This "weird un " happened to me last week, I was just glad that I had read about it a few years back. A woman I work with at the plant, knowing that I had been assigned as a hardware tech on various PC manufacturing lines brought me this troubleshooting problem: her pc would crash when the toilet was flushed. I had been to her house and had noticed a well house. Turns out, the well is her sole source of water and her husband had run the feed for the pc from a overhead light. Everytime her high draw well pump or the window ac kicked in , the lights would brown out causing disconnect or lockup. As I said before , I recalled this being covered in one of my trade pubs 10 years back, just never thought I would see a "toilet go flush, computer go poop" situation. Thought I would share this, as I know many of us have wells. Never tap power for a PC off the light circuit, its usually more vulnerable and use filters on pwr input.

-- Jay Blair (jayblair678@yahoo.com), August 24, 2000

Answers

Hi Jay,

How about the PC printer cable that was run too close to the wall light switch. When the light switch was turned on, the printer would print a test page.

(:raig

-- Craig Miller (CMiller@ssd.com), August 24, 2000.

I'll contribute this one..

There was a jr college in Florida that built a new computer center and installed their mid-range system. Life was good. The campus building program continued which included a new library building.

During this period of time the rage was to have solid copper sheeting for roofing. After the new library was finished, the computer center started experiencing strange things. Every night at midnight the system would freeze-up then IPL itself. Techs took the machine apart and put it back together. No change. Replaced all manner of hardware. No change. They even set with the machine at night and sure enough, at midnight - BOOM; system dumps and then IPLs. They finally brought in additional assistance and began doing site audits and various monitoring. No problem could be found with the building, power, software, or hardware. Then they started doing field strength readings. And thus they found their problem. At midnight a local radio station would shutdown and part of their process was to broadcast a particular tone. Said tone caused the roof of the library to resonate and produce enough RF that was beamed (because of the angles of the roof) right into the data center and into the machine. As I understand it, they moved the machine a couple of feet to one side and their problem ended.

it's always something...

j

-- j (jw_hsv@yahoo.com), August 24, 2000.

Jay! Your post is comical but brought to mind the problem you described regarding the electrical. One of the hardest things to detect, if you've never seen it before, is a 'ground unbalance'. A loose or broken ground wire/connection will cause wierd things to happen. For example: A water pump kickin on and lights on that phase going real dim while lights on the opposing phase will get real bright. I've seen'm in the past when the "high" side would actually burst the bulbs! Now this wasn't a common event but it has happened before. I've also had the voltmeter on those "high" sides while they were high. How about a voltage of 180 volts?!!!!! Lights are rather inexpensive to replace but assumiong a puter is on that "high" side. Without a surge protector it's fried puter. With a surge protector-- it can STILL be fried puter. If anybody has lights dimming while other lights are getting bright--call an electrican--quickly! Don't want any houses burning! Matt. 24:44

-- hoot (hoot@pcinetwork.com), August 24, 2000.

Some comical things happen (unless your the one using the equipment). I have spent 19 years in my field combating EMI around equipment. I am continually amazed at how "stone aged" some "hi tech" minds can be. Once a system we used for testing would always flake out during the early morning hours (midnite to 5 a.m.) the smart ones would scoff at the line operators having problems, saying they ran the system (at 9 a.m.) with no problems. I came in during the time window when the problems were occuring and came to the conlusion that the glitches were happening at the same time all the other data systems were backing up files. The equipment in question used 128 data i/o lines on 1 ribbon cable, no shield grounds. The other tech and myself redesigned the data trunk on the equipment to 8 32 pin ribbons,grounding alternating pins as shield. The early morning gremlins went away. Just another reason to live simply beyond the sidewalks.

-- Jay Blair (jayblair678@yahoo.com), August 25, 2000.

A few years ago, our department at work was in temporary quarters while the renovations were being completed. One afternoon, our heads turned as it seemed someone had started popping some very large popcorn.

It turned out that the sequential popping was the surge protectors on all the PCs connected to a faulty circuit. The culprit was a loose wire shorting to chassis ground of the conduit. It took us a day to replace the surge protectors, but $400 worth of prevention saved $20K worth of equipment and lots of additional headache and downtime.

BTW, the only worthy "personal" surge protection equipment I have come across is made by Tripp-Lite -- they are called Iso-Bars. They are expensive ($40-50) for 6 outlets and a phone jack, but well worth it. You can get what you pay for or you can get what you don't pay for -- it's your choice.

-- Mike O (olsonmr@yahoo.com), August 26, 2000.