Electronic Design


Finding A Home For Old Computers
Great article, right on the money \["Technology Haves And Have-Nots: Can We Share?" May 1, p. 46\]. I'm an electrical engineer who graduated way back in 1984. There has been one heck of a lot of change since then. In the last few years, however, I have opted to stick with an older machine, a Pentium 120 with 32-Mbyte RAM running OS/2. It serves my needs just fine, and I find even the 33.6k net access sufficient. I flatly refuse to get a Pentium 700 with 128-Mbyte or more RAM, because I just can't see the need for it. It seems a lot of others fall in the same boat. You can surf the net very adequately with a 486/100 and a 28.8 or better modem for basic information.

In Ottawa here, we have had a number of "computer garage sales" held at local churches, etc., several times a year. These give people the opportunity to pick up trailing-edge computers at good prices, as well as peripherals and monitors. There's also a store called "Computer Recyclers" where older computers and parts are sold. This is good for those who want to try these things out without having to go for a full-priced new unit, those on a budget, and those like me who just don't want the latest stuff. It also minimizes the number of computers going to landfills and gets them to a home.

Another good example of this is how many golf courses use PCs to keep track of handicaps, etc. You don't need a Pentium 700 to do this. Instead, a 486 or low-end Pentium works just fine.

Over the years, I've repaired many monitors and sold about 30 computers to date, ranging from 486/466s to a Pentium 266. So, there's definitely a need for the trailing-edge systems. Many not-fully Y2K-compliant machines can be fixed by simply installing a software patch or upgrading the BIOS.
Ross Ulan
Electrical Engineer
Nav Canada

Approaching Reliability Limitations
"Can Computers Be Made As Easy To Use As Appliances" \[May 29, p. 46\] stimulated a much abused and frayed nerve I have dealt with in only the last couple of years out of the many since acquiring my first CPU. Sure they are easy to use, but hardware problems can turn that around! I have changed three hard drives in the last couple of years and had memory go bad on a name-brand video card after nine months of use. Are tolerances for read/write on HDs just too tight? Can we realistically cram so much video memory on a card with only a few chips (density)? Yes, it's great that designers/manufacturers are increasing densities in ever smaller packages, but we may be approaching theoretical limits where reliability is concerned, unless a sure-fire method of manufacture to handle those tolerances is achieved!

These problems are on my home computer. I turn the unit off (monitor too) between times. I run a full SCSI system. This allows me to scale back on a main processor because the AHA 2940U2W SCSI (on PCI bus) adaptor takes a lot of the load off the processor. I even turned "virtual memory" off to save disk caching (256-Mbyte SDRAM on the motherboard). I like this as the adaptor vendor doesn't (nor is it required to) turn as many needed iterations in advancement as the main processor industry does. I can use the same SCSI adaptor for many years compared to the need to upgrade a processor in much less time. So, those software vendors can go right ahead and churn out ever more complicated code within increasing file sizes (requiring more real estate on a hard disk). My main processor just loafs along.

I will always start out a new SCSI HD with a low-level format followed by media verification using the utility in the Adaptec 2940 ROM. Hardly ever do I come up with a bad cluster, but sometimes there are a few. When I start having problems with operating, Windows 98 will eventually spit and can't even correct the problem with it's own Scandisk. Using the SCSI utility again to verify media will reveal A LOT of bad clusters! I counted 76 bad clusters on this 2.4G Samsung drive after it would get only 78% into the routine.

This drive saw nine months of use. The disk prior to the Samsung was a 1.2G and lasted about 13 months. I had replaced a 345M Maxtor HD with the 1.2G only to get more room and ended up putting the Maxtor back! The Maxtor is seven years old and has no problems. Now I have a 4.5G Seagate and am keeping my fingers crossed.

I have also gone into a "minimalism" mode. I no longer use my home computer for such uses as Quicken, photo archiving with a digital camera, midi music entertainment, and MP3 functions, as I'm tired of starting from scratch to reinstall devices. I would use every IRQ available and that isn't enough. Now I use the computer to occasionally surf and for daily e-mail.

At work I use a laptop which gets replaced about every two years. Why? They start to break down somewhere within and you know there isn't much modularity with their construction.

I could go on, but I guess I should sum up this dissertation with the notion that a platform with no moving parts would be the most robust. In that way, I could say the computer could be as easy to use (at least last as long?) as appliances!
Tedd Carr
Project Engineer
S.W. Public Service Co.

Food For Thought
I am pleased from issue to issue reading your magazine. Thank you! It may be curious to you, but for me the main courses and side dishes swapped places: I read the magazine mainly for the thoughts presented in the editorial, Don Reinertsen's and L.J. Kamm's columns, and last—not least—RAP's column.... Those are the main courses that I won't find anywhere! The other stories are side dishes (please don't omit anyway).
Christian A. Kranich

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.