HP Forum Archive 13

[ Return to Index | Top of Index ]

32sii Memory upgrade Message #1 Posted by charles on 28 Sept 2003, 8:23 p.m. Hi all, Have any of you attempted to upgrade the memory of the 32sii? Maybe 128k? I am new to circuitry so don't laugh if this is an obvious impossibility, but any thoughts would be nice.

Re: 32sii Memory upgrade Message #2 Posted by Andrés C. Rodríguez (Argentina) on 28 Sept 2003, 8:39 p.m., in response to message #1 by charles I understand it is not possible to do so, but even if you manage to connect such memory to the calculator electronics, the software which runs the HP 32 Sii would not look for such memory, nor "see" it; and hence the memory will not be of any use. Any expandable machine needs to be designed and programmed with expansion in mind to become so. HP 41, 42 and 48 models are able to recognize memory beyond their original capacity, but (AFAIK) other models are not.

Re: 32sii Memory upgrade Message #3 Posted by Tizedes Csaba on 29 Sept 2003, 2:19 a.m., in response to message #1 by charles I'm not a 'nibble-man', but I think the hardware can't to attach more program area: This calculator have got 384 bytes of memory and one operation take 1.5 bytes. So, (384 bytes)/(1.5 bytes/step)=256 steps. The program-counter maybe sized only 1 byte... Csaba

32Sii 3-nibble instructions Message #4 Posted by Victor Koechli on 29 Sept 2003, 8:14 a.m., in response to message #3 by Tizedes Csaba This is interesting! Are you sure there are only 3-nibble instructions in the 32Sii's program space? You would think that the limited number of programmable commands would allow for most instructions to be only 2 nibbles (1 Byte)... Cheers, Victor

Re: 32Sii 3-nibble instructions Message #5 Posted by Christoph Giesselink on 29 Sept 2003, 12:20 p.m., in response to message #4 by Victor Koechli >Are you sure there are only 3-nibble instructions in the 32Sii's program space? No there aren't. There's one exception, most (not all) real numbers have a three nibble prolog (#000) followed by a 16 nibble BCD floating point number. Christoph

Re: 32Sii 3-nibble instructions Message #6 Posted by Tizedes Csaba on 29 Sept 2003, 3:57 p.m., in response to message #4 by Victor Koechli The 32SII haven't got 1 byte long instruction, only 1.5 bytes long. The integers from 0 to 254 use 1.5 bytes, all other numbers uses 9.5 bytes (from the B-2 page of manual). Csaba

Re: "Meta-calc" Message #7 Posted by Paul Brogger on 29 Sept 2003, 9:17 a.m., in response to message #1 by charles I suspect that it would be possible to insert a new layer of logic an memory between the original 32SII chip and its I/O (keyboard and LCD). Some currently unrecognized keystroke combination would bring the new environment to the fore, disconnecting the original chip temporarily. The new layer would replicate the 32sII program editor, but enhance it with named program storage and (perhaps) named register arrays. New commands would cause the loading of a named program into the actual 32sII's program memory by keystroke simulations. Likewise, resulting data (the 27 registers) could be copied out to the new memory as an array of values, perhaps to be reloaded later. The new electronics could include fast stream I/O capabilites and could be supported by PC links and software. The physical ciruit board might be installed in place of the original, with the 32sII board then attached as a daughter board to the new one. Taking it a step further, one might replace 32sII electronics and logic entirely, with a new, more capable implementation. Buutttt . . . H-P appears ready to solve at least some of the 32sII's problems with the 33s (albeit, while introducing others!), so it would seem to make sense to wait and see just how that unit behaves before embarking on an elaborate upgrade for a discontinued model. Another approach might be to get one of the other new models (48GII or 49G+) and obtain or develop a four-level stack emulation for that environment.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall