-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 source= http://www.piclist.com/piclist/2004/01/12/162125a.txt? Hmm. I just noticed the "Quote" button on the PICLIST web page. It gives this: Wouter van Ooijen says: >For example, instead of filling an 8K buffer then ALWAYS erasing and >reprogramming the whole memory, I only deal with the memory that is >represented in the .hex file. >so how do you deal with a code-read-protected chip? and which 8k? there >are PIcs in all sizes, ranging from 1k (0.5k if you include EPROM PICs) to >IIRC 32k. I haven't. I have to re-read the data sheet, but I suspect that I can't do anything without +13V present. In part, I have the posture that anyone using code protection should use a production programmer. If I can clear the code protect word with only +5, then I just have the erase procedure to do a check on the code word. This has philosophical implications on the remainder of the protected page that is erased, and I don't usually advocate philosphically uncomfortable positions. The problem is that clearing the code word erases the protected page, which in turn changes memory that is not addressed by the (sparse) .hex file. This is a silent, non-obvious side effect which will cause lots of mischief. This leaves me to say that a hex file that addresses a protected memory area should fail on my programmer. I think I'm completely comfortable with that. I might be willing to have a "catastrophic erase" command that is difficult to activate, but again I might express that difficulty by having it be a different product. Remember, I'm using parts in front of many students. I'll be using up the erase cycle budget and running out of money if I let erase be cheap. I'll also teach them to "be busy" and not "to plan". > I don't do one thing that otherwise seems universal: I only use low > voltage programming, and I don't make a 13V supply. I have not addressed > or deliberately ignored the variable programming voltage issues. >That makes it a somewhat less general programmer. I agree. It is not intended as a general programmer. You know how difficult it is to verify and maintain such a product. I can't compete with that. My niche is the philosophical position that a group, several specimens of 1 design, should interact peer-to-peer to initialize or restore functionality. So I have a 16F877 programmable product with a 16F877 programmer built in. If I added more functionality, I would add error correction bits here and there in the memory, and a collection of damaged parts would have some potential to recover correct functionality for all, in the field. To me that would be more valuable than an ability to program a part that my shop never sees. >My Wisp628 fits at the high end of the hobby designs / low end of the >seriosu designs. But I sell it, except for one-off hobby purposes it is >not a free design. IIRC Kanda has a similar programmer. Your Wisp628 seems to have favorable reviews. It is one of the ones I had thought about purchasing, but I am being slow to spend money. > If possible, I will use an existing programmer design with my > board. Actually, this seems to narrow down to using a standard connector > and pin-out. >There are a fair number of ICSP connectors, but I would go for an RJ >connector compatible with Microchip's ICD2. The boards that I make for for >instance educational purposes have both this RJ connector and a connctor >form my Wisp628. In my peer-peer concept, this means that I have to provide the RJ plug on the other board, or else provide two RJ sockets and a cable. Which choice does the ICD2 make? Is the ICD2 schematic available? - --- Aubrey McIntosh http://www.piclist.com/member/AM-vima-Y84 PIC/PICList FAQ: http://www.piclist.com -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 7.0.3 for non-commercial use iQA/AwUBQANE4gKlSw8yssF7EQI2vACgmy2uk75hmTRQeLmUdv2Y9GiP320An3Bn HCZDgpkyo6LotPGpvGmg48Z7 =PdG/ -----END PGP SIGNATURE----- -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads