Post by thread:[PIC]12f675 osccal

Hi, The original calibration value of a 12f675 is seems to be erased. The original value it was read it with MPLAB and PS+, and it was "retlw 0x48" at 3fff. I wonder if it is possible to re-write this value. The pic was also missconected "once" on the protoboard: Vdd to pin 2 and gnd to pin 7 (shifted one hole), It was reprogrammed and that pins work nice... The main problem now is when the ps+ finish to verify, there are a lot of bad memory locations compared with the program. If it done twice, the pic is succesfully programmed... and this never happened before, as long as i recall. No icsp is used. Any comments about trust on "that" pic or about how to reprogram the osccal is apreciated. Dennis Crawley Argentina ps: no generalisimos arround,... but is wost.

Just set an org at 03fff, and install the instruction. Try it in MPLAB, compile it, then look at the PGM array. It will be right there. --Bob Dennis Crawley wrote: {Quote hidden}>Hi, >The original calibration value of a 12f675 is seems to be erased. >The original value it was read it with MPLAB and PS+, and it was "retlw >0x48" at 3fff. >I wonder if it is possible to re-write this value. > >The pic was also missconected "once" on the protoboard: Vdd to pin 2 and gnd >to pin 7 (shifted one hole), It was reprogrammed and that pins work nice... > >The main problem now is when the ps+ finish to verify, there are a lot of >bad memory locations compared with the program. If it done twice, the pic is >succesfully programmed... and this never happened before, as long as i >recall. > >No icsp is used. > >Any comments about trust on "that" pic or about how to reprogram the osccal >is >apreciated. > >Dennis Crawley >Argentina > >ps: no generalisimos arround,... but is wost. > > > > > > -- Note: To protect our network, attachments must be sent to spam_OUTattachTakeThisOuTengineer.cotse.net . 1-520-850-1673 USA/Canada http://beam.to/azengineer

Bob Axtell wrote : > > Just set an org at 03fff, and install the instruction. > Try it in MPLAB, compile it, then look at the PGM > array. It will be right there. > > --Bob Maybe check how the programmer handles this. I guess that any programmer that knows what it's doing, would first read the value from h'3FFF' and then add that to the code from the HEX file (so it will be written back). Check that the programmer realy uses the value from HEX file (if any) instead of the read value (that we know is wrong) from the chip itself. Also remember to comment out the ORG and RETLW from the source, or you'll end up with wrong cal values if you use another 12F675 then the one you're using right now... Jan-Erik.

Dennis Crawley wrote: > The original calibration value of a 12f675 is seems to be erased. > ... > The pic was also missconected "once" on the protoboard: Vdd to pin 2 > and gnd to pin 7 (shifted one hole), > ... > The main problem now is when the ps+ finish to verify, there are a lot > of bad memory locations compared with the program. > ... > Any comments about trust on "that" pic How many $10 have you already spent wondering whether it makes sense tossing out a suspect $1 part? How many $100 will it cost if you don't and you were wrong? *Think* about it. ****************************************************************** Embed Inc, Littleton Massachusetts, (978) 742-9014. #1 PIC consultant in 2004 program year. http://www.embedinc.com/products

>Any comments about trust on "that" pic or about how to >reprogram the osccal is apreciated. Well others have said how to reprogram the value, but no-one mentioned anything about what happened to the rest of the chip. What you say about the now unreliable programming suggests that the 13V that is applied to go into programming mode appeared on the incorrect chip pin when it was mis-socketed, and this has damaged the chip, to a a point where I would suggest its reliability is now quite suspect. It will have had ad least 7V between Vcc and Gnd, and possibly 12-13V depending on just which direction the pinout was wrong when in the socket. I would recommend tossing that one (use it in a kids toy if you really mustn't toss it) and getting another for serious use.

Bob Axtell <.....engineerKILLspam@spam@cotse.net> wrote: > Just set an org at 03fff, and install the instruction. > Try it in MPLAB, compile it, then look at the PGM > array. It will be right there. > > --Bob Well, the pic may be locked. I've made org 0x3fff retlw 0x48 but when the pic is read it, it return a nop at that address. I don't know if the programmer settings is related to, because the range is set at start addres 0, end address 3fe, and it can't be modified to 3ff. In the same window, the calibration memory check-box is on. I asume the pic some how fail to write at 3fff. I will try to program it, but with a 12F683 device stting, with the same program, and see what this does. Dennis

Dennis Crawley wrote : > I don't know if the programmer settings is related to, > because the range is set at start addres 0, end address 3fe, > and it can't be modified to 3ff. So the programmer expects that the PIC already has a valid cal-value in 3FF, and does not accept another value from the HEX file. Seems as a logical deafult setting... > In the same window, the calibration memory check-box is on. And what if you de-check it ? > I asume the pic some how fail to write at 3fff. I will try to > program it, but with a 12F683 device stting, with the same > program, and see what this does. But it must be possibole to tell the programer that the cal- value is "lost" and that it should program a new value from the HEX file. Or can it be specified as a command line option ? Jan-Erik.

On 1/18/06, Dennis Crawley <dennis.crawleyKILLspamusa.net> wrote: {Quote hidden}> Hi, > The original calibration value of a 12f675 is seems to be erased. > The original value it was read it with MPLAB and PS+, and it was "retlw > 0x48" at 3fff. > I wonder if it is possible to re-write this value. > > The pic was also missconected "once" on the protoboard: Vdd to pin 2 and gnd > to pin 7 (shifted one hole), It was reprogrammed and that pins work nice... > > The main problem now is when the ps+ finish to verify, there are a lot of > bad memory locations compared with the program. If it done twice, the pic is > succesfully programmed... and this never happened before, as long as i > recall. > > No icsp is used. > > Any comments about trust on "that" pic or about how to reprogram the osccal > is > apreciated. > The PICkit has a button (in the host software) to recalibrate your 12F675. Regards, Mark markrages@gmail -- You think that it is a secret, but it never has been one. - fortune cookie

Jan-Erik Soderholm <.....jan-erik.soderholmKILLspam.....telia.com> wrote: > Dennis Crawley wrote : > But it must be possibole to tell the programer that the cal- > value is "lost" and that it should program a new value from > the HEX file. Or can it be specified as a command line option ? > Jan-Erik. Jan-Erik, I did it, cheating the MPLAB-ps+ with a 12f683 type, with the code that Bob told me to do it. Now the value is in the right place. Analyzing what I did, I forgot to change the pic type the first time I programmed it this morning. I think this is the explanation. Since I won't use this particular pic any more, elucidated by "my own" common sense, confirmed by my soul guru Olin :), I'll repeat what I did this morning, expecting a nop at 3fff, So to be sure. Thank a lot for the advices and helps, to you, extensive to Olin, Bob and Alan. Dennis Crawley Argentina ps: do what I did is not a sin, isn't it?

I seem to remember the procedure to recalibrate osccal was mentioned in the pdf's for the PICkit1,available in the developement tools section of micro's homepage. the early versions of pickit1 had a nasty habit of erasing (i.e writing over) the osccal location.Hope this helps ; )

I would think that you can set a clock out and measure the frequency with a scope and adjust oscal accordingly. On 1/18/06, Steph Smith <EraseMEzapsmithspam_OUTTakeThisOuTblueyonder.co.uk> wrote: > I seem to remember the procedure to recalibrate osccal > was mentioned in the pdf's for the PICkit1,available in > the developement tools section of micro's homepage. > the early versions of pickit1 had a nasty habit of erasing > (i.e writing over) the osccal location.Hope this helps ; )

> -----Original Message----- > From: piclist-bouncesspam_OUTmit.edu > [@spam@piclist-bouncesKILLspammit.edu] On Behalf Of Mark Rages > > The PICkit has a button (in the host software) to recalibrate > your 12F675. > Just want to add both PICkit 1 and PICkit 2 have this function. Another thing is that 0x3400 is the rough value based on my experience with 12F629 and 12F675. Regards, Xiaofan