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.

UCD3138PSFBEVM-027: Problems with UCD3138PSFBEVM-027 (AUXOUT not output)

Part Number: UCD3138PSFBEVM-027
Other Parts Discussed in Thread: UCD3138CC64EVM-030, UCD3138, USB-TO-GPIO2, UCD3138A64CEVM-660, UCD3138A64, UCD3138128A, UCD3138128

Tool/software:

I am trying to evaluate the UCD3138PSFBEVM-027, but I have two questions about it.
① I am inputting 400V, but the AUX output from PWR050 is not outputting properly, and D26 keeps flashing.
Is there a problem with the setup?
*No load is connected yet.
② I connected a USB-TO-GPIO to the UCD3138CC64EVM-030 and started "Fusion digital power studio", but the message "No Devices Found!" appears and I cannot connect.
Is there any setting that needs to be done?

  • Hello Koumei,

    1) PWR050

    Yes, we have had some problems with the PWR050's reliability as an auxiliary power supply. The PWR050 should work regardless if the UCD control card (UCD3138CC64EVM-030) is connected or not. Try disconnecting the UCD3138CC64EVM-030 control card and make sure that the PWR050 is outputting the correct voltages based on the input (Ex. PIN6 of the PWR050 should be proportional to the input voltage). Adding some resistive load on the PWR050 outputs may also help the performance.

    I have attached the schematic, PCB and BoM of the PWR050 to help with your debug.

    7848.PWR050A_SCH.pdf  PWR050A_PCB.pdf  PWR050A_BOM.pdf

    2) Control Card Power

    Are you supplying power on the control card to the UCD3138 via the USB-TO-GPIO2 adapter? You must populate the J2 jumper to supply a 3.3VD power from the USB-TO-GPIO2 adapter to the UCD3138.

    Regards,

    Jonathan Wong

  • Hello Jonathan

    Thank you for the information on the PWR050.

    After asking the question, I did some more research and found that lowering the input voltage to 395V resulted in normal output. For now, I will continue with this evaluation.

    I will investigate the control card a bit more.

    By the way, is the UCD3138-PSFB firmware downloaded to the UCD3138CC64EVM-030 control card immediately after purchasing the UCD3138PSFBEVM-027?

  • Hello Jonathan

    A new problem has occurred and I can no longer detect the ROM. I would like to know how to solve it.

    Here is what I did:

    1) I downloaded "UCD3138_PSFB_PCM.x0" obtained by compiling "UCD3138_PSFB_PCM" with CCS from "Firmware Download" in UCD3XXX DeviceGUI and confirmed that it works.

    2) I wanted to start the firmware as soon as the power was turned on, so I changed the download options and downloaded it.

    3) I wanted to return it to its original state, so I checked "Erace data flash" and downloaded it.

    I don't remember the detailed settings very well, but is it possible to return it to the state it was in immediately after purchase?

  • Hello Koumei,

    Do you have a screenshot of the UCD3xxx Device GUI and the error message you receive?

    If you are currently running the UCD3138 in flash mode, then you can click on DEVICE_ID to ensure communication between the UCD3xxx GUI and the UCD3138. If the UCD3138 is in flash mode, then you can click on the "Command Program to jump to ROM (SendByte 0xD9 to Address 88)" to get the UCD3138 back in ROM mode.

    The UCD3138CC64EVM-030 that comes with the UCD3138PSFBEVM-027 automatically has the UCD PSFB firmware pre-installed. You can download it again using the UCD3xxx Device GUI. You can find the UCD PSFB firmware here: https://www.ti.com/tool/UCD3138FW-PSFB.

    Regards,

    Jonathan Wong 

  • Hello Jonathan

    Thank you for your reply.

    I am attaching a screen capture of when I communicated with UCD3138CC64EVM-030 using UCD3XXX Device GUI.

    The screen capture below was taken when I started UCD3XXX Device GUI and clicked "Scan Device ROM".

    The other one is a screen capture of when I clicked DEVICE_ID.

    It seems like I am not able to communicate with UCD3138 at all. Has it fallen into an unrecoverable state?

  • Hello Koumei,

    Can you take a picture of your whole debug setup? Please have a picture that shows your USB-TO-GPIO2 adapter connected to the UCD control card so I can confirm the setup is correct.

    It looks like the UCD is not responding to any PMBus commands. If the UCD was operating correctly, then it should respond in either ROM mode or FLASH mode. The UCD may be locked in flash mode. Or, you may not be correctly supplying power to the UCD on the control card. Did you populate the J2 jumper like in the first post? Do you have another UCD chip to confirm that your setup is correct?

    Regards,

    Jonathan Wong

  • こんにちは、ジョナサン

    お返事ありがとうございます。

    セットアップ状態の写真をお送りします。
    この状態でROMモードで一度通信できましたので問題ないと思います。

    「フラッシュモードではUCDがロックされる可能性があります」とありますが、フラッシュモードでロックされてしまうと元の状態に戻すことは不可能なのでしょうか?

  • Hello Koumei,

    Adding translation here for searchability:

    Hello, Jonathan

    Thank you for your reply.

    We will send you a picture of the setup status.
    I was able to communicate once in ROM mode in this state, so I think there is no problem.

    "The UCD may be locked in flash mode," but is it impossible to restore it to its original state once it is locked in flash mode?

    Thank you for sending a picture of your setup. The setup looks correct. Note that you should not have both J2 and J6 connected at the same time if you have the UCD control card connected to your EVM. Otherwise, power will be coming from both the EVM and the USB-TO-GPIO2 adapter. Right now, it is fine since you do not have an EVM connected. 

    The UCD is only locked in flash mode if your firmware somehow gets stuck (Ex. in a while loop) and can never reach a code that would allow the UCD to jump back to ROM mode (Ex. "clear_integrity_word() or pmbus_handler()"). As long as you have robust code or a hardware backdoor, then it is unlikely the UCD will be locked in flash mode. However, if your UCD does become locked in flash mode, then there is not much you can do to restore to the original state. You may need a new chip after that.

    Regards,

    Jonathan Wong

  • Hello Jonathan

    Thank you for your reply.

    There doesn't seem to be any hope of recovery in the current state. I will try replacing it with a new chip.

    I think I may have made a mistake in my operation and written the checksum.

    I have two questions.

    1) Could you tell me if there are any precautions to take when installing a new chip?

    2) I believe the same firmware will work with the UCD3138A64CEVM-660, but is that correct?

  • Hello Koumei,

    1) Could you tell me if there are any precautions to take when installing a new chip?

    There are two precautions that you can do to prevent the UCD from getting locked in flash mode.

    The first precaution is to make sure you select the "DO NOT write program checksum (Stay in ROM)" option when downloading the firmware onto the UCD. This means that anytime the UCD is power cycled (turn off and on), the UCD will start in ROM mode and must be commanded manually to execute its firmware. This ensures that if the UCD is locked in flash, then when you power cycle, the UCD will return to ROM mode and you can download a new safe firmware.

    If you are ready for a production firmware, then you can select the "WRITE program checksum (Automatically execute)" option. This means that your firmware will automatically execute when the UCD is power cycled. Thus, when you have a stable firmware that is ready for production, you can select the second option and your firmware will always execute.

    The second precaution is to include a hardware backdoor at the beginning of the main function. The UCD will be locked in flash mode if the UCD is stuck in some while loop or condition in the main code and thus cannot execute any code to return to the ROM state. By adding a hardware backdoor, you can ensure that the UCD will be able to jump back to ROM mode. This is typically done by having some pin be shorted to GND that, if the pin is pulled to 3.3V, will execute a jump to ROM command (Ex. "clear_integrity_word()") after power cycling.

    The PSFB firmware has a hardware backdoor but it is commented out, as shown below. You can reactivate the hardware backdoor by uncommenting the "pmbus_read_write_rom_mode(PMBUS_WRITE)" function. Thus, if you tie FAULT3 and FAULT2 to 3.3V, then your UCD will automatically jump to ROM mode anytime the UCD is power cycled since it is the first action the UCD takes in the main function and will not have the chance to be stuck in a condition.

    You can learn more about the hardware backdoor on this E2E post: https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1356278/ucd3138allcevm150-fault-pins-assignment-in-llc-evm-firmware 

    2) I believe the same firmware will work with the UCD3138A64CEVM-660, but is that correct?

    Yes, the PSFB firmware will work on the UCD3138A64CECM-660. Just make sure to select the build option for the correct part number. The UCD3138A64CEVM-660 comes with the UCD3138A64 installed but is also compatible with any 80-pin UCD variant (UCD3138A64, UCD3138128 and UCD3138128A).

    Regards,

    Jonathan Wong

  • Hello Jonathan

    Thank you for your reply.

    We will continue our evaluation based on the information you provided. Thank you for your support.