I have 1018 - PhidgetInterfaceKit 8/8/8 (http://www.phidgets.com/products.php?product_id=1018) and four 1104 - Vibration Sensor (http://www.phidgets.com/products.php?product_id=1104) connected through the analog interface. They work as designed, but I am trying to use them in a slightly non conventional way. I have roughly a 4 foot by 6 foot 1/4 inch thick rubber mat hanging (6 foot span up and down) on the wall in my garage with the four sensors attached in the approximate center top, center bottom, center left and center right. My intent is to pinpoint an impact (my kids are throwing baseballs at the mat) by measuring the difference in time from one sensor trigger in relation to the other three. It appears that the interface kit can poll on the millisecond but it does not expose the clock time through the API. However I need to work with nanosecond time for reasonably accurate measurements. I tried to pull nanosecond time from the computer but I am getting inconsistent results.

So I have a few questions:
Am I using the wrong sensors and/or interface kit?
Is there anything out there that works with nanosecond precision?

Nanosecond precision is going to be far too fast for any embedded system to handle. In order to measure in nanoseconds your processor needs to operate at a much higher clock speed than any of our products do.

I think that you would have better luck using a 1046 and 4 load cells. If you connect each load cell to a separate port on the 1046 and you should be able to read the value off of each load cell. That will let you find x/y components for each of the cells and ultimately be able to calculate the position of the ball.

_________________
Brian Burley
403-282-7335 ext. 6003
support@phidgets.com

The 8/8/8 samples data form it's inputs at a stable rate.

The speed of a wave through rubber is ~60m/s, so this should give you an accuracy of ~6cm/ms.

It should be easy to correlate the separate sensors into discreet time events - if you use 1ms datarate with sensitivity 0 for all four, then you will get events in this order every 8ms:

sensor0, sensor1, sensor2, sensor3
sensor0, sensor1, sensor2, sensor3
sensor0, sensor1, sensor2, sensor3
sensor0, sensor1, sensor2, sensor3
sensor0, sensor1, sensor2, sensor3
sensor0, sensor1, sensor2, sensor3
sensor0, sensor1, sensor2, sensor3

Each line represents a different 1ms time slice, but all of the data is sent at once, every 8ms.

You should be able to track the time of each 8ms chunk to tell them apart.

The bridge only has a maximum datarate of 8ms, so I'm not sure how useful it would be for this.

-Patrick