6 Replies Latest reply on Mar 10, 2018 6:11 AM by gfoersler_2625781

    FX2LP windows 10 driver problem

    bdawasthi_2660231

      I am new to FX2, not able to install drivers for this board on windows 10, unrecognized USB showing VID=0000 and PID=0002. 

        • 1. Re: FX2LP windows 10 driver problem
          hman

          The device should enumerate with VID 0x04B4 and one of the supported PIDs for the device to bind to Cypress driver. So please check the descriptor file in your firmware.

          Once the device reports the right VID and PID you can bind the device to the Win 10 driver present in the following link:

          http://www.cypress.com/design-guides/cy4500-ez-pdtm-protocol-analyzer-guide

          Install the " CY4500 EZ-PD Protocol Analyzer EXE" present in the above link and after installation navigate to the following folder:

          C:\Program Files (x86)\Cypress\CY4500 EZ-PD Protocol Analyzer\1.0\drivers\DriverBinary\Vendor_Driver\bin\Win10

           

          Please let us know for any troubles.

          • 2. Re: FX2LP windows 10 driver problem
            bdawasthi_2660231

            Hi hman,

            thanks for reply, but i am still not able to fix problem.

             

            fist of all, i am using a third party board for evaluation. image attached

            IMG_20170825_234706512-2.jpg

             

            i have removed EEPROM to ensure that EZ-USB starts in default mode with VID and PID hard coded on chip.

             

            when i connect this board, it appears as unrecognized USB with VID = 0x0000 and PID = 0x0002.

             

            i cant do anything unless this board get recognized, it is not getting detected by any of cypress applications.

             

            how to fix VID and PID so that i could install correct drivers.

             

            dev_man.png

            dev_man_usb.png

            • 3. Re: FX2LP windows 10 driver problem
              hman

              Hi,

               

              - Can you mention the complete part number used in this board?

              -Is this behavior seen in all the PCs?

              If yes, then we can check the following on the third party board:

              - Verify if there are pull-ups on the I2C lines even though there is no EEPROM. Pull-ups are compulsory even if there is no EEPROM.

              - As soon as you plug in FX2LP check whether there is slave clock generated on the SCL line by I2C master in FX2LP? If there is, check whether it is querying the EEPROM addresses 0xA2 and 0xA0 over SDA lines?

              - Check whether the Power on reset timing is followed as per the datasheet spec(mentioned in page 8 of FX2LP datasheet: section 2.9 Reset and Wakeup) : If crystal is used as clock input to FX2LP, then the R and C on the reset pin should be chosen in such a way that the reset is asserted for atleast 5mS)

              - Can you take a USB trace(using LeCroy Analyzer) during enumeration and share it?

               

              Please verify the above checklist and see if there is anything wrong.

               

              Regards,

              hemanth

              • 4. Re: FX2LP windows 10 driver problem
                wuzeg_2738481

                I had exactly the same problem with these boards (bought on eBay).

                The chip gets very old and the enumeration fails.

                 

                My conclusion was bad FX2LP chips (or china fakes)?

                 

                Regards,

                M

                • 5. Re: FX2LP windows 10 driver problem
                  gfoersler_2625781

                  Hi,

                   

                  IMHO the Jumper J2 on these boards (at least the revision I have) just kind of switches address pin 1 between 1 and 0, what might for one position make the large EEPROM being addressed and for the other one make the large EEPROM bahave kind of like a small EEPROM (not investigated thoroughly enough, just my first thoughts).

                   

                  As you have removed the EEPROM that should not be an issue for you anymore (i have disconnected SDA from the EEPROM leaving the pull-up resistor in place, connected to the FX2LP).

                   

                  BTW the pull-ups are 10 kOhm on these boards...

                   

                  With the board that I have, the PC fings the unprogrammed chip, but after programming the RAM the Cypress application examples don't find it anymore. The PC, though, does find the device sometimes. Sometimes saying "working", sometimes "error 10", sometimes just as unknown device even without any possibility to manually install a driver.

                   

                  Only if I use the Bulkloop example from the Cypress Suite USB 3.4.7 (not from the DVK), I can program (RAM) and run the Bulkloop example, but the Streamer Example does not work there either. If I try with the Streamer application the chip may go completely unrecogniceable (eventually it comes up again as NO EEPROM Device after some time off and some juggling with manual driver installation).

                   

                  I also wonder about possible fake chips.

                  Without EEPROM my device enumerates as "Cypress FX2LP No EEPROM Device" "USB\VID_04B4&PID_8613&REV_:001". CyConsole finds a DeviceID of 0xA001. It would be nice to have a table to correlate chip revision of the silicon to the marking on top of the chip, which is on mine:

                   

                  CY7C68013A-56PVXC

                  B  04 PHI 0701

                  CYP 604789

                   

                  Have you tried using the manual reset button? As hemanth suggested, the reset R and C components for the reset may be wrong on these boards, but using the reset button one should be able to ensure at least 5ms reset duration.

                   

                  If you find a fix for the board I would be very interested.

                  • 6. Re: FX2LP windows 10 driver problem
                    gfoersler_2625781

                    Hi,

                     

                    it seems the bulkloop example from the Cypress Suite USB 3.4.7 is the only one that does not try to write to the 7 segment display of the DVK evaluation board. All the others do try that and wait for write success, which never is acheved on boards not having the 7 segment display.

                     

                    If you happen to have a board different from the original cypress evaluation board, you have to remove this writing to the 7 segment display (including the wait for write success) and rebuild the firmware in order to have a running firmware.

                     

                    I modified the streamer firmware from the Cypress Suite USB 3.4.7 accordingly and now the streamer app finds the programmed device and runs with it.