Dr. Ronald B. Standler
Consulting on Surges & Electric Power

Copyright 2000-01, 2003 by Ronald B. Standler

Table of Contents

My brief credentials
Electrical surges
Outages and other "power quality" problems
My experience with protecting computers
My usual recommendations
Conclusion

Dr. Standler's brief credentials

Ph.D. (physics) 1977, J.D. May 1998.
Professor of Electrical Engineering for 10 years.
Author of one book and more than 35 published papers in science and engineering.
My c.v. has full details of my experience and publications.
Attorney in Massachusetts, who has an international practice of consulting to litigators on scientific evidence in torts, especially damage or injuries by lightning or electrical surges.

synopsis of my experience in research on surges:


Electrical Surges

Electrical surges are transient overvoltages that are conducted on wires and cause damage to computers, modems, electronic equipment, and even electrical motors. Such damaging overvoltages are commonly caused by lighting or switching reactive loads. These overvoltages are transient events: they typically have a duration of less than 0.001 seconds, never more than a few milliseconds, which is a fraction of one cycle of the ac mains voltage. Surge-Protective Devices (SPDs) divert surge current away from vulnerable equipment and are the principal means of protecting equipment from surges. SPDs are divided into two classes: (1) surge arresters, which are high-energy devices installed near the point where wires enter the building, and (2) surge suppressors, which are low-energy devices installed near the equipment to be protected.

other problems with electrical power

There can be considerable economic damage from interruptions of electrical power (e.g., blackouts, flicker, outages). For small loads (e.g., a personal computer), use of an uninterruptible power supply at the customer's site is a simple solution. For large loads, a diesel-powered generator at the customer's site can provide electrical power during the rare times that utility power is not available.

There are a number of other "power quality" problems, such as distortion of the sinusoidal voltage waveform, disturbances of the rms voltage (e.g., brownouts, temporary overvoltages), etc. The solution to these problems is to use appropriate technology, such as line conditioners. I put the term "power quality" inside quotation marks, because it is a misnomer: the real problem is with disturbances in the voltage waveform, not power. Nevertheless, "power quality" has become a common term among engineers and technicians who work with delivery of electrical power to customers.

my experience with protecting computers

I have operated eleven personal computers at my home since 1981, accumulating more than 77000 hours of operation, with neither lost nor corrupted data or program file(s), and with no damage to hardware that is attributable to electrical disturbances or surges. This record is even more impressive when you realize that I routinely operate during local thunderstorms, when the electric lights flicker.

I presented a technical paper at the October 1988 Annual Meeting of the Institute of Electrical and Electronic Engineers (IEEE) Industry Applications Society that described how to use a line conditioner and uninterruptible power supply, together with surge suppressors, to protect a personal computer system. This information is repeated in Chapters 19 and 20 of my book, Protection of Electronic Circuits from Overvoltages.

My 1988 paper presented a summary of my operating experience to show that my protection techniques were effective. The following is an updated listing of my experience with computers in my home, to 8 Nov 2003:

total
dates of operating
Computer System operation hours
Hewlett-Packard 9825 Mar 1981 - July 1986 unknown
North Star Advantage Sep 1983 - June 1986 about 1300.
Hewlett-Packard model 236 Sep 1984 - July 1995 > 3900.
IBM PC/XT Apr 1986 - July 1991   5757.
AST '386 machine July 1991 - July 1996  12342.
Apple IIci May 1992 - May 1996   2542.
Austin Pentium machine Dec 1995 - Oct 1998   4509.4
PowerWave May 1996 - July 2001   2421.
Hewlett Packard VL5 June 1997 - current   9234.9
Hewlett Packard Kayak XA Feb 1999 - current  22914.0
Apple G4 Feb 2002 - current  12748.5


During this experience, I have had only a few hardware problems:
  1. several defective 5½ inch floppy drives on the North Star Advantage
  2. one defective 5½ inch floppy drive in the IBM PC/XT
  3. defective RAM integrated circuits in the IBM PC/XT in Dec 1990 and June 1991
  4. power supply failure in May 1997 in the Austin PC.
  5. failure of Seagate hard drive only 3 days after I received my Hewlett-Packard Kayak workstation
  6. floppy disk drive failure in the Hewlett-Packard Kayak workstation after 40 days of use
The failures of the floppy disk drives were probably due to manufacturing or design defects, rather than power problems at my site. The floppy drive failure in my Kayak workstation was definitely mechanical: after the failure, it was difficult to insert a diskette into the drive and the eject button did not come all the way out of the computer case when the floppy diskette was inserted. When I replaced the floppy drive in the IBM PC and HP Kayak with a Teac drive, I had no further problems. I used a Teac 5½ inch floppy drive in my AST machine for more than seven years with no problems.

The power supply in the Austin had been failing intermittently for at least eight days before it permanently failed, and none of these failures occurred during local thunderstorms, so I suspect that it was merely a defective product. The computer worked fine after replacement of the defective power supply: there was no damage to the microprocessor, RAM, hard disk drive, etc.

I emphasize that I have never lost or corrupted a data or program file on any of these machines at my home, despite the fact that I routinely operate during local thunderstorms when the electric lights flicker.

In addition to the experience in my home, I operated a Hewlett-Packard model 310 computer to control instruments that collected data on electrical disturbances at 14 different trouble sites in Pennsylvania during June 1987 to November 1988. This computer accumulated more than 7970 hours of operation at these sites, plus more than 3000 hours at my laboratory at The Pennsylvania State University during May 1989 to May 1990. The only hardware failure during this time was the failure of a sealed Seagate hard drive inside a Hewlett-Packard model 9133H disk drive in November 1987. This failure may have been due to a disturbance on the ac supply mains, or it may have been the result of sabotage (e.g., mechanical shock or vibration) by people at the site who were angry with the electric utility that sponsored my research. The power supply in the disk drive enclosure was not damaged, only the hard drive was replaced, which suggests that the problem was not the result of disturbances on the ac supply mains.

my usual power conditioning recommendations:

  1. Surge arrester at the main circuit breaker panel in the building. I prefer a metal-oxide varistor with a minimum diameter of 40 mm and a maximum continuous operating voltage rating of 125% of the nominal line-ground voltage (e.g., for applications where the nominal line voltage is 120 V rms: 150 V rms).

  2. Optional ferroresonant line conditioner. This apparatus regulates the rms line voltage and removes some disturbances in the voltage waveform. I recommend a ferroresonant conditioner for environments in which there are either: Some engineers and technicians have a strong prejudice against ferroresonant transformers, and I agree that some models are not satisfactory. However, I have had good experience operating four different models of computers from ferroresonant transformers. The trick is to choose both the proper design and the proper power rating of the transformer.

  3. Surge suppressor. There are two basic designs of a surge suppressor that I can recommend:
    1. A low-cost surge suppressor with metal-oxide varistors that are rated for service at least 1.4 times the nominal line-earth voltage, and preferably twice the nominal line-earth voltage [e.g., the varistor should preferably have a maximum continuous operating voltage (MCOV) of 250 V ac for use on nominal 120 V ac service]. One varistor should be connected between the line and neutral conductors, a second varistor should be connected between the line and earth conductors. An optional third varistor may be connected between the neutral and earth conductors.

    2. A combination surge suppressor and low-pass filter. Such a device may contain metal-oxide varistors with a 20 mm diameter that have a maximum continuous operating voltage (MCOV) rating at least 1.25 times the nominal line-earth voltage, provided that there is an inductor (e.g., at least 20 µH) connected between the ac supply mains and varistors inside the suppressor, to help coordinate these varistors with the upstream surge arrester.

    A surge suppressor must also contain protection from surges on the telephone line when either a computer with a modem or a telefax machine is plugged into the surge suppressor. Using a common ground connection for diverting surges on both the ac power and telephone lines is essential; the most convenient way to do this is to use the ground wire in a properly grounded three-prong wall outlet. I prefer a bidirectional silicon thyristor between each telephone wire and ground, and preferably a third silicon thyristor connected between the two telephone wires (i.e., line to line).

    Some surge suppressors that contain metal-oxide varistors can be a fire hazard, although the probability of fire is less than the probability that a surge could damage unprotected electronic equipment.

  4. Uninterruptible power supply, which uses batteries to supply power to the computer during failure of utility power with a duration of less than five minutes. If an outage lasts longer than one minute, the outage will probably last for a few hours, so an orderly shutdown of the computer is desirable, before the batteries are exhausted. (Of course, laptop computers, with their internal batteries, do not need an uninterruptible power supply.)

  5. Other environmental conditions are also important:
To get more information from me, including recommendation of brands and model numbers, you must hire me as a consultant.

Conclusion

Because of my scientific research experience in lightning and electrical surges, and because of my litigation experience, I am interested in consulting to attorneys for legal strategies, scientific evidence, drafting briefs, etc. in this area. I am particularly interested in assisting attorneys who represent either victims of lightning injury or businesses whose equipment was damaged by surges or other "power quality" problems.

I continue to be interested in doing engineering research and development: my change in emphasis from science and engineering to law was motivated only by the annihilation in 1990 of financial support for research and development in the USA in all of my fields of science and engineering. The end of financial support for research was caused by the end of the cold war (which killed funding of research by the military, as appropriations to the military were slashed) and a simultaneous recession (which killed funding of research by utilities).

my fees and terms

my address and telephone

my c.v.

return to my homepage

this document is at   http://www.rbs2.com/pq.htm
first posted 15 August 1998, revised 9 Nov 2003