I don't know how to calculate from adc value to voltage in positive and negative
Hi Kittadon,
Welcome to our forum! Bob Benjamin answered a similar question on the ADS1232 a few months ago. Please review this post and let us know if you still have questions:
http://e2e.ti.com/support/data_converters/precision_data_converters/f/73/p/155786/564470.aspx#564470
Regards,
Tom
Hi Tom
i use formula equal ADC_value/(1L<<24))*(5.0/128)
if negative voltage i read 19mV only
Kittadon,
That formula should get you very, very close, but there will be a tiny error. The formula you use assumes a full scale unipolar value. As this is a differential measurement, you should adjust it slightly:
LSB = (0.5(VREF)/GAIN)/(2^23 - 1)
As the difference between this result and the result from your formula is tiny, you must have another issue. What is the ADC_value that is returned, and what is the actual input value? 19mV is full scale, so maybe the problem is due to common mode. The DC value of the input must be at least 1.5V off the supply rails.
Best regards,Bob B
Hi Bob
Thank.
I make circuit follow datasheets (Figure 42 )
1.I not connect loadcell.
2.reset and send CLK 26 th for cal offset.
3.I read ADC value about 0xFFA0EF ,i don't no why value not equal 0x000000.
You can tell me ? step
My loadcell ,,pull voltage is positive ,,Press voltage is negative.
If you have not connected the load cell, then the inputs are floating. The inputs should be connected to some input voltage, otherwise the input can float to any value. If you connect the load cell, there must be an excitation voltage to excite the bridge, with the other end of the bridge connected to ground. The output voltage of the bridge can be connected to the ADS1232 inputs in either direction, so you don't need to have press voltage as negative ( it can also be connected to press is positive and pull is negative). You also need to make sure that your address select pin is correct for the connected input.
It would be helpful to see your actual schematic, and not just a reference to a figure in the datasheet. This would include how you are planning to connect the load cell, and to which analog input pins.
Best regards,
Bob B
my sch
and
so voltage = adcvalue * LSB correct?
Thank for help.
Kittanadon
The formula is correct. However there are a couple of issues with the schematic. C20 should be 0.1uF, and C16 and C18 are too large. The common mode cap values should be 1/10th ( or smaller) as compared to the size of the differential cap. I know the larger common mode caps help with filtering, but you will see great problems with drift and errors due to mismatch. I would just remove C16 and C18. Large caps will also create large delays in response.
The rest of the schematic should be ok. When you connect the load cell, what is the ADC value with no weight?
Thank for answer.It work
but,voltage is not correct. I measured by Meter 0. 1 mv but read from ads1232 about 0.2 mV
Best regards, Kittanadon
Kittanadon,
The accuracy of the actual measurement will be determined by the exact voltage reference value used and the accuracy of the meter you used to measure the voltage at the input. I expect you have error in both. Unless the meter you are using is highly accurate and is calibrated, you can only expect it to be an approximation. You might also have a large degree of noise that is affecting the result as well. The meter may be averaging the noise out of the measurement.
how to calibrate ADC to remove offset error by software
There is a self offset calibration that can be executed via software. This is a special pattern that is described in the datasheet.
The other option is to subtract out the error in your software. With no weight on the scale, the code returned should be subtracted from the result so that the software then reports this value as 0. When you actually place a weight on the load cell, you will continue to subtract out the code result for 0 weight. The resultant value (code_with_weight - code_no_weight) is the actual value.