0
\$\begingroup\$

Sorry if my question sounds to simple for this forum. I just wanted to confirm the default measure of the output of circular force sensor when it is connected to the breadboard (with resistor)? Does it output force measured in N (newton)? When I slightly touch the sensor with a fingertip, the output value varies around 200-400.

Another question: My final objective is to measure weight categories of objects (light, medium, heavy) using 2 force sensors. Do I understand correctly that the overall procedure is the following:

Calibration:

1) put known weights on top of each sensor and measure the force

2) fit analytical function for each sensor, e.g. some polynomial

Usage:

3) put an object with unknown weight on top two sensors

4) read outputs (force)

5) apply both analytical functions (with known forces) to obtain two weight's measurements

6) calculate an average between two measurements

\$\endgroup\$
  • \$\begingroup\$ There is a link on the page you reference to FSR Integration Guide (Ingles) which looks like it would provide lots of information on the use of that sensor. \$\endgroup\$ – Peter Bennett Mar 1 '16 at 18:36
  • \$\begingroup\$ @Peter Bennett: Thanks. I have read it. Probably due to my beginner's level, I still have doubts mentioned in the question. For instance in this FSR Integration Guide they talk about Voltage measurements... But I obtain values greater than 100 as outputs when simply touching a sensor. So, I doubt that it's a Voltage.I appreciate a lot if someone can confirm/criticize my solution approach. \$\endgroup\$ – JoeBlack Mar 1 '16 at 18:44
  • \$\begingroup\$ Please provide a schematic of exactly how you have connected this, and to what. All components, including power supplies. What does 200 to 400 mean? \$\endgroup\$ – WhatRoughBeast Mar 1 '16 at 18:59
0
\$\begingroup\$

With calibration you are trying to find out the "transfer function" of your system. There is an input and an output, and an unknown function of your sensor. So you vary the input (with the weights) and observe the change in the output. You can then chart this in excel or graphing software of your choice with an x vs y graph.

$$ Output = Sensor(Input)$$

The task at hand is to find out how the output is related to the input. If the system is linear and can be fitted to a first order polynomial you could use fit it with \$ y = m*x +b\$

$$ Output = m*(Input)+b$$

Or it might be a second order polynomial.

$$ Output = c_1*Input^2+c_0*Input+b$$

Once you find a good model for the systems behavior, the systems effects can be reversed and the observed output can be used to find out what the input of the sensor would be.

$$ Input= Sensor^{-1}(Output)$$

Another caveat to this is you may not be calibrating all of the inputs to the system. Temperature is often overlooked because it changes almost all sensors. So you may calibrate the system at one temperature only to find later and find a different result at a different temperature. The difference in accuracy may or may not be critical to your application.

\$\endgroup\$
  • \$\begingroup\$ Thank you. It's more clear now. The first part of my question is related to the measurements. What do we actually measure with this particular FSR? Force in Newtons or in grams or Voltage? I have a very simple circuit: 2 force sensors, Arduino and resistors. \$\endgroup\$ – JoeBlack Mar 1 '16 at 18:50
  • \$\begingroup\$ That depends on what you define your inputs and outputs to be, if your input is a weight set in grams. Then that is the scale you will use for your calibration (you could always convert them to a different measurement set with equations after you calibrated the sensor). For example, you could have a 10g,50g and 100g weight, but this also corresponds 0.098N, 0.49N and 0.98N. Your output is whatever you are measuring the sensor with, if its an ADC, then its probably volts. If you are measuring the current output of your sensor then its in amps. \$\endgroup\$ – Voltage Spike Mar 1 '16 at 18:59
  • \$\begingroup\$ So if you were to calibrate the system you'd put the 10g weight on, measure the voltage and record both values. Then do the same for the 50g and 100g weights. Then plot voltage vs weight (or you could even plot weight vs voltage) in an xy graph. Then find a function that makes sense to fit it to, use the inverse of the function to find the weight on the sensor after calibration. \$\endgroup\$ – Voltage Spike Mar 1 '16 at 19:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.