|Re: RPN Manufacture -- What are the Statuses of HP's "RPN Patents"?|
Message #18 Posted by Paul Brogger on 17 Mar 2003, 9:22 p.m.,
in response to message #17 by Vieira, Luiz C. (Brazil)
Yes -- I experimented a bit with a Pioneer (17Bii) and found that, indeed, it should be possible to simulate keystrokes using (to my primitive notion) magnetic relays. The old Pioneer case and circuit board with a 17-line ribbon cable attached at various points is still in my electronics box.
Tony Duell sort of took over -- he wanted to use some more sophisticated electronics to effect the switching. It ended upon his "back burner" at last report, and I haven't heard anything since.
As a side effect of that investigation, if I remember right, Tony found that the second set of IC pads on the -42s circuit board is for an alternate PROM. The internal ROM could be disabled by tying a certain pin high or low, and the external ROM would be utilized instead. This promising development implied, for example, that one might be able to convert a 17BII circuit board to a 42s with a bit of electonics work.
But, I haven't heard anything for a long while -- maybe Tony will chime in?
I must say, as long as the HP-48 series is readily available, it's kind of overkill to go to extreme lengths to re-make other RPN models. True, the -42s size and 4-level stack are nice, but it's not like the -48 is absolutely unusable . . .
Also, I've chimed in on this topic in another way . . . If someone wants to make an RPN calculator, take a look at the TI-83 and -89. These are in many ways general-purpose computer platforms, with industry standard CPU's (Z-80 and M68000, respectively), assembly language development tools, emulators, computer links, web sites, discussion boards, and (really) tons of documentation -- with all but the calculator absolutely free. . . . and the devices themselves are really not at all bad pieces of equipment!
I wrote a simple four-level RPN program in "TI-Basic" to try interface ideas -- X, Y, Z, T & Last X on the screen, simultaneously represented in both decimal and hex ('cause that's what interests me most), and with most of the typical math functions (not %!) executed when the appropriate key is pressed.
It's too slow and unresponsive to use regularly, but given what how that went, I'm sure something more satisfying could be implemented in assembler. But that's where it stands -- I haven't touched the thing in months, and haven't yet come up to speed on the assembly end of things. (In fact, when I get right down to it, my working 32S is all I need, and there are more important things . . . )
So, I'm sounding shrill and repetitive, but those of you wanting to customize your own RPN device should take a look at the current TI offerings.