>Date: Tue, 05 Apr 2005 16:43:19 +1000
>To: spam_OUTsupportTakeThisOuTmelabs.com, .....ccsKILLspam@spam@ccsinfo.com
>From: Jim Robertson <bugs_839KILLspamnewfoundelectronics.com>
>Subject: Config word address error in your products.
>
>
>Attn: Microengineering Labs and CCS Inc.
>
>FYI
>
>From:
>Jim Robertson
>NEWFOUND ELECTRONICS
>
>Both of you are using an incorrect address for the config word of 12-bit
>core PICmicros.
>
>The affected products that I am aware of are the CCS 'C' compiler and the
>Melabs programmer software. Other software and hardware produced by you
>may also be affected and should be investigated by yourselves.
>
>The problem I refer to is that the config word address specified in the
>HEX read and produced by your products is not per microchip specification
>and in fact corrupts the standard. Each of you is failing to double the
>config word address (0xFFF * 2 = 0x1FFE)
>
>This line was taken directly from a HEX file produced by the CCS Compiler
>and it represents the config word as defined in the source code.
>
>:020FFE00C40F1E  (This is incorrect!)
>
>According to the Microchip standard this line should be:
>
>:021FFE00C40F0E
>
>(Naturally the checksum is also corrected.)
>
>Clearly, this is an error on the part of the 'C' compiler.
>
>The Melabs software however reads the hex file as if it was correct and
>also saves the hex file with the same error!
>
>In fact, what you are both doing is putting the Config word in the RESET
>VECTOR (0x7FF) of a 4K 12-bit core part (eg 16F57) You can see this if,
>using MPLAB and/or the Melabs software, WARP-13 etc..),  you select a
>16F57 and load (or "Import") a file compiled (or saved in the case of
>Melabs) for a 16F505 (or any 12-bit core part?). The config word will
>appear as the RESET VECTOR @ 0x7FF.  The RESET VECTOR is being overwritten!
>
>The attached GIF shows this.
>
>When the memory size of the selected PIC is < 4k, my WARP-13 software
>correctly and wisely flags this condition as an "address overrun error" It
>is!  For 4k pics It will not flag any error as, technically, the hex file
>will be correct.
>
>I notice that other programmer software I tested with (MPLAB, DIY*) did
>not produce any warnings when loading the incorrect hex file. The
>consequence of this is that people are led to believe that the bug is in
>the WARP-13 software and, really, I'm getting a little tired of this...
>
>Please attend to this error in your wares ASAP. Thank-you.
>
>Regards,
>
>Jim Robertson
>NEWFOUND ELECTRONICS
>
>
>*Testing done with a 16C505 as the DIY range of programmers does not
>appear to support even one of the 12-bit core flash parts?!
>
>
>
>_______________________________________
>Home of the famous WARP-13 PIC programmer.
>NEWFOUND ELECTRONICS P/L
>http://www.newfoundelectronics.com
>_______________________________________