Post Go back to editing

Question about AD5410 Iout Fault Register value

Hello,

How often is the value of the Iout Fault bit updated?

How long does the Iout Fault bit hold the value 1?

 

When a disconnection occurs during operation, the /FAULT signal is detected and the Iout Fault in the STATUS REGISTER is read, but sometimes the Iout Fault bit is 0.

It takes time to read the register value because the process of SPI communication 1Mbps (AD5410 x 8 daisy-chain connection) is performed after /FAULT signal is detected.

It is assumed that the /FAULT occurrence register (Iout Fault=1) could not be read because Iout Fault returns to 0 during the process described above.

This is the reason why I asked this question.

 

Regards,

Kazu

  • Hi Kazu, 

    The fault bit is set once the fault is detected. I don't have a eval board for this device atm, but normally, the value on the fault bit remains until it is accessed by a read command. The same should also be the same in daisy chain configuration. 

    It is assumed that the /FAULT occurrence register (Iout Fault=1) could not be read because Iout Fault returns to 0 during the process described above.

    Do you mean that, before the read on the fault register is executed and finished, the fault was already fixed and the /FAULT pin has released its state?

    Best regards,

    Ian

  • Hello Ian-san,

    No, reading the fault register is executed after /FAULT pin has released error state "L".

    My design is to trigger the active signal on the /FAULT pin and go see the register value.

    When I check the register, I sometimes find that Iout Fault is set to "0". However the /FAULT signal remains "L".

    This phenomenon occurs when the output is 0mA, and when the output is 10mA, the phenomenon does not occur and Iout Fault remains at "1".

    I'm currently testing with no load. Is there any relation to this phenomenon? Also, does the time it takes to daisy-chain 8 units to read the data have any effect?

    Regards,

    Kazu

  • Hi Kazu-san, 

    I'm currently testing with no load. Is there any relation to this phenomenon?

    Testing with no load attached should trigger the fault features since this is an open circuit connection. 

    Also, does the time it takes to daisy-chain 8 units to read the data have any effect?

    Ideally, the content on the fault register should be clocked out properly even on daisy-chain operation. You mentioned having 8 devices in the chain, does the missing fault bits happen on all of the devices on the chain? can you try to probe on the SDO pin of the specific DAC that triggered the fault, see if the same problem occurs? 

    When a disconnection occurs during operation, the /FAULT signal is detected and the Iout Fault in the STATUS REGISTER is read, but sometimes the Iout Fault bit is 0.

    How does the disconnection happen? Are manually introducing the fault at the output of the DAC for this?

    Best regards,

    Ian

  • Hello Ian-san,

    Thank you for the reply.

    I have attached the disconnection test condition. Please check the attachment and let me know if you

    notice anything else.

    Regarding probe in the SDO pin of fault chip, sorry I have not tested yet since I don't have the board

    on hand at the moment. 

    Regards,

    Kazu

    PDF