Does anyone know of any issues that might be preventing some PIC16LF819's being reprogammed in-circuit, using a Microchip ICD2 programmer? I have 20 boards each with a PIC16LF819 (SOIC) soldered in. They all programmed without any problems first time around. I then needed to reprogram them and 17 reprogrammed fine, but 3 failed. In fact they fail at the first of the programming cycle, the Part Erase. Even though the erase reports that it succeeded, it takes a lot longer than it should (5 seconds instead of few hundred ms) and a subsequent blank check says it failed. Everything I have tried so far has not got them to reprogram. The config bit settings as shown by MPLAB (v6.60) for the original and reprogramming are: Oscillator: INTRC-RA6 is Port I/O Wathdog Timer: Off Power Up Timer: On MCLR Select Bit: RA5 is MCLR Brown Out Detect: Off Low Voltage Program: Disabled Data EE Read Protect: Disabled Flash Program Write: Write Protection Disabled CCP1 Mux: RB2 Code Protect: Enabled - All protected I have desoldered all the peripherals components in case they were causing the trouble. So now the only connections to the Vpp, PGC and PGD pins are from the board's programming header, but the reprogramming still fails. I've buzzed through and checked for shorts but no problems found. I have tried selecting power from the ICD2 and also power from the board - no luck erasing/reprogramming either way. Checking the signals on a scope: Vdd is 5V (if powered from the ICD2) or 3.3V (if powered from the board); During programming Vpp reaches 13.5V (rise time of 70ns approx); PGC and PGD both produce good squarewaves between 0V and Vdd. I've tried a couple of other ideas I've come across - adding a pull-down of 1k to the Vpp line, and also a pull-up of 1k to the same line (on a different occassion) but with no success. Could anyone please offer any suggestions as to how to get these PIC16LF819's to erase and reprogram? BTW - the codespace is totally full, otherwise I guess I could include a bootloader to get around this. Also, if another programmer is likely to work better for this than the ICD2 then I'm willing to get one. Very many thanks, James _______________________________________________ http://www.piclist.com View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist