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.

ADS8686S: AD7616 Pin2Pin replacement: Met problem

Part Number: ADS8686S

Hi Team,

My Customer is using our ADS8686 to replace AD7616 and we met a problem on one of their design. 

  • issue:

we found that we could directly replace on DSP(TMS32F28335) platform but on another ARM platform, some issues are happened. we could not read the data result from ADC;

From software, there are no different on configuration from code (different platform have different range of ADC input, so there may be some difference on translation sequence);

From hardware, on F28335, we have 22ohom on data bus and on ARM platform, CPU are directly connect to AD7616;

For the usage of AD7616, The software configuration mode uniformly uses one CONVST signal to convert all channels (the channel conversion sequence of different products may be different), that is, burst Sequencer mode is used.

  • Testing:

The test results on the ARM platform show that the CPU can normally write and read out the ADC registers, and the ADS8686S can also convert normally to complete all the configured channels. The basis of this conclusion is that the width of the BUSY signal is correct. After reading the STATUS register after the falling edge of BUSY, the channel number of the last conversion can be read. That is, 11-8bit and 15-12bit in the figure below are exactly the same as the configuration.

When reading the ADC data after the falling edge of BUSY, only the data of V0A can be read, and the data size is consistent with the analog input range of V0A and the configured ADC. (V0A input is DC1V, Ref is 2.5V, and the data read is 13100). In addition, during the test, it was found that when the oscilloscope probe was used to measure only the RD signal or the CS signal, the data of V0A would be read only, but the order of the data was incorrect and not fixed. Sometimes incorrect data appeared. The data was only 1/2 of the correct data. 1/4 (no multiple tests to find a pattern).

Use an oscilloscope to measure DB0—DB15 signals, and find that the waveform of DB cannot be measured correctly. The waveform is as follows:

Yellow for CS  and Green for DB0

  • Other Test:

Testing on F28335 Platform

Yellow for CS and Green for RD

Testing on Arm platform

Yellow for CS and Green for RD

Note: Under the ARM platform, two 16bit data must be read at a time, and the interval cannot be adjusted. That is the position indicated by the red arrow in the figure above.

  • Question:

By comparing the data sheet, it is found that the test condition of the DIGITAL OUTPUT item in the Electrical Characteristics table of ADS8686S is Io=500uA (page 13 of the data sheet), and the test condition of AD7616 is Io=100uA.
The ARM platform products we currently use are mass-produced products, and there is no need for hardware inconvenience to change.
We suspect that it is caused by the insufficient bus driver capability of the ARM platform currently in use. 

Would you help confirm the problem and provide guidance.

BR

Brandon

  • Hi Brandon,

    ADS8686S supports Burst Sequencer mode as well. To be clear, did AD7616 work well on the ARM platform? Is the data bus including /CS signal shared with other devices? Is there any capacitor added on the bus including /CS and other control signals? Was the ADC designed with the ARM processor on a same circuit board or on different circuit boards connected with wires/cables?

    Best regards,

    Dale

  • Hi Dale,

    Yes,AD7616 works well on ARM platform;

    there are some difference:

    On F28335 platform, we connect 22ohom on the bus line and could replace AD7616 with no issues;

    but on our local competitor ARM platform, we have no 22ohom on the BUS lines and all 16 channels' s data are V0A;

    this should be all the different on these two designs.

    BR

    Brandon

  • Hi Brandon,

    Normally it's not a problem for ADS8686S without 22ohms series on the data bus. Can you share the database (sch & pcb) to me in an email? I need more information to address the issue.

    Regards,

    Dale