ADS8686S: ADS8686S数据读取值偏小

Hello, as to your questions, please confirm if the design meets the requirement of the datasheet description:

First,  about the REFCAP, this pin must be decoupled to REFGND using a low equivalent series resistance (ESR), 10-µF ceramic capacitor. Place this capacitor as close to the REFCAP pin as possible. Do not drive any external load from this pin.

As to the REGCAPD, decouple this output pin separately to REGGNDD using a 10-µF capacitor. Place the capacitor close to the REGCAPD pin.

In addition, Decouple the digital LDO (DLDO) with a 10-µF capacitor between the REGCAPD and REGGNDD pins.

At last, the recommended sequence is to  power up DVDD first, followed by AVDD. Hold RESET low until both supplies are stabilized.

Also, care should be attention to the pcb layout which is also the reason which could cause the incorrect voltage. Pleaser refer to the 10.1 Layout Guidelines section.

www.ti.com.cn/.../ads8686s.pdf

I have confirmed that the decoupling capacitors are configured correctly, but the power on timing is not in accordance with your recommendation. At present, the 5V power supply is started first, then the CPU is started, and then the 3.3V power supply for peripheral power supply is started. Is this key to the ads8686s?

Hello, ADS8686 has no strict power up sequence , but proper power up sequence is the basic of normally work. 

The recommend power up sequence is to power up DVDD first, followed by AVDD.

If you power 5V first and then 3.3V , what about REST?  It must hold RESET low until DVDD and AVDD  are stabilized.

I reset the ADC chip after the whole CPU system is powered on and works normally. By measuring the signal waveform, it is found that the reset problem can be basically eliminated. The following figure is the waveform of the data fitting I actually tested. From the shape in the figure, we can see that the signal shape is still correct, but the value is biased and small. The analog signal I actually applied is 50Hz / 0.3V, and the ADC analog channel range is 5V.

The maximum code value I actually received is about 2756. There are differences between different channels, but the shape looks right.

At present, I have tried software mode and hardware mode, and the results are almost the same without obvious difference. I also tried internal benchmarks and external benchmarks, and the results are similar. Anyway, the result is wrong.

Thank you for your feedback information.