PIC ROM erase failure

[shahidali55]

I am unable to clear the 8th bit of data word in entire flash data memory . . .

Softwares tried:-
WinPIC800
ICProg 106A

Data given to program :-
0000 0000 0000 0000 0000 0000 0000 0000

Data read after programming:-
0100 0100 0100 0100 0100 0100 0100 0100

This is 16F628A device . . .
Another 16F628A works fine.

[skimask]

Softwares tried:-
WinPIC800
ICProg 106A
..................................................
This is 16F628A device . . .
Another 16F628A works fine.

Option A) Throw out 16F628A#1 and keep 16F628A#2

Option B) PICKIT2

[shahidali55]

Good point skimask.
Looking the data aquired by oscilloscope, there was a 24v, 15ns pulse on VCC.
That must have fried the chip.

Surely PICKIT2 can't repair damage done to the Flash memory cell.

[mister_e]

Probably not damaged at all, most a timing problem that most/All JDM based software have one day or another. Nothing you can fix yourself. One of your PIC is probably less timing critical than the other.. lucky guy!

[shahidali55]

I've been programming PICs for over 3 years,
I have programmed 12C508A, 12F629, 16F628A, 16F84A, 16F676, 16F873A, 16F877A,
18F2550 and 18F4550 with my JDM.
The problem with this particular PIC is that it was working fine for over 500 rewrite cycles but a simple switchmode step up converter malfunction caused it to fry (not completely).
I still cant figure out why only one bit of all the memory words is fixed to '1'.
Everything erases and writes fine , even the EEPROM.
Is there some type of of "full erase" for PICs ? If yes, how can it be done?
This "fixing" of the bits was caused by a 24v spike so i dont think it can be erased with the normal 13.5v VPP.

I'll just replace it with a new one . . .

Just curious about this behavior of a PIC :-)