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.

PGA309 Failure issues

Other Parts Discussed in Thread: PGA309, PGA309EVM-USB

Hello,

            When using the PGA309 in a pressure sensor application we installed a pressure sensor on a compressor and after 4 hours it quit working voltage went to 55mV read EEPROM and found Ti flag had changed. After reloading calibration system worked fine for a few minutes and problem returned. I am assuming that the change would indicate an error (fault) code. Where can I find information on the meanings of the codes and the best way to diagnose them?

Thanks,

Jim  

  • Hi Jim,

    The PGA309 is a complex device and the two main support engineers that cover it are out of the office right now.  Have you checked through the PGA309 datasheet and user's guide for information on this fault condition (links below)?

     http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=pga309&fileType=pdf

    http://www.ti.com/general/docs/lit/getliterature.tsp?baseLiteratureNumber=sbou024&fileType=pdf

    Here's some text I found in the User's Guide that seems like it corresponds to the PGA309 operation you're seeing.

    3.3 Checksum Error Event
    If at any time the PGA309 detects an invalid Checksum1 from the first part of the EPROM, the PGA309 will disable VOUT, wait for approximately 33ms, and try to read the EEPROM  again from the beginning. It will continue to reread the EEPROM indefinitely. If at any time the PGA309 detects an invalid Checksum2 from the second part of the EEPROM (the Lookup Table data), it will disable VOUT, set the Gain and Zero DACs to their POR values, return to the read configuration register portion (first part of the EEPROM) of the loop, and then try to read the EEPROM Lookup Table again when the next temperature conversion completes.

     There is a great deal of information in the User's Guide describing the Checksum and the answer is likely in there.

    That said, it sounds like there may be some type of disturbance or transient event that's causing corruption of the EEPROM registers which prevents the PGA309 from operating until the EEPROM is reprogrammed.  Is there human or machine interaction with the pressure sensor / compressor during the test that could result in a harmful ESD event? Are all people and machines/test equipment properly grounded in the test area? How often has this event occurred and on how many units does it occur on? Based on the description I'm not sure the likelihood of this, but any chance of large amounts of radiated/conducted energy getting to the sensor during the test time?  

  • Hello Jim,

    How are you calibrating and programming the PGA309? Is this a sample system using our PGA309EVM-USB, or a system of your own design?

    Mysterious corruptions of the EEPROM data are more likely to occur in 3-Wire mode (1W connected to Vout) since noise or other fluctuations at the Vout pin can be misinterpreted by the device as digital communication attempts. This is why we provide the option to disable the 1W interface using the OWD bit (register 4, bit 15 set high). If this bit is set, communication will only be possible if power is cycled and communication begins within 33ms. Please see section 4.10 on p.88 of the PGA309 User's Guide: http://www.ti.com/lit/ug/sbou024b/sbou024b.pdf

    If you are not using 3-Wire mode, or if the OWD bit is set in your application, the EEPROM data is being changed by some other means.

    Best regards,

    Ian Williams
    Linear Applications Engineer
    Precision Analog - Op Amps

  • Hello Ian,

    I am programming using the Muliti-Cal-System for the complete calibration, but when I reload the calibration (Programing)  I am using just a USB DAQ board (PGA309EVM-USB) and the PGA309 EVM software.

    This has occurred in three of the 10 sensors built so far.  One ran for 4 hours and the other at 6.5 hours before they stopped working with this problem. The next day I reloaded the calibration into the  sensors, they worked and ran a full 8.5 hours with no trouble.

    The third sensor was working fine when I welded the end cap on to the electronics cylinder. The sensor was still warm when potting was added to the cylinder, I rechecked to verify operation still working as designed. Went home letting the potting setup and rechecked first thing in the morning and found not operational. Reloaded calibration and have not had a problem since.

    We are planning to replace the PGA309 with the PGA309HT and getting higher temperature components to accompany it. Even though we are not running that high of a temperature.  

    Thanks,

                Jim

  • Hi Jim,

    This is interesting information. When you say "the sensor was still warm when potting was added to the cylinder," do you mean the complete sensor module assembly is warm as a result of being placed in an oven, or the electronics themselves are warm? Over what temperature range are you operating the devices? Have you measured the current consumption of the circuit?

    Finally, register 8, the Alarm Monitor register, may contain more information if there's an error at an internal node. When your modules fail, can you read this register and see if you have a non-zero value? The error codes indicated by register 8 are detailed in the PGA309 User's Guide, section 2.8.

    Best regards,

    Ian Williams

  • Hi Ian,

                The sensor was warm due to the end cap of the sensor housing being welded on. The electronics do not get very warm on their own, but the environment they are subjected to is approx. 100°F. The current draw for the entire circuit is 8.2mA total. I have not had a failure since the first thread of this forum, I will check register 8 and record this for next time.

    Thanks,

                Jim  

  • Hi Jim,

    Thanks for the update, and keep me posted if you have any further issues.

    Best regards,

    Ian Williams

  • Hello Ian,

                We ran the first set of sensors on a field test and found sensors work well until unplugged. Then the sensors would not work until calibration was reinstalled, once calibration was reinstalled the sensor worked until unplugged again. This did not happen every time they were unplugged but it did happen at least 70% of the time. Should we be adding some type of “Watchdog” system?   Or what do think would work best with the PGA309?

    Also, We were unable to retrieve fault data.

    Thanks,

                Jim