despite the likelihood of more squishy keys


Or squishy keyed calculators, I still am not holding out any hope that I'll be finding affordable rpn (or even algebraic) programmable pocket sized models anytime soon.

So- has anyone examined the 20S? I seriously doubt there are any upgrade options there, but if anyone has looked....

an extra 128steps that wouldn't clear when you used a built in would make that little inexpensive calculator a really nice thing- even if it isn't rpn.

On the other side- is this a model where the rom instructions are seperated? could something new be burned and used?



Is this in the hope of using the HP20S as a basis for an RPN hack? Once the HP20S goes you might as well look round for a used 32 or 42...

I can't see the future of the HP20S being very good. All it's elder brothers and sisters have been withdrawn so the chances of HP producing this one for much longer are slim. (OK the 12C has been going for decades but I don't think the 20S has the same niche market).

Sorry if I've got the wrong end of the stick.


Well, it's not so much specifically about an rpn hack to the 20s (though a $30 rpn programmable would just be nice)- it's more about the idea of something maybe being hackable that is available for less than $200 on the used market :)

Granted- the 32sii is there. But the it's not upgradable in ram- which is the only thing I'd really be concerned with in it. It's also running closer to the $10o mark now than it is to the $30- or even $50 mark.

I suspect you are right, and the 20s won't be available inproduction for that much longer. But we've all got ample opportunity to grab a couple, and if there is some basis for hacking... well, it'd be nicer than not, I suppose.



I doubt very much that a 20s can be hacked at all. I haven't taken one apart, but all of the other "later" Pioneer models are single-chip designs, with single-sided exposed copper (not multi-level gold-plated) circuit boards. (The 32SII PCB is a one-component design: the surface-mount chip, with connections to the keyboard, display and battery.)

Just because we can upgrade the RAM on the 42s, don't get your hopes up for any similar 20s modifications. In fact, the 42s has two sets of surface-mount pads etched onto the PCB: the RAM set and one for an alternate ROM. The lower-end and later models have no such avenues of extensibility built in. (Unless I'm missing something REALLY subtle!) Even the 32sII's precursor, the old 32s, which shares the 42s' technology (gold-plated PCB and chip "riding in a hammock" of micro-fine connections) has no external memory pads or other obvious means of expansion.

I bought a TI-83+ Silver recently, but haven't started messing with it yet. (Now that I've bought it, I'm not sure I'll ever have time . . . ) It's a Z-80 based unit in a pretty solid package, with a graphic matrix screen and something like 1.5 Mb of Flash ROM. A Zilog assembler and SDK, a Windows emulator, and (literally) hundreds of pages of TI documentation are available for free. It ought to be available for a while, and it may indeed be amenable to fundamental redesign. (Though there are now only two limited modes offered for software enhancement, and documentation of the calculator OS "kernal" is apparently still proprietary.)

Also available is the TI-89, which is 68000-based, with similar support tools and documentation available. (And both have available PC connectivity, too.)

I think it would be fun to get an interested group cooperating on a next-generation RPN device by all agreeing to work through some available, promising technology with adequate developer support . . .


Well- The ideo of hacking a ti89 isn't bad. I still don't like the keys much, and I HATE with a PASSION the XYZT idiocy in the keyboard layout.

I guess I'll have to open a 20s and find out- it may very well be unhackable, but I'm not sure- it's hard to be positive with some of the stunts that got pulled in later generations. This is why I'm asking if anyone has pulle done apart and looked.

This does not preclude other efforts involving rewriting the ti89 or something. There is room for more than one thing at a time in larger groups :)



There is absolutely nothing that can be done with a 20S, from a hack standpoint. It is a single IC chip on a PCB with no chance of doing anything to it.

I have pictures and can post if anyone is interested.


Ah well, it was a forlorn hope- I just didn't want to take one apart to see.

I wonder if they are really going (gone?) out of production (samson cables seems to think so).


The single-chip-on-a-circuit-board implementation of the 20s is what I suspected (having seen the 32sII), and apparently what it is.

That doesn't mean it's absolutely unhackable -- just that there's more involved in the hack . . .

The tolerances on that circuit board can't be impossibly precise, and the electrical characteristics of the other system components (LCD, keyboard, and batteries) are probably knowable. The task is to come up with a ~4.5 Volt circuit which scans a Pioneer keyboard and generates a readable 20s display. (Some kind of PC interface would be almost essential as well, if it's being programmed from the ground up.)

Now, if a low-power, supported microprocessor with some RAM and Flash ROM could be put on a ~2.5" x 4.5" circuit board and interface to the 20s (or other calculator) infrastructure, you just might have something one could work with . . .


Well, duh, yeah, I didn't think that all the way through. Paul is absolutely right. The board dimensions can be measured and we know there is a lot of available space for a new PCB (alla a 42 or 17). The keyboard is easily decoded as it appears to be a straight 7x6 matrix.

The only issue for me would be the suitability of the 7 segment LCD. I think this might be the limiting factor.


. . . and as we've been reminded (I think by Tony Duell, if I remember correctly) we should keep in mind that there already exists, in very good quantity and quite good quality, an extensible RPN calculator with a PC connection available and a wide body of pertinent expertise, information, and support . . .

The HP-48G/GX.

Really, what would we be buying for all the effort? The difference in form factor (length, thickness) between a 48G and a 42S. Just how much time and effort is that actually worth?


Well, "because it's there"

Alternatively- there are a lot of applications for which I *enjoy* the small, easy, basic, keycode programming of the 20s- I just want it to be RPN and have a big non-overwriteable space on top of the 99 lines where software "loads".

I'ts beyond my skill level for sure- but the amount I could learn helping- even just tracking the process would be worth donating several calcs. :)

Alternatively- this could be the hardware answer for the perennial "homemade rpn" subject. More than a 4 banger, but no so much more that it becomes impossible.

Keep everything in there, make |<-| |x^2| into the the roll key, and it go with it.


I like "Because it's there" as an answer.

I started programming on my HP 29c and on small systems, and have always longed for a general-purpose computer in a Pioneer-sized package.

After half a career of mainframe programming, I still enjoy using my 32s and 42s models. (In fact, I enjoy the thought of using them far more often than in actual practice!) I've used the 28s and 48g some, but there's something compelling about the minimal, and perfectly-realized 42s that strikes a responsive chord with me.

However, my "rational side" keeps nagging, "What, really is all the fuss about? Use what works and get on with life!"

I used to note with chagrin that, on my death bed, I wouldn't be terribly concerned with (for example) how our company database encoded "institution type" values in the high-order four bits of a halfword binary value . . . Am I only lately getting around to accepting and acting upon the implications of that realization?

And yet, I don't think I've ever been so alive as when I've been in hot pursuit of a gnarly technical problem, while trusting I posessed the knowlege, skill and tenacity to (real soon now!) arrive at a diagnosis and suggest a cure . . .


I have similar feelings- when I'm building a desk, fixing the VW, hacking some broken config on my mail server. The whole reason I do so much of this myself when there's perfectly good other ways to go about it is that feeling :)

I've sort of flirted with the rpn 4 banger thing- I haven't done a lot of electronics work (especially compared to some of the people here)- but I've built a set of test equipment from kits, some radios from kits, and then more radios from plans.

Seems to me that the 7seg display and relatively uncramped keyboard are advantages in this case, rather than disadvantages.

I suppose I should start looking for display models or abused 20S calcs to take apart. Or, if the clearance prices are good enough....



(Flight of Fancy) The new cell phones that flip open to a PDA with a keyboard...Maybe those will have potential for someone to write an RPN calculator program. The models I've
seen pictures of have a distinct keyboard...

Possibly Related Threads...
Thread Author Replies Views Last Post
  HP 19BII Versions... Beige Keys and Black Keys? Eric 0 342 03-17-2006, 04:43 PM
Last Post: Eric

Forum Jump: