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.

ADS1261: Synchronizing AC excitation across multiple ADS1261s

Part Number: ADS1261

Hello,

Is it possible to save on component count /complexity by synchronizing multiple ADS1261s so that AC Excitation is controlled by a single set of transistors and thus the excitation source can be shared across chips?

If so, what's involved in the synchronization, I'm assuming this is technically possible if we have a single clock source for all the ADCs and tie the START line together and ensure they are all set to AC/DC excitation in the same configuration. However - I worry that there are many avenues to mess this up software/hardware wise.

Thanks.

  • Hi Alexander Klemenchuk,

    What does your system look like? Are you envisioning having multiple bridges / load cells in parallel such that they all share the same excitation source?

    How would you take data - using single-shot mode where you initiate the conversion, then take one sample, then restart the conversion, take another sample, and so one? Or would you use continuous conversion mode where you initiate the conversion and then continuously take data indefinitely?

    I think if you put all devices in the same mode (2-wire or 4-wire), used the same high accuracy clock source, and tied the START pins together (or used a combined DIN to send the START command), this might work. Your best bet would be to get two EVMs and actually try it out, since the excitation circuitry is already included on them both. You can then control one via the GUI, and then flywire the control signals over to the other EVM after disabling the controller on that board (you would need separate DOUTs however, and a way to pull the data from the flywired EVM because \they cannot both provide data to the GUI)

    However, I would first ensure that AC excitation will actually benefit your system. I have seen this method cause more problems than it solves due to the complexity of the switching and timing. And your system would be considerably more complex.

    -Bryan

  • The current design idea is to use 4 ADS1261 for 4 bridge sensors (for high parallel sampling), all run in continuous AC excitation mode, with selectable 4/6 wire strain gauge bridge measurement (sense lines), making use of the muxable two reference sources of the ADS1261 to pick between referencing the excitation directly - or through the sense wires. The wire to the bridge may be fairly long with poor signal quality. The excitation and the sense lines will have identical filtering to the main strain gauge differential measurement in order to minimize mismatch. The full bridge system cannot be simulated as we provide the equipment to measure, not the sensors or wiring - so it's unclear of the full potential of AC excitation. We've got a single EVM coming.

    I suppose that the issues I'd want to avoid are things such as:

    - All of the ADS1261 needing to be at the same sample rate.
    - The potential for electrical issues if one ADS1261 is expecting a certain reference voltage polarity.
    - The potential for desync (relying on tying the START lines together being flaky).

  • Hi Alexander Klemenchuk,

    All of the ADCs absolutely need to be at the same data rate for this to work. The chopping frequency scales with data rate, so the ADCs will immediately be out of sync if they are not at the same data rate. If this is an issue for you, you would need to identify an alternative solution i.e. having dedicated switching circuitry for each ADC

    The AC excitation feature automatically switches the reference polarity, which is why you need to enable the AC excitation on all devices even if you are only actually using the AC excitation signals from one ADC. You could also perform this action manually if need be, but then again your system is getting more and more complex

    Again, I would suggest testing this with two EVMs (or the EVM you have and some sort of breakout board for another ADS1261).

    -Bryan