ESD Journal - Beyond the lab

Four ESD control veterans share some thoughts on
setting up and running a test program.

Donald Ford
Publisher/Editor, EOS/ESD Technology

After listening in on the salesmen's presentation, this writer asked how garments are selected? "I rely heavily on test data," said Melissa Feeney of TRW, "when buying garments, or any kind of ESD equipment for material, for that matter."

Her comments underscore a commonly held view about the importance of testing. Tests are done to select a material and then to periodically evaluate its performance. Materials are judged and compared based on the following measurable properties: tribocharging, surface or volume resistivity, static decay time and voltage suppression.

Because test labs are important and expensive, EOS/ESD Technology invited four ESD-control veterans to share some of their thoughts on setting up a test lab.

The question of what types of equipment should you buy and why arises. Each person in the forum offered a list of basic equipment and an explanation on why each piece of equipment is important. In addition, they added, here and there, some interesting user notes. Generally, equipment is not identified by company name because some of the participants were concerned about lawsuits. Also, the comments made are personal opinions and should not be considered company policy.

The lists differ. Members of the panel interpreted the idea of "basic" in different ways. But at a more significant level, they learned that there are different ways to do their jobs and different ways to do tests. (These ideas are explored at the end of the article.) Their suggestions should provide a beginning for building a lab or, for those with a lab, a way to compare your approach with four interesting colleagues.

John Kolyer of Rockwell International made the following comments. "I approach testing and a test lab from the material-science approach, rather than the electronics point of view.

"First of all, I recommend you buy a surface-resistivity meter. It sits on a surface and measures surface resistivity at about 30 volts or so. Some people would say these testers have limitations, but the results, in my opinion, are good and quick. For example, when used on a plastic surface, instead of a solid surface, the questions about reliability arise. But, sit the meter on some pink poly and it will read 10¹². Showing there's some feeble current flowing over the bag.

"The next item you will need is a megohmmeter. This instrument doubles as a power supply and goes from 10 to 1000 volts, with 500 volts the current popular level of measurement. With this meter, you can charge up a one-foot-square plate and calibrate your field meters at 1000 volts.

"Incidentally, some people object to hand-held field meters, but I like the tuning-fork type. Calibrated with 100 volts on the plate, these instruments are good enough for most practical purposes...If you worry and split hairs on whether the charge being measured is 500 or 2000 volts, then you are not going to use a handheld meter. If you're checking for orders of magnitude, the handheld meter does the job.

"The good thing about the tuning-fork type of handheld meter is that it does not charge up in a field, such as from an ionizer. You can even check your ionizer points with this meter.

"To use the megohmmeter to measure static-dissipative tabletops, a NFPA99A electrode will be needed. The electrode is a 2 1/2-inch-diameter aluminum foil surface with a layer of rubber and then a five-pound weight. With six electrodes on the tabletop surface, resistance to ground is measured.

"Your contact with the surface can vary the readings: smooth tops give lower surface resistance readings, all things being equal, than a textured surface. In my company, tabletops are checked annually.

"There are a number of divice-sensitivity testers on the market for ranking devices into their classes. Buying one of these is important.

"I recommend testing packaging. We use our own test for this. We put a MOSFET in a bag, and then, with a high-voltage power supply and a pair of electrodes, we zap it at 1500 ohms.

"Last of all, you will need a continuity checker and a wrist-strap tester. Of course, I'd also recommend the continuous monitoring of wrist straps. The continuity checker will test wrist straps, and the wrist-strap tester can be used a a megohmmeter.

Betty Smith is a senior engineer at General Dynamics. She had these suggestions. "My equipment can be used in a lab setting or as a moving lab. If I want to, I can go to the manufacturing floor and do practical tests. In fact, some test equipment stays in the plant.

"I put continuous monitors at the workbench to check wrist straps, the most important item in our program. This eliminates doing the work in the lab.

"In the lab, I test the continuous monitors. We calibrate the monitors with a 2.5-megohm maximum and 275 Kohm as a minimum. This approach eliminates the need for a 9-volt or the 20-volt tester, which I put a big cross-on.

"Work-surfaces, which we test every 60 manufacturing days, are the next most important items to be tested on the manufacturing floor. For this test, I use a Biddle megohmmeter with one five-pound weight. The biddle megohmmeter is also used to measure walking surfaces- conductive floors, floor finishes and even rugs. In this case, you use two five-pound weights, one meter apart, and put 100 volts through the system.

"Still out on the floor, I measure the balance of ionizers with a charge plate monitor. For the bench-top ionizers, I recommend an ionometer. They are portable and easy to use.

"To test garments, I use the EOS/ESD standards as a guide. You'll need a high-voltage power supply, and electrometer and a lot of little clamps, probes and Teflon plates.

"Of course, back in the lab, we keep an oscilloscope and a surface-resistivity meter for testing bags, boxes, DIP tubes and other items with first purchased.

These are the comments made by Melissa Feeney, a member of the technical staff at TRW Inc."I've proceeded based on the point of view that materials need to be tested in a lab, on the manufacturing floor and, sometimes, outside of my plant at a vendor's place. So, some of my equipment needs to be portable.

"Equipment should be selected based on what properties need to be tested. We look at rate of charge generation (tribocharging), rate of charge removal (surface or volume resistivity), length of time to dissipate charge (static decay time) and voltage suppression. We also test charge attenuation in shielded bags.

"We use published standards as guides in doing our tests. For example, we use standards such as EIA IS-5-A for tribocharging and charge attenuation, and ASTM D257 (the latest version) for surface resistivity and volume resistivity tests. In addition, we have established test limits as follows: less than 1.4 nanoseconds for tribocharging, 10¹²W per square for surface and volume resistivity, tow seconds maximum for static decay time, and less than 100 volts for voltage suppression.

"In addition, we have modified some for the basic equipment. Some tumeric (silver-filled) rubber was put on the bottom of the NFP 56-A weight, so I don't have to mess with the aluminum foil or any of the rubber. Now, I don't have any contact-resistance problems and there aren't any residual black spots left on the surface. Secondly, the weight is goldplated, as well as the alligator clips on the electrometer. This gives me good contact.

"I believe that in the lab you should have an ionizer and a dry box, for humidity control, to keep a constant environment in the test lab,when that is necessary. Also, some device for measuring capacitance is necessary so you can always go back to Q = CV relationship to see what you are really dealing with in terms or a deliverable charge off a part."

Here are the highlights of the comments made by Ben Baumgartner, Lockheed Corp. "As I put my list and reasons on paper, I realized there were areas where equipment is used to measure things that shouldn't be measured. So I started again and produced two minimum ESD testing equipment lists; one for a small, commercial plant and the other for a large plant. In the small plant, the parts are not hi-rel parts and they probably won't be doing research or analysis on the sensitivity of their product.

"First of all, they would need a high-resistance meter. It doesn't have to be fancy, just any handheld meter that gives an indication of whether the material is way up in the insulative range or in the low-dissipative range.

"The second-iten is a voltmeter, a common item. It should be capable of delivering 4.5 volts or higher at the terminals. Digital meters should not be used for this purpose because they deliver very low voltages, as low as 0.5 volts, and you can get misleading results. Also, an electronic environment might interfere with these meters. So stick with the standard, old-fashioned voltmeter.

"A third item that one should have is an electrostatic voltmeter where one knows the input capacitance and the resistance. And this is an appropriate spot to mention that I do not like handheld meters because they lack the capability to tell the capacitance for a given voltage.

"A power supply that goes up to at least 2000 volts is the last item. With this item, you can calibrate the electrostatic voltmeter and charge up materials to see how fast they discharge.

"My second list, for a large plant doing hi-rel work, includes a resistance meter with a wide range of settings, from 10 volts to 1000 volts. Similar to a small plant, a resistance meter will be needed and a voltmeter- but with a chart recorder, so you can observe and record voltage readings and enter them in your reports.

"Last of all, in order to simulate the human-body capacitance, you will need a human-body tester, an oscilloscope and a high-voltage probe. The probe, used to calibrate the tester, permits you to tell that a pulse was actually delivered and at what voltage the device broke down. You can observe the method of breakdown: was it sudden or was it slow?"

Quantification from testing offers a haven for making decisions, a safe way around the subjectivity in judging the reliability of the ESD-control materials used to improve product yields. Measurements, at the heart of all science and technology, offer a golden promise of true information.

Thus, it is disconcerting when tests results aren't repeatable, as is the case in some ESD testing. When the anticipated measure of objectivity fails, one feels duped. The following analysis of why surface resistivity measures are often not repeatable is enlightening.

Why aren't measurements more reliable? Is the equipment a problem? Are there special techniques or measuring tricks that a neophyte needs to learn? Are the measures simply not applicable to the materials? The panel looked at this issue by examining the test equipment, the testing conditions, the testing techniques, the meaning of surface resistivity, and the goals of the test.

"Surface resistivity measurements have been made for a long time in physics," Said Ben Baumgartner. "The problem is adapting this measure to the materials we use. The definition of surface resistivity is derived from homogeneous materials. But what is homogeneous? Is pink poly homogeneous? Is a foam surface that is porous homogeneous?

That's why I propose that people use a two-point resistance method and call the measurement an apparent resistance and apparent surface resistivity."

In addition to Baumgartner's two-point method, there are two other types of testers: one uses a guarded ring and another has two parallel bars. "There are three different ways to measure surface resistivity, and each way gives different results," said Betty Smith.

Melissa Feeney, who had done an extensive study on surface resistivity testers, concluded "if I get as much as an order of magnitude of repeatability, that is as good as you can expect." She had compared testers that have different distances between the guarded rings and the parallel bars, as well as equipment from different suppliers and the varied voltages put across the materials.

Kolyer suggested that numbers can't be compared unless one agrees before testing on an accepted order of magnitude based on what the number should mean. "What are you looking for? If it is a piece of polyethylene, we only need t know if a feeble current will run across it. So trying to split hairs on the exact resistance is not very productive thinking."

In addition, there is always the human equation: How much care was taken when doing the tests? Good training should overcome this problem, but it doesn't. And, did the tester intend to keep all environmental conditions, while testing, constant; or did he intend to include changing environmental conditions as part of the test?

These last question involve more than testing technique. They relate to deeply held convictions on how a testing lab fits into an overall ESD-control program. The following section records the thoughts of panel members on these questions.

Items such as continuous wrist-strap monitors and ionometers measure the reliability of ESD-control tools. These items are placed, not in an isolated lab, but on the manufacturing floor. The majority of the panel felt the materials should be tested where the materials and equipment are used: where there is changing temperatures and humidity, and people of all different sizes, shapes and attitudes.

Although an isolated test lab is one component of a testing program, the lack of confidence in any given material and the huge variation implicit in a manufacturing-floor environment dictate that some testing equipment be carried and used there.
Baumgartner, who runs his program from a lab, disagrees with the others. He has engineering specs and logic to support his view. Here is his reasoning.

"We tell engineering that our limit for control is 100 volts and then we guarantee that no device will ever see anything higher than 100 volts. No matter what voltage you might put on a worker, using ESD control tools, he will never show more than 100 volts," said Ben.

"Let me use a hot-water heater as an analogy," he continued. "If I want to be sure that the temperature of the water in a hot-water heater will never rise beyond a certain point where it could burn somebody, I'd install a controller. If I wanted to be fancy, I could even calculate heat losses and so on to take into account heat and temperature variation and have the controller work with great precision. But the point is, I could guarantee that nobody would ever get scalded.

"My ESD control system is like the controller on the hot-water heater. I calculate what voltage is possible when a worker wears a wrist strap. Now if the worker takes the wrist strap off, all bets are off; but from an engineering point of view, I can set a safe condition."

So the question about where one test involves whether one can trust the engineering of a hot-water-heater controller. And can you trust the engineering of your ESD control system? Melissa Feeney thinks she'd want to do some tests on the hot-water controller, such as checking the temperature of the water in the basin now and then, and she and the others intend to keep checking their ESD control systems.

So your degree of faith in your materials will dictate how you define your job. And that will determine what test equipment you'll buy and how you'll use it.