• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 63 subscribers
  • Views 280 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.

LM5066I: Vref is out of spec. - 4.1V

user6504975
Prodigy 50 points
Part Number: LM5066I

Dear Support,

In my design I use multiple LM5066I devices on one board. I have attached the schematics of one LM5066I circuit. 

I am able to communicate with each Hotswap controller over I2C, and switch ON/OFF is possible via UVLO/EN pin.

The issue is with the measured voltage and current. In case of voltage I measure 20.3V instead of 28V input voltage.

I checked the internal register value, converted it and I receive 20.3V as the GUI shows.

After further measurements I realized that Vref pin is out of specs - 4.1 V instead of max. 3V. I am sure this is the root cause of the wrong input voltage measurement.

Beside the higher voltage of Vref, the ripple on the 1uF capacitor is high. 

I have attached a few Oscilloscope figures. I measured the Vref in AC and DC mode as well, on my custom board and on LM5066IEVM-626 Evaluation Module.

Do you have any suggestion? Could you explain what causes a higher Vref voltage? I have 12 LM5066I controllers on the board, and all of them have 4.1V on Vref pins. 

Thank you for your support.

Best regards,

Alpar

  • 0
    Prodigy 50 points

    Vref on my board - measured on C504 - Vref = 4.1V

    Vref on LM5066IEVM-626 Evaluation Module. Vref = 2.92V

  • 0 in reply to user6504975
    TI__Guru** 105361 points

    Hi Alpar,

    Thanks for reaching out!

    Have you done the ABA swap test to confirm whether the issue is related to the board or device ? Looks like some offset on your board causing the issue.

    Are these voltage measurements are across C504 capacitor with respect to LM5066i ground. ?

    Best Regards,

    Rakesh

  • 0 in reply to Rakesh
    Prodigy 50 points

    Hi Rakesh,

    The ABA swap test was not performed yet. 

    The attached measurements were performed across C504 (see attached schematic), with the "paperclip" method to minimize GND loop. The ripple of Vref measured on C504 had a frequency of ~1kHz. (There is one other buck converter (28V-->5V) on the board, but it's switching frequency is 700 kHz, so this is something else.)

    The attached SCHEMATIC is the circuit of 1 Hotswap controller on my board. In total there are 12 Hotswap controllers. GND/AGND, 28V_VIN, I2C lines are common for each LM5066I.

    Each Hotswap IC on the board has the described Vref issue: 4.1V..4.2 V instead of max. 3.02V, with high ripple on it. ( I checked multiple boards from the same production as well)

    On the other hand Hotswap controllers are communicating over I2C, OV/UV protection works as expected. Only the measurements are not correct, due to Vref issue. (Measured VIN is 20.3V instead of 28V; Measured output current 2.6 A instead of 3.6 A)

    I tried to add smaller(330nF) and higher (10uF) Vref capacitor to the circuit. With 10 uF the ripple disappeared, but the Vref voltage was still around 4V. Smaller capacitance was not helping eather.

    VDD line (4V85_1) is not noisy, and the voltage is ~ 4.9V which seems OK.

    I suppose the internal regulator for Vref is using the VDD line as a supply. 

    Have you seen something similar before? Do you have any other suggestion? 

    Thank you!

    Best regards,

    Alpar 

  • 0 in reply to user6504975
    TI__Guru** 105361 points

    Hi Alpar,

    We have not seen this kind of behavior in the past. 

    Please test that sample on EVM. If the same behavior is observed, please send the unit to us for further analysis.

    Best Regards,

    Rakesh

  • +1 in reply to Rakesh
    Prodigy 50 points

    Hi Rakesh,

    Meantime I have found the root cause of the issue. 

    Components for temperature measurement were not mounted to the board, and DIODE pin was floating. This caused the Vref issue. Pulling DIODE pin to GND solved the issue.

    "If the temperature measurement and protection capability of the LM5066I are not used, the DIODE pin should be grounded."

    I think "should" needs to be changed to "must" in the datasheet.

    Thanks for the support!

    Alpar

  • 0 in reply to user6504975
    TI__Guru** 105361 points

    Hi Alpar,

    Good to hear that.

    Best Regards, Rakesh