Hi i am Electronics Engineering student. I am using 10Kg load cells in my final year project. But the problem is that the Load cells give very small output. My load cell gives just 0.5mV/Kg change at its output and i need to give it to microcontroller. So it needs to be amplified first.
OK, used only in tension. Generally the output of a load cell is specified as mV/V based on the applied excitation voltage. Typically numbers like 2 mV/V or 3 mV/V are popular with a maximum excitation voltage called out. What this means is if I have a 10 Kg load cell with an output of 2 mV/V and I apply 10 Volts excitation that with 10 Kg of force (full scale) my output will be 20 mV. That is 2 mV per Kg.
Years ago I built my strain gauge amplifiers but today I buy them. I did have good luck using the AD524 which can be found here.
As to resolution and accuracy don't confuse them. A very large factor of your resolution will be the A/D process. If your final amplified voltage runs into you u-Controller A/D you will have only so many quantization levels to resolve the voltage. I understand where you are going with only using maybe the bottom 1/5th of your load cell, however even if you scale your amp so 0 to 2 Kg = 0 to 5 volts you still face the A/D process and depending on bits the number of quantization levels.
Your accuracy will be a function of the load cell along with the overall amplification and the overall linearity.
The chip I linked to isn't inexpensive but works quite well for these applications. It also allows for a load cell bi-polar input. Few external parts and easily setup.
Actually my project is a bit simpler I need not to make weigh scale
I need some of desired weigh lets say i want just to know that if 700g weight is reached and then compare it with comparator and give microcontroller the signal 1 or 5V
So i need to know just output of load cell at say 700g and just to compare it
So you amplify and signal condition your signal from your sensor (load cell) and that gives us a voltage. If you have a set of precision weights to hang from the load cell that would be nice. So let's say in the tension mode you hang (weigh the hanger) enough weight to equal your 700g example. Measure your voltage. Next build a comparator circuit around the LM339 (though you don't need a quad) around what you see here. Allow for a variable reference like a pot. The comparator will give you a High or Low depending on if your input voltage (load cell) is above or below your reference voltage set by the pot. This is sort of a yes/no approach. My choice would still be the AD724 instrumentation amplifier.
As we know that Load Cells have very small output voltage as said earlier so it has noise issues also. Do i need any capacitors or something else or i can directly interface it with AD524 ??
If you review the data sheet see figure 41. Personally I never used it but set things up like figure 52 for a typical bridge input. Just make sure you use shielded cable and note how they ground to prevent ground loops. That worked for me as I recall. Keep leads as short as possible and things should go fine. Mostly a matter of applying good practices and it worked fine for me. Typically with signals of 20 to 30 mV full scale I amplified X 1000. Then firther amplified to 5 or 10 volts FS.
Here's what I use for amplifying the output from a strain gauge, I needed about 1 volt and had to amplify my signal by about 3000. Don't seem to have much problem with it. It's a Texas Instrument INA128.
Kinarfi
I am using 2 stage amplified load cell signal. First amplification from AD624 and then from non-inverting op amp. Now i need to compare it with a threshold.
But the problem is that when i give amplified output into the comparator, the change in weight on load cell does not affect the output of amplifier rather it remains constant.
Why it happens ?