i have the load cell of 100kg i have a ADC NAU7802. i have to calculate weight through this ADC. can anyone have idea how to convert ADC value to weight.
Tell me your sensitivity of our Load cell and also the excitation voltage range... Your ADC resolution..Supply Voltage.. Please give the details fully.. Go through the data sheet NAU7802 and see the how many analog channels it provides.. take the output of your load cell wires into the analog channel pin of your ADC.. Connect to the host via SPI interface.. And do the calibration of the load cell for your requirement like 100 kg after the offset process.. I hope you understand the layout..Go ahead
sensitivity of our Load cell=2mv/v excitation voltage range = 0-5 volt Your ADC resolution =24 bit Supply Voltage= 5v
2 analog channel i used only 1 channel
i used i2c protocol
without load cell when i read the value from the ADC it give me more fluctuation
an open input will, unless you are within a Faraday cage, give "more fluctuation"
anyhow I have some doubts here, since the code for a weigh scale, with which you have problems, is fairly simple the issue to make a reasonable precise scale is LAYOUT, LAYOUT, LAYOUT
Erik
My questions is : 1.Apart from ADC do you have any other slave in your board?? 2.What is the essential part of using I2C interface with the ADC in the board??.. 3.SPI provides faster communication compare to I2C.. Why it is neglected?? 4.Are you using 24 bit completely or 16 bit of ADC.. I mean tell me your total counts like 65536..
May well be true, but seems unlikely to be relevant in this application...?
i get the value from the ADC but there is more fluctuation .but don't know how to convert this value into weight. Do u have any idea
Are you saying that you have started this project not having _any_ idea how to be able to convert an ADC value into a weight?
And you haven't tried with five different weights and collecting the five ADC readings and then tried to plot them on a paper to see if there seem to be any pattern? Couldn't there possibly be some form of relationship between ADC reading and the weight?
Let's say one coin weighs 7 grams and 47 coins weights 329 grams. If you at a later time measures 133 grams - wouldn't you then have enough information to be able to come to a reasonable conclusion how many coins there must be? Or what the weight would be if you had 38 coins?
What do you find so magical about converting ADC readings into weights? Not all sensors are linear but wouldn't the first step be to see if your sensors follows a normal A*x + B formula? How do you convert temperatures between °F and °C? Are there really any significant differences in computing weights from ADC readings?
About the fluctuation - you can improve it a bit by taking multiple samples and compute the average. Or you could spend time trying to figure out what is causing the fluctuation and see if you can improve your wiring or circuit or supply voltage or reference voltage or ground plane stability.
What is vital to do is to figure out if the fluctuation will make a significant difference to your computed values, and also if the fluctuations are may be affected by the environment so that you sometimes may get good values and at other times can get much worse results - a good instrument should either give a measurement within the guaranteed tolerances or avoid to present any measurement at all. It's a question about trust. If the values can't be trusted, then it's time for recycling.
Give a fixed input voltage to Analog input of ADC and check if you get fluctuations in the readings. Monitor the fluctuation in the power supply at the ADC supply pins (between Vref & gnd), if fluctuations still persist. Also, there will always be some minimal fluctuations (5-10 counts) Hence perform averaging of more than 5 (average of 5 is good, but you may reduce or increase depending upon the speed of conversion - dont exceed 16 and perform average of 3 readings atleast).
Load cell=2mv/v ... Supply Voltage= 5v Which simply tells me that at full load, the output of the load cell will be 10mV ( = 10000uV). i.e. 65535 count for 10000uV. By using simple mathematical calculations, calculate the counts vs voltage. Now, the load cell will generate 5000mV for max load. So calculate mV vs weight and then integrate everything.
i want just calculation of weight from ADC output.
That calculation is either a 5 minute task for you. Or it's a showstopper for you. But it is still your task. You have been given clear instructions how to do it. Now is the time to let the world know if you are a man or a mouse.
The important thing here is that you have been given clear information how to calculate your conversion. It's now all on your table. Have the schooling been for nothing?
"dont exceed 16 "
Why put a limit at 16?
Averaging is basically a form of low-pass filter, and depending on needed time constant and sampling speed, a good solution can be to use way more than 16 samples.
Note that rolling averages means that the CPU load doesn't increase with the number of samples. It's just a trade-off on amount of RAM for the sample buffer.
Clear calculations procedure are provided to you. Follow them. Even my brother, an under-graduate diploma engineering student, would have performed the calculations (given the load cell capacity).
gain =128 excitation=4.5 v 24 bit adc so 2^24 -1=16777215;
//convert adc value into volt volt=(((float)present_adcdata*excitation)/((float)16777215*gain)); volt=present_adcdata*.000002095; // i get in mili volt 128 gain
load cell capacity is 100kg load cell sensitivity=2mv/v excitation=4.5 v
100kg -> 9mv
1:100,00 = x *9 mv
x= 9/100,000
x=0.00009 1 gram want 0.00009 mv
not convert adc volt into gram
gram=volt/0.00009;
is my calculation is right or wrong