How to construct components with a 314030003 load cell to read starp tension.
Hi, did you mean to say strap tension?
I think this loadcell has three wires, white, red and black. The two strain gauges (on either side of the bit that bends) are between white and red and black and red. They each have a nominal resistance of 1k ohms. So you need to build a bridge with two 1k resistors connected in series between white and black. If you connect 5V to white and ground to black and measure the voltage between red and the point between the two resistors you should have a signal which varies with load. There may be a very small voltage without any load due to resistance mismatch. Even with your maximum load you’ll only get a small voltage which needs to be amplified. The gain of that amplifier depends on the signal you read at full load and the maximum analog input of 3.3V with a bit of headroom for overload. If this is a partial answer to the question you are asking I recommend you try this first and note the voltages measured. I’ll look out for a response from you.
1 Like
Just to complete this story, unless you are looking for something really accurate, you can get away with just one Op Amp and four resistors to make a differential amplifier, The gain is the ratio of the two (equal value) feedback resistors divided by the two equal value input resistors. I recommend LM321/LM358/LM324 which are inexpensive single/dual/quad amplifiers respectively. They will work from a single 5V supply but they are not rail to rail so the trick is to adjust one of the bridge resistors (mentioned in my first reply) upwards from 1K and one downwards to make the output near to one rail when there is no load, and close to the other rail at full load. This may get to a maximum of maybe 4.5V with a 5V supply so remember to use a potential divider on the output to reduce the voltage to about 2/3 to be sure the voltage fed to the processor is below 3.3V. You can buy modules which do a better job than this but it all depends on your budget and how important the accuracy. Make sure your bridge resistor differences and gain value is not putting the output too near the rail voltages to avoid being non-linear as you approach the rails. Your A/D will always give a number bigger than zero and (hopefully) less than full scale, so you are losing maybe 1 bit. Calibration needs to be done at zero load and full load as a minimum. Other points are maybe worth calibrating by using a lookup table. Correction for hysteresis (is the measured load going up or down) is probably more important than using an expensive load cell amplifier.