0

I have a setup where I need to measure the weight at the end of an arm.

Setup

The arm is clamped between two pieces of aluminum at one end. The Sensor S is mounted a short distance (5cm?) from the aluminum base. The distance from the sensor to the end of the arm is 25cm.

The weight at the end of the arm ranges from 1kg to 100 kg. I'd like to measure it with an accuracy of 0.1kg.

How do I calculate what kind of range my sensor should have?

How do I select the best sensor for this application?

Dutch2
  • 153
  • 7
  • 2
    Strain gauge... – Eugene Sh. Nov 20 '17 at 20:48
  • I am not able to mount a strain gauge on the arm. That is why I was thinking about measuring the weight this way. – Dutch2 Nov 20 '17 at 20:52
  • I doubt what you have in mind will work. What else can you mess with? The Clamp perhaps? – Trevor_G Nov 20 '17 at 21:25
  • 3
    So you want one part per thousand ***accuracy*** at full scale? From that location? Are you kidding? What's the drift spec? What about temperature effects? Etc. How will you calibrate this against a standard (DIN, NIST, etc?) – jonk Nov 20 '17 at 21:27
  • Yes, I can modify the clamp. – Dutch2 Nov 20 '17 at 21:27
  • I have a micro with a 12-bit A/D converter. Resolution would be 25g, I was hoping for 100g accuracy. – Dutch2 Nov 20 '17 at 21:32
  • We are talking across the bow from each other (cross-purposes.) Do you understand the terms? See [Accuracy and Precision](https://electronics.stackexchange.com/questions/310436/how-to-improve-resolution-and-presicion-of-sensor-reading/310500#310500) and then repeat to me what you want of those. Note that accuracy requires periodic calibration to standards. And precise tracable-accuracy weights providing 1 part per thousand accuracy are well over US$100 each!! – jonk Nov 20 '17 at 21:42

1 Answers1

1

You will never get the kind of accuracy you need detecting the deflection in the arm like that.

Instead modify the clamp. Perhaps something like this.

enter image description here

The longer you can make the clamp block, the better the accuracy since a longer block lessens the effect of manufacturing tolerances between the hole and the sense point. Though you can likely calibrate that out.

Of course you also need to make it long enough that 100kG on the other end does not swamp or crush the force sensor. I guess the ultimate length should get you just under max sensor force at your max arm load.

Trevor_G
  • 46,364
  • 8
  • 68
  • 151
  • Thank you, that looks like something I could do. Why does the length of the clamp affect the accuracy? – Dutch2 Nov 20 '17 at 22:17
  • @Dutch2 A longer block lessens the effect of manufacturing tolerances between the hole and the sense point. Though you can likely calibrate that out. Of course you also need to make it long enough that 100kG on the other end does not swamp or crush the force sensor ;) SO I guess ultimate length gets you just under max sensor force at 100kG – Trevor_G Nov 20 '17 at 22:35