• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 672 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Daisy Chained ADS131 converters

Jarret Adam
Prodigy 75 points


Looking at the ADS131 as it  is very attractive since it operates in a chain-able fashion; also, the sample results is stored and can be accessed after the START signal is sent.

Question is: if I send a START signal to 5 devices configured in the Standard Device Configuration, will all inputs sample synchronously and can we "leisurely" query each device one-at-a-time to pull off the sample data? We'll be sampling at rates between 1 and 400 Hz.

  • 0
    TI__Mastermind 23235 points
    Hello Jarret -
    For reference, you can refer to the Multiple Device Configuration of the datasheet on page 43 (if you have not already).
    To synchronize, you need to tie the START pin together and additionally provide the same clock to all parts - either with one as the master or externally for all (details in datasheet).
    To read the data, I am not sure what 'leisurely' means but you need to perform read operations within the same datarate period for synchronized data. Depending on you reading mode, you could overlap the next DRDY a little (if it start in the previous). You can review RDATAC and RDATA modes on page 24 & 25 of the datasheet.
  • 0 in reply to Greg Hupp
    Prodigy 75 points
    Thanks for the feedback Greg. The RDATAC feature seems a great way to pick up all the samples. For a system with 5 ADCs running at 1 ksps on a 10 MHz SPI bus, this sounds like it provides plenty of margin on the serial channel.
    As an alternative, some of our data paths only require sampling at 50 Hz. Would you advise running the 1 ksps mode and dropping 19/20 samples, or manually poll the device? Clearly we'd introduce delay with the latter method; would the delay be t_settle from Table 7? For 1 ksps, that's about 5 ms, which is almost a quarter of the sample period. Do you know what the jitter spec is on that number?
  • 0 in reply to Jarret Adam
    TI__Mastermind 23235 points
    Hello Jarret -
    For your design, it probably just depends on the bandwidth of what you need to do. Just letting the device run and collecting samples generally is less work since you can set up DMAs to handle the data collection. Be careful with running at a faster datarate and just 'picking' out a single data point and 'assuming' that is a different datarate.
    One additional item of note, when you are using the multi-device mode, make sure you review the Daisy Chain Modes section of the datasheet and the items of note. This mode inserts a single bit between the device data, which can cause problems when using standard serial interface micro-controllers/processors. FPGAs usually just have to be programmed to account for the extra bit.