Fowler Associates Labs

Monitoring Your Monitors

Your ground monitors may be fooled by their own sensitivity and the "human-battery effect." The result is that your wrist straps may be better than you think.

Edward H. Russell
Quality Assurance Specialist, Santa Barbara, CA

A vital part of an ESD-control program is grounding workers via wrist, heel, or foot straps to electrical contact between ground and the worker. With personnel grounders, voltages on workers are reduced to levels safe for ESD-sensitive devices.

But simply put using grounders isn't enough. Workers must be taught to use them properly, and grounder performance should be continually monitored to assure that resistance to ground stays within acceptable limits. Unfortunately, if improperly understood, monitoring can lead to what seems like high rates of grounder failure. In reality, some facilities are probably rejecting wrist straps needlessly.

Background

During a recent study of wrist straps from more than 30 electronics manufacturers, ESD consultant Roger J. Pierce spotted a number of deficiencies. He properly emphasized the importance of frequent monitoring to ensure the integrity of wrist straps and grounding cords worn by the worker.

However, engineer Betty Smith found that some monitors were better suited than other to given applications.

It is now clear that high failure rates, normally thought of as due to defective or poorly fit wrist straps, can also be the result of test-method anomalies. Effective wrist straps may be failing because monitoring instruments are giving false alarms.

In our case, an excessive number of failures during wrist-strap tests at General Dynamics (Pomona, CA) led to a study that uncovered corrective actions that could prevent future rejection of acceptable wrist straps.

All failures occurred using testers that measured an overall series resistance consisting of: the contact resistance between the strap's cuff and the person's skin, the resistance of the grounder's current-limiting series resistor, and the resistance of the ground cord.

The testers were made by various manufacturers but operated on much the same principle. Powered by a 9-V battery, one of three lights on the tester indicated whether series resistance was less than 750 kOhms greater than 10MOhms , or between the two values. The testers had been previously checked for accuracy and met manufacturers' specifications. All failing wrist straps had an expandable fabric cuff and were from the same manufacturer. The length of time the wrist straps had been in service ranged from one week to six months.

The most common failure mode was indication of an overall series resistance exceeding 10 MOhms . Replacing the failed wrist straps with new ones resulted in a pass condition at some workstations.

The wrist-strap-and-cord combination was then removed from the worker involved and its resistance measured using an ohmmeter. Measuring from the fabric portion of the cuff to the end of the ground cord, all of the grounders measured approximately 1 MOhm, that is, near the value of the current-limiting resistor. This was acceptable performance. At this point, the testers were rechecked and they were found to be operating acceptably too. Why, then were the wrist straps "failing?"

Human Batteries

This isn't the first time such confusing results have surfaced. In 1982, James R. Huntsman and Donald M. Yenni, Jr. found similar false-resistance measurements when using low-voltage ohmmeters. Huntsman and Yenni described a combination of galvanic action and human biopotential effects that resulted in a small flow of current from a person dependent on the wrist-strap cuff material and which resulted in false-resistance readings.

Measurement errors due to this so called human-battery effect, according to Huntsman and Yenni, could be reduced significantly by using ohmmeters with 20- to 30-V power supplies. And the effect itself could be reduced by assuring closer contact between a worker's skin and the grounder's cuff.

To determine whether the human battery effect was the cause of the failures in our case, measurements were conducted on ten of the failed wrist straps with a Triplett Model 630 ohmmeter which uses a 30-V battery. The ohmmeter showed as low a 1 MOhm total series resistance when worn by some people and as much as 6 MOhms on others. In each case, the associated 9-V tester had indicated grounder failures. Apparently the human-battery effect was at work.

Adjustable fabric-cuff straps were substituted and adjusted snugly; causing the failure rate to drop slightly but remaining high. In quest of lower contact resistance, an expandable metal cuff was substituted; it measured acceptably on all ten people.
These tests results supported Huntsman and Yenni's findings about the human-battery effect, and the investigation was carried further in a special study.

For the study, thirteen assemblers were selected whose skins ranged from fair to dark complected, from dry to moist, and who offered various amounts of hair on their wrists. The assembler's body weight also ranged widely. The selection was eleven females and two males.

Humidity ranged from 43 to 58% for the four-week testing period. The group agreed to wear a specific wrist strap and ground cord unless they became defective. All straps were of the expandable fabric-cuff type, and were made by the manufacturer whose straps had failed previously. For comparison, a new expandable fabric cuff, a new expandable metal cuff, and a new adjustable fabric cuff were later substituted and their resistance measured on each assembler.

The overall series resistance of each assemblers' skin, wrist strap, and ground cord were measured and recorded weekly. Measurements were made using two 9-V testers and the Triplett Model 630 ohmmeter. The results are shown in Table 1.

When reviewing the data, it became obvious that the only real failure was a broken ground cord; all the other wrist strap failures proved to be false alarms when double checked with the Triplett 630 VOM. The adjustable fabric cuff showed a small improvement because the human-battery effect was reduced when slightly more intimate contact was made with the skin than had been the case with the assembler's original wrist strap. The majority of the assemblers' straps were snug, and some were pushed up the arm to obtain a snug fit; however, they still had "failed" despite these efforts to assure contact.

Some of the fabric cuffs were adjusted so that the metal snap area contacted the inner side of the wrist where there was less hair; all measurements of these straps were acceptable. The expandable metal cuff passed all tests.

The conclusion was that all wrist straps used in the study were acceptable, but that the test method was flawed. To deal with the ambiguity, immediate corrective action was replacement of all fabric cuffs with expandable metal cuffs until a more reliable test method could be decided upon.

The study's results also created a concern for the human-battery effect using popular constant monitors. After analyzing several different types of these monitors, the following conclusions were made:

1. Resistance-type monitors operate similarly to the 9-V testers- because they are low-current devices, false alarms can occur.

2. Capacitance and impedance-type monitors are low-current devices also, but are much less sensitive to the human-battery effect.

3. Capacitance and impedance-type monitors, however, have their won problems. False alarms can be triggered by capacitance changes such as those occurring when workers lift their feet off the floor or move to a standing position form a sitting one. Some manufacturers have already corrected this problem.

4. The upper alarm point is set at 3 to 5 MOhms on some monitors and at 3 to 10 MOhms on others. Thus, false alarms can occur if a small person with very dry skin is being monitored. Such a person's skin resistance is typically near the alarm point and yet still acceptable. These false alarms can be eliminated or reduced if the upper alarm limit is adjusted to 6 to 10 MOhms.

So the question "How does your wrist strap measure up?" is not a simple one. If you experience a high rate of grounder failures, it could be caused by the test or monitoring method.

Guidelines

A high-quality fabric-cuff wrist strap should provide many hours of service if cared for properly (e.g., by wearing the strap at all times during working hours to reduce wear and unnecessary cleaning.)

If a fabric cuff is preferred, the adjustable type is recommended, and to ensure intimate skin contact, adjust the wrist strap so that its metal snap plate lines inside the wrist.

Creams designed to lower skin resistance when applied to the skin will only improve contact for a short time and are thus not a long-term or complete solution to false-alarm problem.

Expandable metal bands will provide longer service than fabric grounders, will test accurately on low current testers, and minimize false alarms during monitoring.

For More Information

For more information see the following:

J.R. Huntsman, D.M. Yenni, Jr., "Tests Methods for Static Control Products," 3M Co. Presented at the 1982 EOS/ESD Symposium, Orlando, FL.

EOS/ESD Standard No. 1, Standard for Protection of Electrostatic Discharge Susceptible Items: Personnel Grounding Wrist-Straps.

Betty Smith, "Constant Monitoring: Quality You Can't Afford to be Without," General Dynamics, Pomona, CA. Presented at the 1987 EOS/ESD Symposium, Orlando FL.