Don’t fall in the marketing trap! What most companies won’t tell you about specsmanship.

Q: You're surfing the web for a low-cost accelerometer. Both IMI and their competitor basically offer the same item, it is even marketed the same, low-cost permanent mount is the category. Yet one sensor has an 80 g acceleration range and the other says 50 g’s It’s the same technology, same sensitivity (100 mV/g), and same supply voltage (18-30 VDC). So why would one be 80 g and the other 50 g?

A: Specsmanship. IMI has always had the policy of "spec’ing" for worst case scenario. Other companies do not…and frankly they do not tell you that.

The IMI sensor has max bias voltage of 12 VDC and min supply voltage of 18 VDC. The dynamic range is supply voltage - bias - 1 volt (it gets "eaten" by the electronics). Thus, worst case, you have 5 volts to use for your signal (18-12-1 = 5 V). So a 100 mV/g sensor has, worst case, 50 g range. Hence the IMI spec.

The competitor’s sensor has max bias voltage of 14 VDC - by the way you're not in control of that, what you get for bias is what you get…for both companies - and min supply voltage of 18 VDC. Thus, worst case, dynamic range is 18-14-1 = 3 V. So with a 100 mV/g sensor you’d have 30 g range. But they spec 80 g on the assumption that:

-          you probably will not send the sensor the min voltage

-          you probably will not get stuck with a 14 VDC bias from the factory

-          you probably will not ask them about it

-          and you probably will never notice

Many companies play games with specsmanship and not just in measurement range either. It happens all the time in a fierce market where each manufacturer is trying to get that little edge. Frequency response is one of the most liberal specs and often isn’t tested just assumed or calculated. An example is provided at the bottom of the page.

We ran a calibration on a competitor's sensor and found it to be out of spec. The sensitivity of the model is supposed to stay within 10% of 100 mV/g all the way through the frequency span to 600,000 cpm. If you click on the download at the bottom of the page you will see the sensor had greater than 10% deviation before its specified high frequency limit. The deviation at 600,000 cpm was almost 80%, producing a sensitivty of 180 mV/g!

If you should have any questions on any sensor spec please call one of our Application Engineers – all certified vibration analysts through Vibration Institute. We buy many high end items on specs and with sensors it is no different. It is important to know what the specs mean and when something just doesn’t add up.

Remember with IMI Sensors your model is guaranteed to meet spec. We give you the worst case scenario. We think it is a more honest and realistic approach.

If you want to learn more about the effects of bias voltage on range we have it on our site: http://www.pcb.com/techsupport/tech_signal.php (scroll halfway down the article).

Thanks again for working with IMI!

Calibration Certificate - Failed Competitor's Sensor
This calibration certificate shows a competitor's sensor failing a frequency sweep test at the high frequency limit specified on their spec sheet. The sensor is supposed to stay within 10% of 100 mV/g to 600,000 cpm. The cert shows nearly 80% deviation at 600,000 cpm.