 1998wrote: > The real win of the SMOS cardpc is when you're able to make > your own carrier board with the pins you want drawn off the > cardpc; otherwise a regular pc104 cpu card is a far superior > solution. The cardpcs have the lowest power specs for cpus and vga I've seen anywhere. Which I went this route. > The cardpc has a couple of annoying disadvantages > I became intimately familiar with over time: > > 1. The bios battery backup is not included, so you have to > make up your own solution. My best efforts ended up substantially > less robust than a soldered-on-the-board coin chip would've been. > And, since the bios is on the cardpc, if the unit ever separates > from the battery for any reason, you lose your bios setup and must > restore it manually. I've had this problem as I've rewired my wearable a dozen times, but lately its been no problem, and resetting the bios the way I want takes less than 10 seconds. I just have a pair of wires running to a small board with a lithium ion button battery and my dc-dc converter. > 2. The units (at least the 486/100 units) dramatically overheated > during regular use. I had the misfortune of giving a demo in a > slightly warm room in front of a customer and having the board > lock up on me. It had been ventilated, but had no heat sink. I'm using the 486/40 and even after a couple of hours running in an enclosed fanny pack with no circulation the SMOS card is just *warm*. > 3. The cardpc itself is a slightly heavy unit with no locking > connectors. As a result, it tends to dislodge itself naturally > during jostling. It's possible to have a cardpc that looks like > it's firmly seated that is in fact not making a connection. What board have you used? Intelec's or Parvus or something else? I've had no problems with it getting loose. I've even wiggled it in its socket while its running without any problems. :) > I strongly recommend that anyone trying to make a wearable avoid > the cardpc line for these reasons; it's just too much headache for > what you get. comments just for counterpoint. I agree with most of your concerns, and would only recommend this cpu for a wearable with many of the reservations you mention. -Paul R. Paul McCarty / DARS Coordinator /
/ x52059 317 Lattimore Hall, University of Rochester, Rochester, NY 14627 Computers don't make mistakes;what they do,they do on purpose.-Dale/KOTH
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail