Post Go back to editing

LTC2949 exhibits unstable VREF during grid connection, and abnormal large currents are detected during charging and discharging processes

Category: Hardware
Product Number: LTC2949

Hello, 

Using LTC2949 for battery monitoring, the system voltage is 870V, and the current is collected through a shunt resistor. Please assist in confirming the possible causes of the following issues and suggest some solutions? Thanks!

  1. During grid connection, the LTC2949's VREF reference voltage fluctuates, changing from 2.997V to 2.970V.
  2. After the inverter is connected to the grid and charges or discharges the battery, the shunt resistor (0.1mΩ resistance) occasionally detects a large current exceeding 200A (normally around 20A), lasting approximately 0.5 seconds, with a probability of occurrence of 5%. However, when using an oscilloscope current probe to measure the current on the charge and discharge lines, there is actually no such large current present.
  • During grid connection, the LTC2949's VREF reference voltage fluctuates, changing from 2.997V to 2.970V

    Is the fluctuation AC or a change in DC voltage?  If there are loads on the VREF pin the pin voltage may change as the load varies.  The internal reference voltage is 1.2V and multiplied by 2.5x then buffered before being output on the VREF pin.  If the link measurement is referenced to VREF to support negative voltage measurement then this would explain why it changes.

    After the inverter is connected to the grid and charges or discharges the battery, the shunt resistor (0.1mΩ resistance) occasionally detects a large current exceeding 200A (normally around 20A), lasting approximately 0.5 seconds, with a probability of occurrence of 5%. However, when using an oscilloscope current probe to measure the current on the charge and discharge lines, there is actually no such large current present.

    During this 0.5 seconds does the PEC match?  If it does are any of the FAULTS or EXTFAULTS registers set? 

  • Hi, 

    Thank you very much for your response!

    The cause of the VREF variation has not been confirmed yet. If subsequent calculations are based on the measured value of VREF rather than the theoretical value of 3V, this should help avoid errors.

    Currently the software does not implement the functionality to check PEC matching status and FAULT register status. The plan is to reduce the supply current from 12V to 5V and filter out this unreal current using software.

  • I suspect this is a communication problem in the software and that the PEC will provide more insight.  Without PEC the we can't be certain of the values in FAULTs and EXTFAULTS.