HP Forum Archive 09

[ Return to Index | Top of Index ]

15c CPU Message #1 Posted by Sean McNamee on 12 Dec 2002, 2:07 a.m. On the MoHPC's 15C page, there is a link to "Technology and Packing", which then takes you to a 1x series page. This page in turn contains a link to CPU info -> Does this page contain the information for the 15C?? It looks like it does, but it doesn't actually mention the 1x series at all on this page. The reason I'm asking is I'd like to hack the 15c a bit. Has anybody done this?

Re: 15c CPU Message #2 Posted by Ex-PPC member on 12 Dec 2002, 6:09 a.m., in response to message #1 by Sean McNamee http://www.finseth.com/~fin/hpdata/hp15c.html

Re: 15c CPU Message #3 Posted by Sean McNamee on 12 Dec 2002, 7:01 a.m., in response to message #2 by Ex-PPC member Ok... So what's a "1LE2 SACAJAWEA" ? Google doesn't come up with anything - except a newsgroups reference to it being related to a BERT??? What's a BERT???

Re: 15c CPU Message #4 Posted by Raymond Hellstern on 12 Dec 2002, 7:51 a.m., in response to message #3 by Sean McNamee 1LE2 is the ID of the chip, SACAJAWEA is it's name, and BERT is an HP internal name for some pioneer series calculators (10B,20S,21S), so BERT may not be what you need. Compare http://www.hpmuseum.org/collect.htm#series . Another source for HP calculator internals is www.hpcalc.org , mainly for the Lewis/Clarke/Yorke chips (HP-48 & 49G), but maybe you should first read http://www.finseth.com/~fin/hpdata/names.html to get an idea how it all belongs together. Raymond

Re: 15c CPU Message #5 Posted by Bill Wiese on 13 Dec 2002, 1:40 p.m., in response to message #1 by Sean McNamee Sean... You're gonna have a hard time hacking your 15C at the lowest firmware level. In theory it is possible. But in practice you're gonna have to rip things apart and use quite a bit of logic (maybe PLD/FPGA) to interface some other (P)ROM. SP (synthetic programming), commonly used in 41C series, essentially is the merging of user-level commands/operands not normally possible due to error-checking, etc. for 'safety reasons'. Thru a bit of trickery, access to other user-undocumented registers, modes, etc. is possible; this does not include lowest-level CPU programming at assembler level. Most HP calculator external RAM/ROM is 'weird' - IIRC, serialized streams, possibly a combined RAM&ROM chip, and the instructions are 10bits wide. It took quite a bit of hardware, etc to do CPU-level code development on the HP41C - and that was a machine with ports! Bill Wiese San Jose, CA

Re: 15c CPU Message #6 Posted by Sean McNamee on 13 Dec 2002, 4:48 p.m., in response to message #5 by Bill Wiese Well, I was only interested in seeing what was possible at the software level. I just acquired my 15c, and I'm not really interested in tearing it apart! The question in my post about SP was because, due to the MoHPC CPU page, I wasn't sure if these two calculators shared the same CPU. I have found it's possible to browse the 15c's memory using the tricks found on http://www.jlw.com/~woolsey/15Cnotes.html but I was interested in getting further info (ie. whether ROM space is available through this trick, instruction format, etc...). The page does seem to indicate that it's so, but doesn't give the details. This page is the ONLY location I've seen any such details on the 15c, so I thought this would be the next logical place to try... Thanks for your response.

PS Message #7 Posted by Sean McNamee on 13 Dec 2002, 4:56 p.m., in response to message #5 by Bill Wiese Even if ROM isn't available, it is evident that the programmable memory is shared with the registers, so that "self-modifying" code could be entered using this trick, if the instruction format was known.

Re: PS Message #8 Posted by Vieira, Luiz C. (Brazil) on 13 Dec 2002, 5:52 p.m., in response to message #7 by Sean McNamee Hi; I've been using an HP15C since 1984/1985 (I'm not sure about the precise date) and I was known about this sort of information only last year, when I foud the MoHP. I decided to find as much info as I can, and I'm dealing with HW. As I find something that's interesting, I'm posting. For now, I'm delving into some programs for the HP41 series I intend to place at the Library for downloading. Cheers.

Re: PS Message #9 Posted by Bill Wiese on 13 Dec 2002, 8:26 p.m., in response to message #7 by Sean McNamee Sean... > Even if ROM isn't available, it is evident that the > programmable memory is shared with the registers, so > that "self-modifying" code could be entered using this > trick, if the instruction format was known. Umm, not quite - at least at machine level. Memory space is NOT really "unified" on such calcs; machine code in ROM is in a different 'space' than data (user programs, data memory, stack, display regs) in RAM. Also, machine ROM code, IIRC, is 10 bits wide. There is no mechanism (instruction, data path, etc.) to move data in RAM/regs to ROM space to execute machine instructions. At 'user code' level - where you program the calculator w/keystrokes - you aren't really programming per se: you're entering *data*. This data is interpreted by a machine code routine that fetches 'tokens' and calling appropriate ML routines. With a lot of handwaving you might be able to 'execute' data. The whole premise of SP is to form normally unpairable commands/operands by using tricks - entering data, then fooling the machine and making it think it's program steps. Bill Wiese San Jose, CA

[ Return to Index | Top of Index ]

Go back to the main exhibit hall