• State TI Thinks Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 534 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

AMC131M02: SEC_FAIL error bit frequently set

Franz Boeni
Intellectual 545 points
Part Number: AMC131M02
Other Parts Discussed in Thread: AMC131M01

Tool/software:

Hi team

In my design a use 9 AMC131M0x (I have 1 - 3 channels per chip) 8 chips run more or less well but 1 AMC131M02 has frequently (about 90% of reading) the SEC_fail Bit set in the status of chip. On the same SPI (shared with individual CS) I have 4 other AMC chips which don't have any problem. All these AMC chips on the same SPI bus has 100% equal setting. I tried follow things:

  • different CKIN frequencies Result: 7.8MHz around 90% error rate, 4.76MHz around 85% error rate
  • different SPI speed:  Testet from 3.1MHz - 12.5MHz: result no difference
  • I add additional capacitors on primary and secondary side according data sheet application in 1206 housings --> no difference
  • I add 1k load to SPI CLK, SPI MOSI and SPI MISO. no difference
  • different parameters DCDC_Freq: Best result with default value 0000. If reduce the frequency I in increase the number of fails und under around 3MHz there are only errors

Schematic of AMC131M02 with problems

Initialization at power up:

MODE_ADDRESS: (MODE_DEFAULT & ~(MODE_RX_CRC_EN_MASK | MODE_RESET_MASK))  | MODE_RX_CRC_EN_ENABLED | MODE_RESET_NO_RESET
CLOCK_ADDRESS: (CLOCK_DEFAULT_2CH & ~CLOCK_OSR_MASK) | CLOCK_OSR_2048
DCDC_CTRL_ADDRESS: DCDC_CTRL_DEFAULT
GAIN1_ADDRESS: GAIN1_DEFAULT
CFG_ADDRESS: CFG_DEFAULT | CFG_GC_EN_ENABLED | CFG_GPO_DAT_HIGH | CFG_GPO_EN_ENABLED

Reading ADC signal AMC chips:
Every 100us the status is read. In case DRDY0 is active (new data available) the ADC register is read.

The description of SEC_FAIL is a little poor. Maybe you can give me more information about this error. Maybe this will give me a hint to find what is the problem source.

Thank you for helping me.

Franz

  • 0
    TI__Expert 6700 points

    Hi Franz,

    Could you provide some details on your application and use case?   What is being measured by the devices?  What kind of high frequency/high voltage transients are present in the system?

    The SEC_FAIL bit is set based on many internal checks on the device.  It indicates some issue with the high-side supply.  This could be due to issues on the high side in which the supply is not stable or is fluctuating a lot.  It could also be an issue on the low side, if the low side supply dips enough a sec fail could occur.  Are you able to monitor the high and low side supply of the failing device?

    Thanks.

  • 0 in reply to Saleh Ahmad
    Intellectual 545 points

    Hi Saleh
    Thank you for your fast replay.

    Red curve: High side voltage Pin 5 - Pin 1 ripple under 0.1Vpp
    Blue curve: Low side voltage Pin 15 - Pin 13 ripple approx. 0.25Vpp

    Layout of AMC131M02 with SEC_fail problem:
    red top side, yellow and blue inner side layer. No bottom side layer, because of high volt insulation. below chip air gap.

    I tried to add different capacitor on low and high side voltage. AMC131M02 with problem has the longest distance on the board for the low side voltage and also for the SPI Bus of the group. The group on the same SPI bus has 5 AMC131m0x chips. 
    But I have another AMC131M01 chip on external board. This board is connected to main board with 1.27mm flat cable, length 30cm. This is a single AMC has his own SPI bus, without sharing with another chip. This AMC chip on another board has for sure longer distance for SPI bus and also for low side voltage. This AMC on external board don't have any problem.
    Has the number of SPI nodes on the same bus an influence to the longest distance of SPI bus? Interesting is a lower SPI CLK don't improve problem.

    The application is a BMS (battery management system), mostly used in home solar application. But for the moment I work only with 50VDC power supply. (no battery, no load) The problem happen without any load or any high voltage. It is not depending of any current.
    So we don't have any transient on the input yet. The AMC131 with problem measure the insulation resistance from battery to ground.

    Hope this will give you a hint.
    Thanks
    Franz