Post Go back to editing

ADRV9002 POWER DOWN ISSUE

Category: Software
Product Number: ADRV9002

Hi,

I am using CUSTOM ADRV9002 board with ZCU102 EVAL board.

In our board we are using TWO ADRV9002 ICs.

There is no problem , while working with individual ADRV transceiver,

but when I am trying  to program two ADRV transceivers at a time I am getting voltage drop and load current is increasing.

After that the board is not working .

How to solve this issue ?



moved and updated form
[edited by: GenevaCooper at 1:07 PM (GMT -4) on 26 Sep 2023]
Parents
  • Hi LAXMIPRASANNA,

    Please verify that your power supplies remain within tolerances at all times while programming the parts.

    Kind Regards,
    Michał

  • Hi,

    All the power supplies are within range.

    The power supply sequence for our board is attached here

  • Hi LAXMIPRASANNA,

    That's an ambiguous answer. Can you verify (e.g. with an oscilloscope) that during programming, the supplies don't drop below their tolerance?

    Any filtering (including ferrite beads), sense resistors, and trace resistance between the supply output (or the feedback point) and the ADRV9002 input can cause a DC voltage drop. So for example, if at idle your 1.3 V supply is exactly 1.300 V,  it only takes 0.035 Ohms of DC resistance to bring it out of spec under a 1 A load.

    Kind Regards,
    Michał

  • Hi,

    Thanks for your response.

    I have checked voltages before ADRV9002, all are as expected.

    In our board there are TWO ADRV devices, 

    While working with one ADRV9002 device, every thing is working fine and getting results as expected.

    for second ADRV9002 also got same results.

    But if I program TWO ADRV9002 devices one by one, (working in NOOS)

    after two minutes of programming the voltage drop has been occured.

    Is there any cause for POWER DROP in software configuration?

  • Hi LAXMIPRASANNA,

    Can you be more specific about what you mean when you say "voltage drop" and "power drop"? Which voltage (and power) are you referring to, and how are you measuring it?

    Kind Regards,
    Michał

  • Hi,

    I am using custom ADRV board with EVALZCU102 board .

    I am giving power supply for adrv9002 externally through DC power supply (12V and 1.5A. our board supports upto 6A current )

    And observed current consumption in DC POWER SUPPLY 

    While working with two adrv9002 devices,

    Initial static current consumption is 0.16A observed in DC power supply.

    after programming (in NOOS)  two adrv9002 devices the consumption is 0.6A .

    But after two minutes of programming , the device voltage is dropped by 3V in DC power supply .

    After that Observed short in ADRV9002 power rails(initially it is not short, Have checked ).

  • Hi LAXMIPRASANNA,

    I have never seen this happen on the ADRV9002 before. Are all of the voltage rails in the ADRV9002 at ground when the failure occurs?
    I think it might warrant a second look at your power solution:

    1. According to its datasheet, the ADP1762 requires a minimum supply voltage of 1.1 V. Here, you state that you supply it with 1.0 V

    2. According to its datasheet, the maximum current output current of the ADP1754 is 1.2 A. It looks like you built your design expecting it to supply 2.0 A

    3. Only the first two channels of the ADP5054 are capable of supplying 6.0 A of current per channel. Channels 3 and 4 can only supply up to 2.5 A per channel. It looks like your design expects channels 3 and 4 to supply 6.0 A, the same as channels 1 and 2

    Since you are using No-OS, I can't help you with debugging from the software side. In the SDK that we supply, we have two APIs for running a basic system test (in-system debugging) to verify that basic functions are operating as expected: adi_adrv9001_Utilities_SystemDebugPreCalibrate(), and adi_adrv9001_Utilities_SystemDebugPostCalibrate(). It may be worthwhile checking if No-OS has anything similar, and trying to run it on your system.

    Kind Regards,
    Michał 

  • Hi,

    Yes, All of the voltage rails in the ADRV9002 at ground when the failure occurs.

    The updated power supply sequence for our board is attached here

  • Hi LAXMIPRASANNA,

    A failure mode that manifests by shorting all supplies to ground is usually permanent as it would cause significant and permanent internal damage, and this would not something that you can fix by simply resetting the part and programming it again. At least, not on the ADRV9002. Therefore I must maintain that the issue is likely external, and probably caused by the power supply.

    It's clear to me that your power supply solution requires more work. You fixed the issues that I pointed out in my previous reply, but did not pay attention to any other issues in your design. For example, the ADP1754 requires an input voltage at least 0.4 V above the output voltage, whereas you're supplying it with 2.1 V and expecting 1.8 V output.

    This forum is intended to support customers with challenges that they encounter when using the ADRV9002/3/4, and debugging your power delivery solution is outside the scope of what we offer. Therefore, I cannot help you until you can guarantee that the power supply that you have built is capable of supplying the needs of two ADRV9002s.

    Kind Regards,
    Michał

  • Hi,

    Individually our adrv9002 is working properly. We are generating all supplies properly on the board. When programmed at a time SPI was configured successfully. During Debugging we found short in 1.3V path all other power supply paths did not have an issues. As per our design ADP1762 1.3V LDO supports up to 2A. We are working on TX1, RX1 and RX2 for both adrv9002 at a time in FDD mode. Our 1.3V  layout path supports up to 2.5A at 40 C temperature.

    The required current for 1.3V is 1.234A for 2 TX, 2RX we think 2 amps is sufficient for this path.

    While working at a time on both ADRV9002 only ADP5054 is common for both ADRV9002 in supply path and for 1.3V path it gives up to 6A. Both 1.3V LDO's current will require up to total 4A. ADP5054 is able to support both the ADRV9002 at a time. The difference in evaluation and our design is that in evaluation schematic 1.3V supply  is directly given  from DC-DC Converter instead we are giving through LDO does it cause any issues due to this approach? or we should go for mutually independent supply path for both ADRV9002?

Reply
  • Hi,

    Individually our adrv9002 is working properly. We are generating all supplies properly on the board. When programmed at a time SPI was configured successfully. During Debugging we found short in 1.3V path all other power supply paths did not have an issues. As per our design ADP1762 1.3V LDO supports up to 2A. We are working on TX1, RX1 and RX2 for both adrv9002 at a time in FDD mode. Our 1.3V  layout path supports up to 2.5A at 40 C temperature.

    The required current for 1.3V is 1.234A for 2 TX, 2RX we think 2 amps is sufficient for this path.

    While working at a time on both ADRV9002 only ADP5054 is common for both ADRV9002 in supply path and for 1.3V path it gives up to 6A. Both 1.3V LDO's current will require up to total 4A. ADP5054 is able to support both the ADRV9002 at a time. The difference in evaluation and our design is that in evaluation schematic 1.3V supply  is directly given  from DC-DC Converter instead we are giving through LDO does it cause any issues due to this approach? or we should go for mutually independent supply path for both ADRV9002?

Children
  • Hi LAXMIPRASANNA,

    If your current supply individually works for one ADRV9002 at a time, then yes, it sounds like two separate solutions might work. Ultimately however, it depends on how carefully it is designed, and if it can stay within tolerances while supplying both minimum and maximum load current, and also at transients.

    One last question, You mentioned that you are using the ZCU102, and you are supplying 3 A at 12 V to your dual ADRV9002 board. Is this 12 V / 3 A supply coming from the FMC connector?

    Kind Regards,
    Michał

  • Hi,

    Thanks for your response.

    We are giving supply to our adrv board from external DC power supply not through FMC connector.