Announcement

Collapse
No announcement yet.

[Solved] Dead U2 Programmer?

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

  • #16
    It has been sent.

    Best Regards,
    Paul

    Comment


    • #17
      I can't reproduce the problem with your file, but I did succeed in an "Erase Failed" message when the chip was previously code-protected.

      Can you confirm that the configuration settings on the programmed chip match what is in your hex file? Are you making manual changes to configuration that are not stored in the file?

      My theory is that this is related to specific configuration data. We're studying up on the code-protect circumstances, but I have no evidence that this is what is happening on your end.
      Charles Leo
      ME Labs, Inc.
      http://melabs.com

      Comment


      • #18
        I've just tested the bulk-erase at 3.3V and it works fine for me. The low-voltage erase should not be required. The programming specification supports this finding, saying that the bulk-erase should work down to 2.7V. For subsequent testing, please uncheck the low-voltage erase option.

        I'm told that some adjustments were made for the 87K90 in our last beta release. If you would, download and try the beta version of the meProg software:
        http://melabs.com/support/progsoft.htm
        Charles Leo
        ME Labs, Inc.
        http://melabs.com

        Comment


        • #19
          Originally posted by Charles Leo View Post
          I can't reproduce the problem with your file, but I did succeed in an "Erase Failed" message when the chip was previously code-protected.
          I have never code protected anything in a PIC. Ever. Not even "just to try it out."

          Can you confirm that the configuration settings on the programmed chip match what is in your hex file? Are you making manual changes to configuration that are not stored in the file?
          Yes, I've learned the hard way that it's best to make any config changes directly in the PBP file. PBP3, of course, makes it far more convenient to do that...

          My theory is that this is related to specific configuration data. We're studying up on the code-protect circumstances, but I have no evidence that this is what is happening on your end.
          If I took the same PBP3 hex file and loaded it into the PIC with the non-U2 programmer, then read it back and saved it, would that be of any use to this? Perhaps there are some defaults that it sticks in for bits I haven't explicitly set.

          Best Regards,
          Paul

          Comment


          • #20
            We just noticed an anomaly concerning the watchdog timer postscaler configuration. Not sure if it's related at all, but its behavior is quite mysterious.

            I'll get back to you on this...
            Charles Leo
            ME Labs, Inc.
            http://melabs.com

            Comment


            • #21
              Originally posted by Charles Leo View Post
              I've just tested the bulk-erase at 3.3V and it works fine for me. The low-voltage erase should not be required. The programming specification supports this finding, saying that the bulk-erase should work down to 2.7V. For subsequent testing, please uncheck the low-voltage erase option.
              Since the description of the problem behavior I've used only the default settings unless otherwise specified.

              I'm told that some adjustments were made for the 87K90 in our last beta release. If you would, download and try the beta version of the meProg software:
              http://melabs.com/support/progsoft.htm
              That is the version I am using. Re-downloaded and re-installed yesterday,just in case. About box says "4.50 beta". Incidentally, the download file is "meProg_4_50_1_Beta_setup.exe-". The final "-" has to be removed from the filename before the installer will run.

              Best Regards,
              Paul

              Comment


              • #22
                I've sent a new EPC file for the 87K90 with instructions via email.

                We found that the WDT postscaler setting in the configuration wasn't being implemented correctly. This could cause verification errors, though it doesn't explain why I couldn't repeat the problem.

                To fully duplicate your environment, I will need a hex file that represents the chip before you try to program it. This is created by reading the chip and saving a new hex file. You should have the new EPC installed when doing this.
                Charles Leo
                ME Labs, Inc.
                http://melabs.com

                Comment


                • #23
                  Regrettably I can see no changes to performance.

                  Not 100% sure what you need for the hex file, so I erased and blank-checked using the non-U2 programmer, then used the U2 and mePROG Beta to download and save the file. It has been sent by direct email.

                  Best Regards,
                  Paul

                  Comment


                  • #24
                    After erasing with the Microchip programmer, does the U2 still give you the "Erase Failed" error when you try a simple erase?

                    My goal is duplicate exactly the state of the chip and the process you are using.
                    Charles Leo
                    ME Labs, Inc.
                    http://melabs.com

                    Comment


                    • #25
                      Originally posted by Charles Leo View Post
                      After erasing with the Microchip programmer, does the U2 still give you the "Erase Failed" error when you try a simple erase?
                      Yes. I went through all of the same processes discussed earlier in this thread with no changes from what has already been reported.

                      My goal is duplicate exactly the state of the chip and the process you are using.
                      Want me to ship you the board?

                      Best Regards,
                      Paul

                      Comment


                      • #26
                        I've tried everything I could think of and I can't repeat the failure. Shipping the board may be our only option.

                        If the programming cable is custom - other than melabs 10-pin or Microchip 6-pin - I will need a cable.

                        If you can do without it for a few days, send your U2 programmer. This is optional. If I can't make your board fail with my programmer, I'll send you another U2 for testing.

                        Ship to:

                        > Charles Leo
                        > microEngineering Labs, Inc.
                        > 2845 Ore Mill Road STE 4
                        > Colorado Springs, CO 80904
                        > 719-520-5323

                        I'll test as soon as I receive it.
                        Charles Leo
                        ME Labs, Inc.
                        http://melabs.com

                        Comment


                        • #27
                          It will be in your hands tomorrow. I will send the Eagle files (schematic and layout) via email. I'll also include the U2 I've been using just in case, although I have already tried it using a second U2 (with identical results). Let me know if there is anything else I can provide.

                          Sorry to be such a pain...

                          Best Regards,
                          Paul

                          Comment


                          • #28
                            No worries, Paul.

                            I'm afraid I will be out of the office tomorrow, though. You might save the overnight charges and ship 2nd-day instead. I'll be able to look at it on Monday.
                            Charles Leo
                            ME Labs, Inc.
                            http://melabs.com

                            Comment


                            • #29
                              Thanks for letting me know, it will arrive Monday then.

                              Best Regards,
                              Paul

                              Comment


                              • #30
                                Problem Absolutely Solved

                                Hello Everyone;

                                If you've made it to this point in the thread, the problem has been found and it was 100% my own fault. Not sure why but I've always put a 0.1µF cap to ground on the reset line between the programmer and the PIC. At Charles' suggestion I removed that capacitor. Lo and behold it all works perfectly now...

                                Mea culpa.

                                The U2 programmer remains a slice of perfection. I am profoundly grateful for time Charles has invested in remedying what was clearly my own ignorant mistake.

                                If possible, I suggest modifying the title of this thread to add "SOLVED" at the end. For completeness I've attached a corrected schematic.

                                "The only way to avoid mistakes is to gain experience. The only way to gain experience is to make mistakes."

                                Best Regards,
                                Paul

                                Working_ICSP.pdf

                                Comment

                                Working...
                                X