So I have a little project I want to do. It has been done on a smaller scale :
I want to do something similar but on a 15x15 inch surface and measure the pressure in lbs for each sensor. I have never used arduino but I do have some programming skills in java and visual basic which I am hoping transfer over well to the libraries in arduino and python. I am about to purchase the arduino kit as well to get started on my project.
What I was hoping the community help me with along my journey is on what approach is best to take for measuring pressure in this large of a surface area. Basically I am going to load the 15x15 grid with about 225 pieces of velostat pressure sensors and see the readouts on the screen. Not sure what velostat produces output wise but if I take a weight and compare it to the velostat readings then I could convert my output to lbs.
So does anyone know how accurate velostat is and if its not that accurate then what are the alternatives? I need to keep this as thing as possible so velostat seemed great but not sure what the maximum pressure is on this type of sensor either.
Any guidance would be great because this arduino concept has just opened up so many doors for my imagination its incredible!!!
Thanks!
Velostat is not very accurate.
The Force Sensitive Resistors (FSR) are better, but still not good enough to measure weight.
This question has been asked a number of times, and for myself I would like to measure the weight of each square for example for a mattress. But there is no cheap solution.
Some manufacturers of mattresses have such a grid-like mat, which measures the weight. I'm not sure what they use. Perhaps a lot calibrated FSRs.
Thanks for the tip. I have read about FSRs and as you said their supposively very innaccurate. However, when you indicate that the mattress ones could be calibrated it raises a new question.
Say I have a 5 pound weight on a FSR and calibrate my readings to specifically that sensor meaning this many pounds. Well if I sit that weight on the fsr sensor ten minutes later should my readings be exactly the same as before or is that what is meant by the inaccuracy, that the results change over time for the same sensor?
I have found:
which reads a 3% error percentage which isnt bad but would prefer a cheaper alternative.
bradyboyy88:
Say I have a 5 pound weight on a FSR and calibrate my readings to specifically that sensor meaning this many pounds. Well if I sit that weight on the fsr sensor ten minutes later should my readings be exactly the same as before or is that what is meant by the inaccuracy, that the results change over time for the same sensor?
Your sensor says it has a 3.8% drift per logarithmic time scale (whatever log that might be). It'll continue to be less accurate the longer the item sits on it.
In that Sparkfun page is a link to the datasheet.
According to the datasheet, a single FSR has a Single Part Force Repeatability of 2% to 5% of established nominal resistance.
And the Part-to-Part Force Repeatability is 15% to 25%.
So after calibration of each FSR, the repeatability is 2% to 5%. That's not so bad.
The mechanical way how the weight is applied to the FSR is also important. Will it put force on the FSR every time in exactly the same way ?
Chagrin:
Your sensor says it has a 3.8% drift per logarithmic time scale (whatever log that might be). It'll continue to be less accurate the longer the item sits on it.
Interesting. I did not realize there is a drift in error. I got my masters in financial mathematics but sure wish it was physic right about now lol. I can remember in grade schools calc based physics there is a graph related to resistance over time but I could just be thinking of something else. Is there some theory behind this less accuracy over time such as a curve which I can put back into the calculations/measurements to recorrect or a consistene curve which I should be able to predict from?
Peter the repeatability concept you just mentioned really helped me more than you can think. I need to read the specs more to these sensors. yea 2 to 5% is not bad but that site also seems to have this with less than 3% maximum error:
but its 20 bucks a pop. What makes one sensor more accurate than another? Is it the protective barrier tor reduce interferance/outside variables or a completely different material inside? The reason I ask is maybe there is a cheaper way to manufacture this DIY?
I am purchasing my arnuido now so im taking the plunge!!
I think the material is about the same. Those large differences in prices is normal, even for the same component from the same manufacturer. I'm sure that Adafruit selected good quality components.
The 3.50 dollar FSR look different, they may be from a cheaper manufacturer, but they could be just as accurate.
I think the material is about the same. Those large differences in prices is normal, even for the same component from the same manufacturer. I'm sure that Adafruit selected good quality components.
The 3.50 dollar FSR look different, they may be from a cheaper manufacturer, but they could be just as accurate.
Well I ordered my arduino uno yesterday night. i am going to get these sensors now. I may buy the 7 dollar one and the 20 dollar one to test. The 3.50 sensor has a weight limit of 2 pounds basically which is much less than I need. The 7 dollar one may be too low as well. I should be expecting around 50lbs per square inch of my grid thus I need to keep that in mind.
I also will be heating these up to about 400 degrees so hopefully that is not a problem lol???
Basically my buddy bought some cheapo heat platen which doesnt have a pressure sensor on it like most of them so we are just trying to make a grid of sensors which will measure the pressure distribution across it when clamped down. The FSR sensors are thin enough to imitate the thickness and feel of the cloth fabric we would be pressing so they are perfect; however we do want to be able to do this with the heat press on since I may throw in some temperature sensors as well. We are going to be using that matrix concept in the first link I gave to get a visual. The flexicare sensor:
FlexiForce HT201 Sensor
Seems to be perfect for temperature and pressure but is insanely expensive. I could do the temperature sensors on top of some temperature barrier and the pressure sensors on the bottom. Is there some sort of thin material that acts as a temperature barrier to protect the sensors?
Is there any flat thermocouples which can withstand up to 400 or more deg fahrenheit? I am looking and coming up short in regards to adafruit and sparkfun . The reason for a flat one is because i do not want it to interfere with the pressure readings if the thermoucouple is hard and tall.
Scratch doing the pressure and temperature at the same time. I feel as though its impossible given the thermocouples I have found on line. I cannot find a thin enough one which would not interfere with the pressure readings. I will just test them seperately.
The FlexiForce HT201 Sensor seems to be the one for you. Yes, it is very expensive, but you want to do something special.
A heat barrier, hmmm, that might be possible. Look at clothes for fireman and for people handling melted metals. You need a combination of shiny metal aluminium heat reflective material and isolation that can withstand high temperatures.
Combine that with remote contactless temperature sensing, and you have something that might work.
Well I am going to do the measurements seperately so I do not have to pay an arm and a leg for the pressure sensors. I am wondering though if there are any other fra sensors you came upon which allows up 50 psi? Only other one I have found is for about 20 bucks a sensors from adafruit.
As for temperature probes, I may be able to go cheaper now given I will only be apply small pressure for testing the temperature.
I wanted to do 9 pressure sensors and 9 temperature sensors total. Did not realize how pricey these could get!
I don't know a cheaper solution, beside the clothes for fireman.
When I start with a project, I buy a number of things to learn what it does and to get some feeling with the project. The resulting project could cost only 1/20 of everything I bought to test.
Peter_n:
I don't know a cheaper solution, beside the clothes for fireman.
When I start with a project, I buy a number of things to learn what it does and to get some feeling with the project. The resulting project could cost only 1/20 of everything I bought to test.
Well to test I bought 4 of the square pressure sensors you mentioned previously from sparkfun. I am going to see if I can calibrate them and predict the drift error to help smooth out my estimates. I honestly think with some statistics or some sort of modeling technique I can narrow down that 2-5% error. Also, I read there is some stuff you can dip the sensor into that will remove other variables as well which fluctuate output. Some sort of rubber dip or spray?
Also, just curious about the surface area. How does surface area effect these FSR sensors? I was debating the smaller circle one from sparkfun vs the square one but did not know if the smaller surface area produce greater accuracy?
I have a few FSRs, but I never tested it (a project waiting for me), so I don't know how the surface area has influence.
Is there some kind of dip for the FSR sensors ? I didn't know that. Perhaps some latex or silicone caulk.
The numbers in the datasheet are for a nice smooth equal pressure of course.
Peter_n:
I have a few FSRs, but I never tested it (a project waiting for me), so I don't know how the surface area has influence.
Is there some kind of dip for the FSR sensors ? I didn't know that. Perhaps some latex or silicone caulk.
The numbers in the datasheet are for a nice smooth equal pressure of course.
I guess I kind of figured that the surface area would pay a key roll given the distribution of pressure effects the resistance, thus the smaller the surface area the less likely of unequal distirbution of pressure for repeated cycles hence lower repeatability error. Unfortunately I already ordered the larger surface area so now I am kicking myself lol. After I do some testing I will see if its worth buying more sensors.
What is your project on? The reason I ask is because maybe we can share results or improvements of using these things. When my sensors come in I plan to go full throttle on getting this done!
You really haven't told us what your project entails; all I can gather is that you have a heated plate which, I assume, you're clamping things onto and are trying to get pressure readings for that clamping.
. The reason for a flat one is because i do not want it to interfere with the pressure readings if the thermoucouple is hard and tall.