Tekscan Flexiforce sensors
frank freedman
Posts: 1,975
Has anyone out here used the Tekscan FlexiForce pressure sensors? If so did you run into any sort of gottchas or unexpected results? I am considering using these as force/balance sensors for an assistive technology project. Thanks to all.
Comments
I do know, not directly again, but I do know enough about them. About the only ones that are available are that of connecting them. After that I believe that they can be used the same as any variable resistor would be used.
I did use flexiforce sensors upon a time to help out a biologist who was studying small rodents (kangaroo rats) out in the desert. The idea was to place lots of trays around in different types of habitat and to monitor the visitations. Why do that, you will have to imagine. But it had to be relatively cheap with lots of food trays, which doubled as scales. The trays were made in the shape of a triangle, with a flexiforce dot under a silicone pad at each corner. Unlike your request, the goal here was to check for changes of total weight, not to check for shifting around. The fact that the electrical resistance was inversely proportional (roughly) to the force was helpful in adding up the distributed forces to find the total weight on the tray.
The pads are fragile against bending, so you'd have to be careful with the platform, and watch out for the load limits and drift. The rats were at the low low end of what could be detected, but 200# spread around on several sensors should work okay. The product offerings were different then, and far less expensive than what I see on the tekscan web site. They do have a lot of design notes there. Parallax sells a demo kit.
Given the price of these at Mouser, the TE units tubular mentions may be the way to go. Those have an odd (to me at least) readout value given in mV/V, which I would have to guess the x mV output would be determined by the applied source voltage. @$30 each, they may prove easier to work with than the FlexScan sensors. Probably I will play some more with the FelxScan and maybe get a couple of the TE units.
Edit:
It may be that since I don't care about the absolute value from each sensor, possibly I could set all four in place under the load, allow for a settling time followed by a reference acquisition for a baseline to compare against over time.
Your guess is correct. Typical load cells are strain gauges connected to form a Wheatstone bridge. Typically an AC voltage is connected to one pair of opposite contacts and the signal measured between the other pair. This is used in scales for a range of loads from micro grams to kilo tons.
PS - There are front end (analog + adc) chips available for working with strain gauge load cells.
The TE strain gage bridge would be far less ambiguous, more quantitative.
You might have an option with inductive sensing, using something like the TI LDC1000, or similar idea using tricks with the Propeller. It could involve the inductance of springs that hold up a platform, or anything really that affects inductance as the platform shifts or flexes. (Think like forumista @Beau Schwabe)
Next up is interface electronics and mounting to frame feet.
Edit: These are 0-100# sensors, plan to use four though may go to six for the mid-line of the platform.