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.

INA226: Response time of alert pin

Part Number: INA226
Other Parts Discussed in Thread: INA3221, INA231, INA321

Hi team,

In the datasheet, there are figure shows that the response time of the alert pin is related to the conversion time.

1. Do you have the response time for the alert pin when the shunt voltage or the power limit happen in INA226 and the INA3221 in different conversion time?

2. Is the response time for the alert pin to de-assert same as the alert pin to assert?

3. After the alert pin has to be assert, if the shunt voltage don't exceed the limit, will the alert pin to be de-assert automatically?

Lacey

Thanks a lot!

  • Hello Lacey,

    Thanks for using the forum.
    1. Looks as though you may have found an error in the datasheet. I suspect the original writer accidentally swapped the plots. I will need to check on this. As for a different conversion time, I will need to check this in the lab.
    2. I will also need to check this in the lab.
    3. This depends on the mask/enable register. If the LEN bit (INA226) or the WEN or CEN bit (INA3221) is set to transparent, the alert pin will automatically de-assert.

    For the items that I need to check in the lab, this may take a day or so to collect in the lab.
  • Hi Patrick,
    1. Do you have the data for the relationship between the response time and the conversion time?
    In my customer's application, they use the INA226 just to detect the current over-limit.
    So they need to the alert pin to respond as quick as possible.
    Their requirement is the response time need to be less than 800us.
    Please help to provide the data in short time, customer need the data to decide if they can use the INA226.


    2. Also for the conversion time, I have some questions:
    What's the internal design for the conversion time? Is it related to the ADC calculation?
    Why does the long conversion time has less noise?


    Lacey
    Thanks a lot!
  • Hi Patrick,
    Any update for below question?
    Lacey
    Thanks a lot!
  • Hello Lacey,

    1.  I spoke with one of our design engineers and he said the device should trigger within 2 conversion  cycles.  For the INA226, the shortest conversion setting, the maximum time is 154us.  So we would expect that the ALERT will trigger after 308us, which is well below 800us.

    2.  The reduced noise can be attributed to the increased sampling involved in the longer conversion time. If you look at the example probability distribution curves, you will see that one is skinnier than the other.  The skinnier one is produced from a smaller standard deviation.  The standard deviation can be calculated from the formula below. For uncorrelated noise samples the distribution is centered around 0.  N corresponds to the number of samples.  With more samples the standard deviation begins to converge towards zero.  Noise values at zero and near zero will minimize the impact of random large noise values, as the small noise values’ frequency of occurrence is considerably higher, especially in a larger sample set.

  • Hi Pactrick,

    Thanks so much for your reply!

    For the response time, if the response time will be less than 2 conversion time, so it means the curve in the INA226 is not correct.

    Because this shows when the conversion time is 140us, the response time is about 1ms.

    So can you help to test the correct response time and send me your test results.

    Because the response time is the key factor for customer to use the INA226, need our prove to show it meet customer's requirement.

    2. One more question for the alert pin.

    When the average number is not 1, such as 10. For the INA226, it will check the shunt voltage every conversion time and report or need after 10 conversion time when it get the average shunt voltage?

    Please help to look into the question, it's important for me to let customer to use the INA226.

    Lacey

    Thanks a lot!

  • Hello Lacey,

    Yes, you found a mistake in our datasheet.  Whoever placed the graphs in the datasheet swapped the tct.  Below I set the INA226 to trip at 71mV Shunt Voltage Under-Voltage alert.  The conversion time I set for the shunt voltage is 140us with no averaging.

    2.  The alert will not assert until all conversions in the set of averages are complete.

  • Hi Patrick,

    Thanks so much for your support!

    But from customer's requirement, they need to monitor shunt voltage in every conversions and also need to use the average shunt voltage.

    So from your description, it seems both the INA226 and the INA231 can't meet customer's requirement.

    But the datasheet description seems the INA226 will compare the voltage in every conversion.

    Please help to confirm!

    If the INA226 and the INA231 can't meet the requirement,  do you have the recommendation of the part which can meet the requirement above?

    I found that the INA3221 has the alert for each shunt voltage conversion and the average shunt voltage conversion. But because the INA3221 has three channel, so it just can monitor three channels serially.

    So had better can have one channel device that can meet the requirement.

    Lacey

    Thanks a lot!

  • Hello Lacey,

    Thanks for your persistence, I misspoke on the last post. The Alert should always trigger within 2 conversion cycles if the trip point is crossed. When I was testing in the lab the other day, I accidentally changed the conversion setting and mistakenly thought the longer alert delay was due to averaging.
  • Hi Patrick,

    To confirm this for customer, can you help to test the response time when you set the conversion time to 140us, and the average to 1024, as below:

    The results maybe the best way for customer to prove this function.

    Lacey

    Thanks a lot!

  • Hello Lacey,

    Here is the setup and the measurement:

  • Hi Patrick,
    Thanks so much for your continue support!!
    One more question: Does the INA231 has the same performance in response time and average function with INA226?
    Lacey
    Thanks a lot!
  • Yes the INA321 has the same response time and average function as the INA226. However, the INA226 has better shunt offset voltage.