Sensors Mag

Check It Out!

September 23, 2008 By: G. Raymond Peacock, Temperatures.com Inc.


E-mail Ray Peacock

This is the ninth essay in a series expanding on an article I wrote for Sensors titled "A Twelve-Step Sensor Selection Checklist". In July, I talked about acquiring the sensor and the expert technical services you require. This month, I'll focus on making sure you get what you asked for.

Things Happen

Taking steps to ensure you get what you ordered seems like a no-brainer. After all, when you finally get the sensor you've spent so much time and effort specifying and selecting, you'd think you would make certain the device meets your specifications. But not everyone does, and therein lays the problem.

In my experience, there are a few dozen things that can go wrong between the time you order and when you finally take delivery. I used to work for a premier sensor maker. Its manufacturing plant was in England, and we did sales and services in North America and Mexico. We took every precaution—including an incoming inspection and calibration test—before we shipped the goods to the end users. But even with all our efforts, something went wrong on occasion.

Belt and Suspenders

Keep in mind there's a lot at stake. Modern, high-efficiency, high-speed processing can make a lot of scrap in a very short period of time. If your sensor is not right, it could end up costing far more than the expense of testing. That's why it is essential that you inspect your sensor on delivery. Be sure you get the right device and that it meets your specifications.

I know that seems to go against the principles taught in SPC and Six-Sigma programs, but one-shot events are outside the scope of such disciplines. A new sensor is a one-time event, and a very critical one that needs to be 100% within specification when it's used. The responsibility for the deployed sensor's performance is yours, not the supplier's.

Checklist for Success

So what precautions should you take? First, make sure the new device meets the promised specifications. Assume nothing.

Unfortunately, proving that the new device works correctly is not always easy. You may need some special tools or simulation devices to mimic the process variables and conditions of your measurement. If you're unsure of how to perform the tests, start by asking your vendor how to evaluate the device. The vendor is the expert, and most often, its staff will be eager to help. After all, they want the sensor to work. Besides, they usually don't get paid for the device if you do not accept it.

So you hold the reins of power. Use them wisely. Make friends among the people who supply you, who you must satisfy in your plant, and who make up the QA groups. Document what you do, and keep a record of the tests. It's likely to be the first of many records that are developed for the device, as I'll discuss in a later segment of the series.

But do not blindly follow the vendor's suggestions. Take the time to understand how the test proves the calibration and the specific performance features you're concerned with. Do not hesitate to think for yourself.

For example, ambient temperature is one of the major influences on the performance of any sensor that has electronic components. If your specification includes an ambient temperature performance requirement, you'll need to perform a test on a representative portion of that range.

Low-cost heated and refrigerated water baths of variable temperature, plus a little creativity, help enable temperature tests to be run relatively easily and quickly. But be certain you allow the device to fully "soak" to a constant temperature, and record both the maximum transient and the steady state response. The latter is most often the one of concern in actual use, but both provide an excellent way of determining whether the device requires a larger heat sink/cooler to cope with rapid changes in its environment.

If you do an ambient temperature test carefully, you can also collect a series of time-temperature response data—assuming step changes in temperature—that enables you to calculate an average response time constant for changes in ambient temperature, as well as verify the basic performance in the range of test values. Similar considerations apply for tests on vibration, ambient humidity, smoke, fumes, and other environmental conditions. Many can challenge your ability to devise effective tests.

If the vendor's suggestions, your imagination, or experience do not help you create adequate tests, get a third party, measurement consultant, or independent testing laboratory to advise you or even help perform some of the tests.

The American Society for Testing & Materials International has a free searchable online directory of consultants and testing laboratories. They are great resources, and often you can get professional advice for little or no cost.

The responsibility is yours. I can't say it too often—test, test, test!


Add Comment




IIoT University


Deep Learning for Vision Using CNNs and Caffe: A Hands-on Tutorial – 9/22/16 – Cambridge, Mass


IDE






Sensors 2017 Call for Speakers


Sensors Midwest


Advertise


Subscribe



Twitter Feed

Find It Fix It Forum

Sensors invites you to join the Findit-Fixit Forum, where you can get answers to your sensing questions—concerning technologies, products, methods, applications, and services--and also offer help to your fellow engineers. The Forum covers all kinds of topics, from the basics to the extraordinary.

Join the discussion!


© Copyright 2016 Questex, LLC. All Rights Reserved. Sensorsmag. Privacy Policy | Terms of Use

If you are having technical difficulties or considerations, please contact the webmaster.