This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
An 8-channel MUX is connected to the input of the ADC121.
The ADC is using auto mode and the MUX channels are switched every 1ms before reading out with SDA.
I would like to ask about the disadvantages of increasing the sampling rate of fconvert.
I actually did a comparison between 1.7ksps and 13.5ksps settings.
The 13.5ksps was more accurate as a reading.
I would like to know if there is a recommended setting method.
Hello,
The cycle time setting used is dependent on your system requirements and the input circuit. The input circuit needs to be able to drive the input and settle within the allotted time, thus the faster the device samples, the shorter time the input circuit has to settle. The device performance will remain the same across the range of sampling rates.
it is odd that the faster setting resulted in a more accurate reading compared to a lower sampling rate. I would recommend using the sample rate that meets your system requirements, and then optimizing your input signal circuit to settle within that.
Can you share your schematic, including input signal circuit
Regards
Cynthia
Hello,
Cynthia-san
First, for our purposes, we want to read all 8 channels of the MUX during a 10ms period, and we switch the MUX channels every 1ms.
The ADC read timing occurs just before the MUX channel is switched. Therefore, there is a 1 ms read wait time after switching the MUX channel.
The right side of the attached schematic is connected to the MUX input. SWBAT in the schematic is 12V.
The other 6 channels are the same circuit, SWBAT and D5.0V are basically the same voltage.
I have a question about your answer.
I am aware that the conversion result is overwritten in the conversion result register unless the conversion result is read out. In this usage, for example, 13.5ksps would mean that 13.5 conversions are performed during 1ms. When data is read back 1 ms after switching channels, we assume that the data converted during that time will be discarded and the data remaining in the register will eventually be read back. As a result, I thought that the data readout would only read back stable data after 1ms, but is this not the case?
Regards
Katsuya
Katsuya-san,
In automatic mode, the ADC will only hold the conversion value until the next conversion, which will happen automatically, and will overwrite with the updated measurement. In normal mode, the ADC will hold the value until it is read out. This means that how you are planning to use it will not work, if the device performs 13 conversions, and then you read out the value, you will read out the last measurement, not the first one.
From what you shared, it seems that for your application, normal mode may be the best fitting, instead of automatic mode.
Regards
Cynthia
Cynthia-san
Thank you for your response.
Again, in my usage, the input signal settles down after enough time has passed after switching channels on the MUX, so no matter how fast I set the cycle time in Auto mode, it keeps overwriting the previous value and eventually only the data remaining in the register can be read out.
What mechanism in the ADC causes the input signal to become unstable even when the cycle time is set to a high speed?
Regards
Katsuya
Katsuya-san
I am not sure there is an issue, this is how the device is designed to function.
At higher speeds, the input circuit has less time to drive and settle the input signal. This will be seen as offset errors in the output measurements. To fix this the input circuit will need to include an amplifier, or the speed will need to be reduced.
I think a phone call will be best to make sure the issues addressed. Would you please connect with your sales or field engineer to set a phone call.
Regards
Cynthia
We were able to find the best value here and solved the problem, so there is no need to call us.
The optimum value was 3.4 ksps, which can be converted three times during 1 ms.
Thanks for your support.