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.

UCD3028 Clear Faults PMBus command (0x03) Issue

Other Parts Discussed in Thread: UCD3028

Hello, one of my customers is experiencing an anomaly when I execute a ‘Clear Faults’ PMBus command (0x03) to TI’s UCD3028 controller (please refer Additional Issue" information below, which descibes the error and bus captured data using an I2C Bus Analyzer as well as the TI’s SMBus/I2C/SAA tool (GUI).

Summary:

Unable to clear fault/warning status bits even though the cause was eliminated (ie., moving limits back outside normal operating regions). Additionally, although the UCD3028 controller does ACK the ‘Clear Faults’ PMBus command (0x03), it NACKs the very next PMBus transaction to it. The UCD3028 controller has the exact same response regardless of whether it’s our GUI or TI’s GUI interface. The problem persists even if the Intermediate Bus power device’s out is disabled.

Are you aware of this problem? If so, do you have a solution, perhaps a UCD3028 controller FW update or different sequence of PMBus commands?

Also, is it possible to receive a list of errata regarding UCD3028 PMBus operations?

Additional Issue information:

Issue:

1.       Unable to CLEAR a warning or fault flags.

a.       Using the CUI GUI

b.       Set OT_WARNING to 0°C in DevCfg tab

c.       Device and LSB Fault/Warning LEDs were asserted

d.       Reset OT_WARNING to 115°C in DevCfg tab

e.       Clicked the ‘Clear Device Faults’ in DevCfg Tab, I2C Bus Capture

 Expected Results:        [Start]#36[W][A]03[A][Stop]

Unexpected:        [Start]#36[W][N][Stop]

Unexpected:        [Start]#36[W][N][Stop]

 f.        While board remained powered, closed the CUI GUI and invoked TI SMBus/I2C/SAA tool.

g.       Executed a SendByte with no data to IBC using PMBus command 0x03.

 07:18:00.080: SAA #1: SendByte (Address 27d, Cmd 0x03): ACK

 h.       Initiated a WordRead of IBC’s Status Word using PMBus command 0x79

 07:18:40.212: SAA #1: ReadWord (Address 27d, Cmd 0x79): NACK

07:18:54.658: SAA #1: ReadWord (Address 27d, Cmd 0x79): ACK 0x0006

 What’s interesting to note is that for some reason, although the TI Controller ACKs the ‘Clear Faults’ command (0x03), the Warning/Fault does NOT get cleared. Moreover, both the TI GUI and the I2C bus analyzer verify the TI Controller’s NACK of the next PMBus command (command type irrelevant). Unable to execute using Ericsson’s GUI because they don’t have a ‘Clear Faults’ option.

Additionally, using the TI SMBus/I2C/SAA tool, I’m able to successfully execute a ‘Clear Faults’ command followed with ACKs for any other immediate PMBus command on a NDM2Z-25H POL.

07:38:01.241: SAA #1: SendByte (Address 4d, Cmd 0x03): ACK

07:38:02.791: SAA #1: SendByte (Address 4d, Cmd 0x03): ACK

07:38:04.192: SAA #1: SendByte (Address 4d, Cmd 0x03): ACK

07:38:04.696: SAA #1: SendByte (Address 4d, Cmd 0x03): ACK

07:38:06.072: SAA #1: ReadWord (Address 4d, Cmd 0x79): ACK 0x0840

07:38:06.745: SAA #1: ReadWord (Address 4d, Cmd 0x79): ACK 0x0840

07:38:07.233: SAA #1: ReadWord (Address 4d, Cmd 0x79): ACK 0x0840

 Questions:

  • Why is the TI Controller not clearing the Faults and Warnings with the ‘Clear Faults’ PMBus command?

Why is the TI Controller NACKing the very next PMBus transaction?

Hope to receive a detailed response from the UCD3028 TI Apps team.

Thx,

Leo

  • Leo,

    Is this the same issue as what was posted here: UCD3028 Fault Processing Issue?

    Please advise. Thanks!

    Regards,

    Brandon Vonk

  • Hi Brandon,

    There was no such issue you described, and the link provided is broken. I'm inclined to say that the initial answer to your question is no...

    Leo

    Leo Champagne

    Sr. Field Application Engineer

    Arrow Electronics

     

  • Leo,

    Sorry, Richard Elmquist came to us with this issue a few days ago. Please sync up with him. He was going to go back to the customer to clarify some information including the following:

    What is the CUI GUI? What firmware is on the UCD3028?  Is it for use in one of our EVMs? If it is our EVM code, we will need to know which code it is.

    Regards,

    Brandon Vonk

  • Team,

    Here's addtional information from the customer regarding the questions above:

    • CUI GUI: customized UI for accessing and controlling various POLs and IBCs (Intermediate Bus Components) using PMBus commands propagating across the localized SMBus. One of the features of the GUI is to allow the user to clear POL and IBC warnings and/or faults. This is accomplished by the GUI sending the PMBus command 'Clear Faults' (0x03) along with the address of the POL or IBC.

     

    • What firmware is on the UCD3028? The UCD3028 controller is embedded in a 3rd party design for an IBC. They'd be happy to provide the information but are unsure how to best get that information. Is it contained within the controller's non-volatile memory? If so, what PMBus command register should be accessed?

    Any feedback here would be appreciated.

     

    Thanks,

    Fern

  • The PMBus commands on the UCD3028 are entirely firmware controlled.  We do provide an example set of PMBus commands in our EVMs, but we have no control over how our customers use them.  We often have no knowledge which commands they implement or the details of the implementation.  If we do know, we are often constrained by non-disclosure agreements.  Since this is a third party code, not one of our EVMs, we can't really provide any help directly to CUI.  We could help the third party. 

    Sorry I can't be more helpful,

    Ian

  • All,

    It was discovered that the 3rd party SW/FW combination requires a PEC checksum data byte to be sent with any PMBus "SEND BYTE" COMMAND.

    This explaines what was captured on the SMBus analyzer