Post by thread:[PIC]: 18F Programing problems.

I'm having trouble getting my programmer to actually write to a 18F device. It can read the ID bytes correctly, It can read the config bytes correctly (they are at the default "erased" values), It can read the code memory from an already programmed device. What value should it read from the code memory of a brand new device ? I'm getting all zeros and I expected all ones. Also it always reads zeros from the code memory even after running "bulk erase" and programming. I've even resorted to looking at the clock and data signals on a 'scope and all looks good. Are there any "gotchas" in the programming guide ? Peter Onion.

What programmer/software? On Friday 22 July 2005 04:58 am, Peter Onion scribbled: {Quote hidden}> I'm having trouble getting my programmer to actually write to a 18F > device. > > It can read the ID bytes correctly, > It can read the config bytes correctly (they are at the default "erased" > values), > It can read the code memory from an already programmed device. > > What value should it read from the code memory of a brand new device ? > I'm getting all zeros and I expected all ones. > > Also it always reads zeros from the code memory even after running "bulk > erase" and programming. > > I've even resorted to looking at the clock and data signals on a 'scope > and all looks good. > > Are there any "gotchas" in the programming guide ? > > Peter Onion.

On Fri, 2005-07-22 at 09:17 -0500, PicDude wrote: > What programmer/software? > "I'm having trouble getting **my** programmer ........" I've been looking at the sources for several other programmers and I can't *see* anything different between what mine does and what the others do (same PIC commands in the same sequence). I'll have to get the scope out again and check more closely the timings on the clock and data signals and the Vdd and Vpp signal voltages. Peter

Peter Onion wrote : > > "I'm having trouble getting **my** programmer ........" > Well, I don't know, but maybe there *is* a reason that most just uses a programmer designed and tested by someone else. Best Regards, Jan-Erik.

On Fri, 2005-07-22 at 15:36 +0100, Peter Onion wrote: > On Fri, 2005-07-22 at 09:17 -0500, PicDude wrote: > > What programmer/software? > > > > "I'm having trouble getting **my** programmer ........" Well after much head scratching and pulling out of hair it is working (for small values of working) in that it has written a count to the first 8 bytes of the code memory. I'm not sure what the problem was in the end, infact I think there were several ! Mind you, the fact that my "waitWms" function failed with W=1 didn't help ! Also looking at the Wouters code in wisp628.jal seems to suggest that some of the steps in the programming specification aren't needed (some of the twiddling of bits in EECON1 before configuring for single pannel writes for example). I've got code running in the PC end to load .hex files so it (should) be an easy step now to actually program a 18F458 to "blink a led". Anyway, I need to go and do something else now before my eyes turn square from looking at these LCDs for too long ! Peter :-)

> Also looking at the Wouters code in > wisp628.jal seems to > suggest that some of the steps in the programming specification aren't > needed (some of the twiddling of bits in EECON1 before configuring for > single pannel writes for example). I you are brave you can take the programming specs at the same level as the self-programming examples in the datasheet: probably correct (for small values of probable), but not necesarry the way you would do it. And you of course realised that you are just force-feeding instructions to the CPU. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

> What value should it read from the code memory of a brand new device ? > I'm getting all zeros and I expected all ones. I recall having a similar problem when I first started programming 18F's but I can't remember exactly, nor what the cause or solution was. But I strongly recommend that you get a known working programmer, program the PIC with it, and then try to read it with your own progger-in-development. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

On Tue, 2005-07-26 at 10:40 +0200, Wouter van Ooijen wrote: > I you are brave you can take the programming specs at the same level as > the self-programming examples in the datasheet: probably correct (for > small values of probable), but not necesarry the way you would do it. > And you of course realised that you are just force-feeding instructions > to the CPU. I've not looked at "self programming" techniques at all ... yet. But boot loaders are getting close to the top of my "interseting things to investigate" list. My programmer is now working nicely for 18F devices. I can now start to investigate using sdcc. Peter

On Tue, 2005-07-26 at 10:40 +0200, Wouter van Ooijen wrote: > I recall having a similar problem when I first started programming 18F's > but I can't remember exactly, nor what the cause or solution was. But I > strongly recommend that you get a known working programmer, program the > PIC with it, and then try to read it with your own > progger-in-development. Luckily I have a programmed 18F452 in one of these... http://www.crownhill.co.uk/product.php? prod=1306&PHPSESSID=2b7a27e382006f133bc20b6f274ee9df So I could check the "read back" was working in my programmer. Peter