• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 191 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LMP90100: Disturbances at temperatures near lower operating range.

Miguel Cabrita
Prodigy 50 points
Part Number: LMP90100

Dear Sir or Madam,

 

In one of our devices we use 96 LMP90100.

 

The LMP90100 is configured for multiple channel continuous conversion. We are facing problems using the LMP90100 near the external climate chamber temperature of -40 °C. Our measures indicate an internal device temperature near the ADC about -35 °C. So we are not operating outside of the specified temperature range.

 

Our problem is that on some of the ADC (10% of 96 ADCs) we measure disturbances up to 12 uV. We use a ratiometric measurement and therefore we can exclude problems with our reference voltage.

 

An interesting fact is that those disturbances only appear when we operate the device at a temperature cycle from -40°C up to 85 °C and down to -40 °C. The problem doesn’t always appear on the same ADC. When a new temp cycle is done, it is possible that ADC with problems are good and some good ADC may cause problems. The disturbances appear only at -40°C (climate chamber temperature ). At the other calibration points (5°C, 23°C, 45°C and 85°C) the results are perfect!

 

We’d like to know if there are any limitations or known issues operating the LMP90100 on such alternating temperatures or what you think in general about our operating mode. If you would like to have schematics or source code for the ADC registers please contact me personally.

 

Kind regards

Miguel Cabrita

  • 0
    TI__Mastermind 47625 points

    Hi Miguel,

    Thanks for your post and welcome to the forum! 

    I am not away of any specific concerns in regards to temperature testing with this device, however I know that all circuits are subject to certain concerns when performing tests like this. I have a few questions to help us diagnose this issue. 

    How quickly are you going from -40C to 85C back to -40C?

    Are you opening the chamber door at some point during these temperature changes so that the humidity can normalize? 

    How long is the circuit soaking at these temperatures before data is taken? 

    I want to confirm that the devices do not show issues before undergoing this temperature stress test? 

    Do the devices continue to show issues after the device stress test? i.e. permanently damaged as a result of the test? 

    If these transitions are happening too quickly, there could be moisture accumulating that may have an effect on the measurement. I would recommend opening the chamber for a few minutes once room temperature is reached during the transitions, then allow a significant amount of time (20mins, depending on chamber size, quality of seal, etc.) for the temperature to soak once the final temperature is reached. 

  • 0 in reply to Alexander Smith
    Prodigy 50 points

    Hi Alex,

    Thank you for your answer.

    > How quickly are you going from -40C to 85C back to -40C?

    We start at 23C and configure the climate chamber to -40C and then from -40C to 85C. From 23C to -40C the climate chamber needs around 90 minutes and from -40C to 85C it needs around 60 minutes.

    > Are you opening the chamber door at some point during these temperature changes so that the humidity can normalize?

    No, we do not open the chamber door.

    > How long is the circuit soaking at these temperatures before data is taken? 

    On -40C an 85C we wait approximately 180 minutes until the device acclimatizes and we take data.

    > I want to confirm that the devices do not show issues before undergoing this temperature stress test? 

    The devices do not show any issues before and they go back to normal after a cold reset at 23C.

    > Do the devices continue to show issues after the device stress test? i.e. permanently damaged as a result of the test?

    No, at the next temperature cycle the problematic ADC may work well again, but other ADC may work wrong.

    Regarding the transitions and the moisture accumulation.

    Our analog input channels have a tropicalisation coating.

    Kind regards

    Miguel

  • 0 in reply to Miguel Cabrita
    TI__Mastermind 47625 points

    Hi Miguel,

    Thanks for the clear feedback. 

    The devices that do show errors, when they undergo the temperature stress test a second, third, fourth time, do they continue to show the same error results? Is the failure reproducible?

  • 0 in reply to Alexander Smith
    Prodigy 50 points

    Hi Alex,

    when we run multiple tests, without resetting the hardware, then we measure the same disturbances on the affected device, but it's not necessarily the same device as in the previous cycle. So it's not reproducible for a one specific device. We only know that when we run a cycle, there will be devices with this problem, but we can't tell which it'll be.

    We thought that we'd have a misconfiguration and made many tests with various configurations but the disturbances were independent from the configuration.

    Kind regards,

    Miguel

  • 0 in reply to Miguel Cabrita
    TI__Mastermind 47625 points

    Hi Miguel,

    This may be a SPI timing issue. It's possible that there is already a violation of some sort and as the test over temperature is performed, the violation gets worse. This is a difficult ask, but is it possible to get scope shots showing your SPI communication over temperature? If not over temperature, room temperature is a good place to start. Raw data from the MCU at cold temperature may help as well. 

    It is possible that a bit is being missed or there is a false SCLKs of some sort that is causing this issue. 

  • 0 in reply to Alexander Smith
    Prodigy 50 points

    Hi Alex,

    The disturbances are constant. We measure them over a few minutes. If this would be an SPI SCLK issue, wouldn't the disturbances be like spikes when the data kinda randomly changes due to shifted bits?

    Another observation that this seems not to be an SPI issue, is that when we run a cylce, at the first -40C the values are ok. The problem comes after going to 85C and then back to -40C.

    Why we do this is because we measure a high precision resistor, which is not inside the climate chamber, at both temperature points. We expect to get the same values, with very small disturbances, in the second measure at -40C and in most of the devices it is the case.

    If you'd like to see any data of our measures, let me know.

    Kind regards,

    Miguel

  • 0 in reply to Miguel Cabrita
    TI__Mastermind 47625 points

    Hi Miguel,

    Possibly, unless it is consistently happening at the end of a string of SCLKs. 

    If you can get it, SPI communication with a device showing the error would be great. I would like to see what data you are receiving and know what data you are expecting to receive. From there, we can see if there's a missed bit somewhere or if an SCLK is being tripped when it shouldn't be. Then we can look at the timing of the communication to ensure that the temperature change doesn't cause a timing violation.  

  • 0 in reply to Alexander Smith
    Prodigy 50 points

    Hi Alex,

    In our internal meeting we discussed the possibility of an SPI Problem. The fact that this disturbances only appear at one channel, when we are measuring three channels in "ScanMode2: One or more channels Continuous Scan", makes us think, that the SPI probably won't be the problem.

    Alexander Smith said:
    If you can get it, SPI communication with a device showing the error would be great. I would like to see what data you are receiving and know what data you are expecting to receive. From there, we can see if there's a missed bit somewhere or if an SCLK is being tripped when it shouldn't be. Then we can look at the timing of the communication to ensure that the temperature change doesn't cause a timing violation.

    We'll need some time for this to happen due to the actual pandemic situation. I'll let you know when we can provide you the data.

    Thank you for your patience.

    Kind regards,

    Miguel

  • 0 in reply to Miguel Cabrita
    TI__Mastermind 47625 points

    Hi Miguel,

    Thank you for getting back to me. 

    This is certainly a good point and we are continuing to discuss internally what the issue could be here. 

    In addition to the SPI scope captures, can you provide a set of good and bad data (raw ADC data)?

    A more complete description of the system including why type of sensors are being used? 

    A schematic?

    I understand, things are a little more logistically difficult than normal on our side as well. Please let me know when you have the data and stay safe.