Accelerometers

[Micko..]

Hi guys,

Just wanted to ask a question about accelerometers. We are looking around at what is available at the moment and I am seeing a pretty big range of quoted accuracies, from 2% to 10% (usually associated with a respective price range) I wanted to know if the error associated with most accelerometers was random or steady state (i.e. could be rectified post sample). I figure that if it is random then a 10% difference in results could be quite frustrating.

Does anyone have issues with chassis vibration affecting their acceleration logging or do frequency filters rectify the issue.

Lastly does anyone have any pcb mounted 2 axis accelerometers that they would recommend I have had a fair bit of a look around (analog devices, Davidson etc) but any suggestions of ones that have been used successfully would be greatly appreciated.

Cheers

[Micko..]

Hi guys,

Just wanted to ask a question about accelerometers. We are looking around at what is available at the moment and I am seeing a pretty big range of quoted accuracies, from 2% to 10% (usually associated with a respective price range) I wanted to know if the error associated with most accelerometers was random or steady state (i.e. could be rectified post sample). I figure that if it is random then a 10% difference in results could be quite frustrating.

Does anyone have issues with chassis vibration affecting their acceleration logging or do frequency filters rectify the issue.

Lastly does anyone have any pcb mounted 2 axis accelerometers that they would recommend I have had a fair bit of a look around (analog devices, Davidson etc) but any suggestions of ones that have been used successfully would be greatly appreciated.

Cheers

[drivetrainUW-Platt]

We have a Tazzo Motorsports acc. that is susposed to be one of the most accurate ones, not sure if thats what you are looking for since this is stand alone and does 0-60, 1/4 mile, g's ect for a street car with no external outputs for data logging

[J. Schmidt]

A while back Car and Driver did a comparo of various acclerometers available to the average consumer. It can be found here:
Do Inexpensive Performance Meters Work? (http://www.caranddriver.com/article.asp?section_id=4&article_id=9380)

[B Lewis @ PE Engine Management]

We have used Analog Devices PCB mount accelerometers with good luck. Generally, you can improve on the listed accuracy by calibrating them yourself.

[Micko..]

thanks for that guys,

excuse my ignoance brian, but how do you calibrate them? i was thinking on a vibrating table rig but i am guessing there is a better way.

cheers

[alfordda]

Depends on what type of accel it is. If its a DC/capacitive type of accel (typically used in low frequency aplications, think vehicle dynamics), you just need to flip it over and record the voltage change, because it is exposed to Earth's gravity. But if its a piezoelectric accel (typically used in higher frequecny applications, think vibration tesing or modal analysis), you can use a hand held calibrator, which vibrates at one frequency and g level, and you measure the voltage output from the accel. There is also a drop calibration method for this type of accel, but its been so long since I have used it that I don't really remember all the steps.

[B Lewis @ PE Engine Management]

Many of the "chip" accels are DC so they can be calibrated using the "flip cal method" like Dan suggested. Another alternative is to use a known good accel as a reference and shake both transducers at the same time. Then use the known good accel to calibrate the new one.