|Re: HP-45 simulator for the HP-20b|
Message #12 Posted by Scott Newell on 9 Aug 2008, 12:13 p.m.,
in response to message #11 by Walter B
True. But it may demonstrate the opportunities -- one working model is worth a thousand words.
Hate to throw cold water here, but I'm not sure the 42S is a feasible candidate.
First, the display is radically different. Even if you were to somehow transplant hardware from a donor calc, the CPU doesn't have enough lines to drive more than 400 LCD segments--the HP20B is maxed out. If you're going to replace the display and add a driver, you might as well start from scratch with a new PCB design.
Doesn't the 42S contain 8kB of RAM? That's 2kB more than what we have to work with on the 20B, and there's going to be some overhead in the simulation. Could you even run a simulated 42S with 2k or 4k? Sure, you might add RAM with the internal GPIO pads, but now you're targeting not only people who can re-program the 20B, but those that are willing to open it and perform surface mount surgery.
How large is the 42S ROM? I've seen mention of 64k, but I'm not sure if that's words or bytes. If bytes, you might squeeze it in. If words, well, there goes your entire flash just to hold the simulated ROM. Adding memory is an option--see above.
To me, the keypad is the only bright spot--the layout appears to be a match in many respects.
I'm not even going to bother researching the 42S (CPU speed? legal status of firmware image?) until a viable display workaround pops up.