In most high-technology fields today, the pace of change is mind boggling. If you buy a new PC, it's obsolete before you get it out of the door. New models are waiting in the stockroom to replace the one you bought, offering faster speeds, bigger hard drives, and features you never dreamed of.

The same can be said of your cell phone service, cable television system, and even your vehicles.

What? You don't have a built-in satellite navigation system? Even those of us who are "techies," and I do consider myself one, are often overwhelmed.

Sometimes, it is hard to resist the urge to try something new. Many years of advertising leaves us with an immediate reaction that "new" and "improved" always go together. But we also know too many instances of new, unproven products that reached the field and failed badly.

A few years ago, a contractor I knew was bidding a giant office network with over 10,000 fiber optic connections. He had plenty of experience in fiber optics and knew how to install the job properly. He could also estimate costs accurately.

One of his suppliers told him of a new type of fiber optic connector that could be installed much more quickly than current designs. A salesman and application engineer visited him in his office and showed him how their connector could be installed in less than a minute, instead of the 5 minutes he was used to. He was sold on the idea.

He submitted his bid based on these new connectors. He bought enough termination kits for his crew and quantities of connectors. But when he got into the field, he was in for a rude awakening.

The connectors were installed easily and rapidly. His installation times were much shorter, although absolute time savings were lessened when setup and knockdown times were included.

But when they started testing these connectors, only one-third passed loss testing. Most connectors they installed had losses that were unacceptable to the customer and would not work in a typical network. Before he was through, his crews ended up making three times as many terminations as they had estimated, and the manufacturer had to come help him solve their problems.

Sometimes similar problems can occur with field-proven products, but they are new to the installer. I know another situation where a contractor installed a large fiber optic network on a government facility. It was a single-mode fiber optic system, and he had only done multimode fiber before, but he assumed both were installed the same way. He also had to test with an OTDR (optical time domain reflectometer), an instrument his crews were not familiar with.

This became a classic example of the old adage, "Two wrongs do not make a right!" The single-mode fiber was installed and terminated just like multimode fiber, which led to high losses at every connector and splice. And when they got their OTDR, the salesman trained them. Unfortunately, all he knew how to do was turn it on and hit the "autotest" button.

When they started testing, the results were garbage, but they did not understand that. Their conclusion was that all the fiber optic cable was bad. By the time someone who knew what was going on showed up, they had ripped out over half of the installation and dumped it in a mud puddle on a nearby field.

The losses incurred by these two mistakes approached $100,000 each.

What did these contractors learn from these situations? Never use a new product for the first time on a big job. Never take on a project unless you really understand what is needed and have been properly trained. And a salesman's claims and demo are no substitute for personal experience and quality training.

A few years ago, after almost 20 years as a manufacturer and trainer in fiber optics, I started developing a training program in Cat 5 cable installation. Being used to fiber optics, where everybody has good application notes and training materials, I asked Cat 5 suppliers for material I could use in my course.

I got two consistent reactions. Those manufacturers who offered training for their certification programs all told me their material was proprietary. The others told me "everybody knows how to do that."

"If that is true," I asked, "why does one read that between 80 and 90 percent of all Cat 5 is improperly installed and won't meet specs?"

After much investigation, I concluded that this assumption was part of the problem. What "everybody knew how to do" was install telephone cabling. It looked a lot like Cat 5, but if you installed Cat 5 like you installed telephone cabling, it would ruin the performance of the Cat 5.

Whenever you are dealing with new products or processes, high tech or low, make absolutely certain that you know exactly what you are doing. Start by getting everybody trained and let them experiment in the office until everyone is comfortable with it. Never experiment on your customers; you both may end up being losers.

HAYES, a frequent contributor to Electrical Contractor, is president of Fotec in Medford, Mass. He has written widely on fiber optics, including The Fiber Optics Technician's Manual. He can be reached at jeh@fotec.com.