Announcement

Collapse
No announcement yet.

820-00840-A CD3215 short

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    820-00840-A CD3215 short

    Is the CD3215 chip bad if it has pp3v3_upc_xb_ldo at 2.6v and 5.5ohm resistance? I removed C3208 and the ROM chip and the short did not go away. Just checking before I remove the CD3215.

    Meter shows board is pulling 5V 330mA
    Last edited by powerblackout; 01-19-2022, 03:43 PM.

    #2
    Re: 820-00840-A CD3215 short

    Place meter in diode mode.

    Remove all power to the logic board.

    Red meter probe to ground. Black meter probe to each LDO output of CD3215. Post the readings for each output and CD3215.

    Comment


      #3
      Re: 820-00840-A CD3215 short

      Yep it was indeed bad. I replaced it and am now getting every LDO except for 1.1V. Repair wiki says there could be communication issue between the ROM chip and the CD3215.

      Voltage in MISO is 3.3V, but MOSI is anywhere from 0.8V to 1.2V. Shouldn't both of these be the same voltage for communication to work?

      I removed resistor on MOSI line, still same voltage coming out of CD3215. no short to ground detected between CD3215 and ROM chip.

      Comment


        #4
        Re: 820-00840-A CD3215 short

        It is possible that your instrument (meter?) is unable to properly measure this signal line. Keep in mind that it is a serial communication device interface.

        The line could very well be toggling high(3v3) and low (0v) and the multimeter is unable to properly measure this line. An oscilloscope will be the proper tool to use here or a low cost logic analyzer (some are ~$10-$20 on Aliexpress / Amazon).

        In either case, do you have the proper LDO voltages on every CD3215 device ?

        Ironically on the CD3215 rom, there is a new widget in Asia that offers the full tools to reflash the firmware. Was trying to get it shipped out before CNY but may be too late for now.

        With the replacement, what are the results for the power delivery ? Your box remains @ 5V ? What is the current draw ?

        Comment


          #5
          Re: 820-00840-A CD3215 short

          TBT/USB-C ROM are standard 25-series SPI EEPROM, no need for anything fancy to flash them, just correct adapter (socket or board) and the dump (can be different between boards and sides, still there's some cross-compatibility).
          OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

          Comment


            #6
            Re: 820-00840-A CD3215 short

            Yes I get proper LDO voltages out of both chips, except for the 1.1V LDO. I'll hook my scope up and see if any traffic is happening.

            All ports are currently 5.17V 0.01A pull. They were close to 5.05V 0.35A before the CD3215 replacement, with the CD3215 getting pretty warm compared to the rest of the board.

            Comment


              #7
              Re: 820-00840-A CD3215 short

              Ok measured with the scope, I see a bit of traffic happening right at powerup, but none after, sitting at 0.8V - 1.2V

              I plugged the ROM into my reader, I was able to dump it, meaning that the chip itself works fine?
              Attached Files
              Last edited by powerblackout; 01-24-2022, 01:03 PM.

              Comment


                #8
                Re: 820-00840-A CD3215 short

                TBT/USB-C ROM are standard 25-series SPI EEPROM, no need for anything fancy to flash them, just correct adapter (socket or board) and the dump (can be different between boards and sides, still there's some cross-compatibility).
                @Piernov, thanks. Is there a public database of dumps for these models? Thinking it may be worthwhile for the tool for access to the flash dumps.

                Comment


                  #9
                  Re: 820-00840-A CD3215 short

                  with the CD3215 getting pretty warm compared to the rest of the board.
                  Any chance of soldering issues with the replacement ? Lots of flux + hot air and make the BGA dance with a gentle nudge of the tweezers to remove any possible short underneath the device.

                  Comment


                    #10
                    Re: 820-00840-A CD3215 short

                    @mon2
                    aren't these CD3215s supposed to be replaced in pairs from my recall?
                    Also pay special attention to the actual part number:
                    CD3215B03 will not work in place of a CD3215C00

                    Comment


                      #11
                      Re: 820-00840-A CD3215 short

                      Yes, mine are CD3215C00 and I replaced them with the same model #

                      It's certainly possible I didn't install it correctly, this is the first CD3215 I have done.

                      I ordered 2 off of ebay https://www.ebay.com/itm/193840530946
                      I tried reflowing the first replacement, then removed it and tried the 2nd replacement. Very careful to clear all solder from pads, using amtech flux, heating until they dance, giving a very very gentle nudge. Both replacements have behaved the same.

                      I have a CD3215 stencil ordered but it is not here yet.

                      Could the ROM be corrupt? Where could I find a replacement ROM file to flash?
                      Last edited by powerblackout; 01-24-2022, 01:54 PM.

                      Comment


                        #12
                        Re: 820-00840-A CD3215 short

                        Personally do not have enough experience with the CD3215 / 3217 devices, yet. From assorted readings, it is a custom Type C controller deployed by TI. TI has many public documents on similar devices.

                        Can the firmware not be applied over the CCx lines ? Did read about the UART interface.

                        Will be really nice to not have to desolder the flash device and just program it in-circuit.

                        On the firmware / ROM corruption - my thoughts exactly but not sure if there are public dumps of the flash for the assorted models. The tool in Asia is likely the usual flash programmer but I think they also implanted the mix of the flash dumps. Unless shown otherwise, it may be a good investment - just to get our hands on the flash rom dumps.

                        However, the heating up of the CD3215 should be investigated as this may be a soldering issue.

                        Comment


                          #13
                          Re: 820-00840-A CD3215 short

                          That was the original issue, the CD3215 heating up. After replacement it no longer heats up, 5V 0.01A on the ammeter.

                          Am I correct that PPBUS_G3H should be 0V without the battery connected at this point?

                          Comment


                            #14
                            Re: 820-00840-A CD3215 short

                            What is PPBUS_G3H with the external Type C adapter connected ? Leave off the battery ?

                            Is it 0V ?

                            What is PPDCIN_G3H with the external power adapter ?

                            Remove all power. Meter in resistance mode (2k is ok).

                            Check & post the resistance to ground for F7000. Is it a low value ?

                            If yes, then you have a bad cap on this rail and is causing U7000 to turn OFF.

                            If you see a low resistance on PPBUS_G3H rail, consider to remove F7000 off the board. Then check the resistance again on the pads of F7000.

                            Is the low resistance on the PRODUCER side (ie. U7000) or the CONSUMER side (downstream rail = F7000, pad # 2) ?

                            If you find the resistance is OK (not too low) on F7000, then check the DC voltage on:

                            SMB_SMC_CHGR_SDA
                            SMB_SMC_CHGR_SCL


                            These are the I2C (SMBUS) communication lines used to configure U7000.

                            Comment


                              #15
                              Re: 820-00840-A CD3215 short

                              PPBUS_G3H is 0V when the battery is disconnected and Type C adapter connected. 2.3Mohm resistance when disconnected
                              PPDCIN_G3H is 5.18V with external power. 4.8kohm resistance when disconnected
                              F7000 is 0.6Mohm resistance to ground on both ends.

                              My board doesn't have SMB_SMC_CHGR_SDA or SMB_SMC_CHGR_SCL
                              I'm not familiar enough to know if that is a typo or if I'm searching incorrectly lol!

                              EDIT: PPVIN_G3H_P3V3G3H is 4.8v while the PPCIN_G3H_CHGR_R is 5.17v, is that a problem, the voltage drop across the diode?

                              EDIT 2: Reflowed U7000 now getting 0.1V on PPBUS_G3H and PPVIN_G3H_P3V3G3H is 5.05V
                              Last edited by powerblackout; 01-24-2022, 03:51 PM.

                              Comment


                                #16
                                Re: 820-00840-A CD3215 short

                                These are good readings. No shorts on the inspected rails.

                                The SMBUS (I2C) interface pins are linked to the U7000 regulator.

                                See attached. Proceed with great caution as these are very tiny parts but you should be able to locate a sweet spot to take your measurements.

                                To confirm, your board is 820-00840-A ?

                                See attached on some test pads on this logic board for the SDA & SCL measurements.

                                Attach your external Type C adapter and very carefully check the voltages for the SCL and the SDA pads. Post their values.
                                Attached Files

                                Comment


                                  #17
                                  Re: 820-00840-A CD3215 short

                                  PPVIN_G3H_P3V3G3H is 4.8v while the PPCIN_G3H_CHGR_R is 5.17v, is that a problem, the voltage drop across the diode?
                                  Yes, a diode does have a voltage drop. Not uncommon to have upto 0v6 of a drop from input to output.

                                  As an alternate confirmation of the diode area, remove all power - meter in resistance mode (2k is ok). Check the resistance to ground of PPVIN_G3H_P3V3G3H. If the reading is low then there could be a shorted cap here but you have 4v8 so that is not a bad reading.

                                  Next, configure your meter in DIODE mode.

                                  Black meter lead on pin #3 of D6905; Red meter lead on pin # 1.
                                  Then reverse the meter leads.

                                  Check again with pin # 2 of D6905 & #1 of the same dual diode.

                                  Post each reading.

                                  Finally, confirm the VOLTAGE reading @ PP3V3_G3H which is the output of L6995, pin # 2.

                                  Comment


                                    #18
                                    Re: 820-00840-A CD3215 short

                                    Yep 820-00840-A

                                    Oh It's SMBUS not SMB my bad.

                                    Both SDA and SCL are 0v

                                    PP3V3_G3H is now 0v.. it was 3.3 before.. hmm Maybe when I reflowed U7000 I messed up the pins underneath, causing the enable for 3v3 to not work.
                                    Last edited by powerblackout; 01-24-2022, 05:34 PM.

                                    Comment


                                      #19
                                      Re: 820-00840-A CD3215 short

                                      If you have reflowed U7000 then consider to apply flux onto the top side of the device -> high heat -> medium amount of air and allow for the balls underneath to melt naturally. Do NOT press down on the top side of the IC as this will squish the solder balls and make a mess.

                                      Keep moving in circles on the top of the device till you can see the balls sink down - naturally - test the condition using tweezers across 2 edges (corner).

                                      With a very gentle nudge, the IC should 'dance' and then spring right back to their proper position on the pads. Repeat this once or twice to confirm the balls are being corrected with this process and they will bounce back on their own.

                                      If you push too hard with the tweezers, then U7000 will end up in your neighbor's backyard. Just take your time here.

                                      The above should remove any solder faults underneath U7000.

                                      After this, allow the board to cool. Do not apply power.

                                      Check the resistance to ground of the SMBUS SDA & SCL lines.

                                      Post the readings.

                                      Comment


                                        #20
                                        Re: 820-00840-A CD3215 short

                                        Ok reflowed 2 more times and I have PP3V3_G3H again.

                                        Now that it is back, the SDA and SCL lines are at 3.2V and I see them jumping around about once a second (meaning traffic?)

                                        Comment

                                        Working...
                                        X