It's time for the user community to design a real calculator for HP and present it for manufacture.

Here's what I want:

Clamshell case slightly smaller than HP28S and rotated 90 degrees.

entire left (top) side is a large lcd taking up most of the space on the lid

the bottom half is a keypad with a layout similar to Voyager, six rows by 12 columns. A large ENTER key at position 3, 1. An additional row of smaller soft menu keys closest to the top edge.

They keys are injection-molded and reminiscent of the HP 34C. Shift keys in yellow and blue labeled f and g. Alpha key labels are in black on the downslope of the keys, in querty format accessed with a black shift key marked with a greek alpha.

Most salient math features available on keycaps and with f- and g- shifts, including percentages, statistics, exponentials, trig, stats, object delims and STO, RCL, GTO, GSB, LBL, PSE, R/S, PGM, etc.

Keystroke and RPL programmable.

Runs on 4xAAA or 6V wall wart.

Compact flash slots on each side, IR window on the right RJ45 on the left. Uses stock digital camera memory for ROMs in the same form factor for memory expansion or aftermarket application software. Yes it has a tulip or other ethernet chip so that PC coms go via FTP.

Runs a StrongARM uP at ~200Mhz with a meg of internal ROM, 8 meg of internal flash.

Generally behaves like a blindingly fast 48GX with the added bennie of keystroke programming in 41CX/42S style.

CAS might be nice, contact manager, appointments and todos would be even better.

The damn thing is a Calculator with a couple of PDA features that geeks and gnurds might like. It is fully field programmable with outside assistance from a PC. It pretty-prints your RPL for editing.

Who wants to do a mock up of the art work?


YES! YES YES YES! This is the kind of thing that would put HP on top again in the calculator world! I'd gladly pay a few hundred for such a beauty as this...


I'd be willing to pay up to about $600.

Here's what I have in mind:

Somebody do some art work so that we can play around with how it looks so that we can have a mock up on a number of web site to recruit the substantial talent required to pull this off.

I've applied for a project over on sourceforge.net so that we can archive the work. Let's plan on using eCos or a radically stripped version of Linux 2.4 for the starter on the platform core. Design the core calculation engine as a GPL application that runs on the platform core.

Recruit dedicated volunteers to work out the core hardware architecture:

- CPU and memory

- keyboard interface

- display interface

- ir port interface

- ethernet (or USB) port interface

- power management

- system clock

- expansion card interface

Recruit software developers (#include ) to develop an emulation bench for the core software system that runs on arbitrary Linux platforms.

Recruit mechanical engineering talent that grew up on Real HP calculators to help out with packaging

Manage the sharing of all of this under the terms of the GPL.

If it looks like it's coming together, start to shop it around at HP, TI, Canon, Casio, Compaq WRL, Palm, Transmeta, etc to get it built.

It may not take ten years. Let's shoot for being able to get something to power up on July 1, 2001 and to be available at retail from some manufacturer one year after that.


It's most interesting, you thought about building an own calculator. Unfortunately I don't believe, a lot of people would pay $600 for a plain calculator (I would !).

I've had such thoughts too, but without Linux (which would nail the project to a very powerful and battery-hungry processor).

I'd prefer an embedded processor and programming the calculator in assembler (no flames please). For selling an advanced calculator, the device shouldn't cost more than double the price of a HP-49 or TI-89. The power factor is very important (would you like a calculator, which needs battery refill every week, think about your mobile phone, that's annoying enough).

Anyway, this would be a large project and I don't know, if we could acquire such a large number of designers for developing a complete machine. A university could help here, if some parts of this device could be handled as diploma thesis.

Might be, that a StrongArm could be the right processor for a powerful solution, for very cheap devices there would be the PIC or AVR processor (which could prohibit a HP-49G like device, but it would be ok for testing some concepts)

I'd think, that a calculator needs a well-suited, especially written OS, not an emulator running on a general-purpose OS kernel.



We'll need electrical & mechanical engineers. We'll need software developers for the kernel and for the calculator engine.

I suggested StrongArm because it's used in the Apple MessagePad 2000 series machines with favorable battery life and there's already a full-scale port of Linux and the compiler tools for it. A nice feature of the StrongArm is the lack of an FPU that would tempt us to take short-cuts. We'd have some flexibility to explore software implementations of the math engine. The Mototola chip used by the Palm devices (and the TI-92?) also has a port of kernel and tools. And then there's Crusoe from Transmeta.

The only thing I think would remain of "linux" after bootstraping this thing is parts of the boot code and perhaps part of the MMU.

I've been granted the project at sourceforge. Before next weekend I'll have a page up for it and maybe a few drawings. You're welcome to participate in whatever capacity you feel suited for.


I'm an electronics engineer (communications engineering). I could dive into the datasheets and try developing a concept. We'd need some mechanical drawings.

I could also try writing some software, but my knowledge of the high concept of programming is rather limited (I've written a multitasking system for an embedded application long ago for a project, but I'm not an expert in math algorithms). Anyway, I'm able to learn. I have been looking around for algorithms some time ago (You see, I also wanted to do such a thing).

I can think about the construction possibilities after we have some other guys, who want to make suggestions abt. the capabilities of the future device.

We'd also need some drawings and design studies. We should get a mechanical engineer rather fast, so that we can see, what's possible.

I really hope, that there are enough people with enough time to build such a device.


Don't want to dampen anyone's enthusiasm for this project, but it seems wildly ambitious at the moment. Maybe you need an intermediate step like upgrading an already existing calculator series. Maybe a retrofit for the voyager series. Anyone can buy a 12C and perhaps do a board change and key swap to create a XP 6x HX calculator. Seems like a faster way to market, and a first step to prove your point. I would love to see an upgrade to the 41C and believe that would be very successful.


An 41C clone would be much easier to design and would be a very nice unit for a lot of people here. Unfortunately I fear, that a HP41HX (or whatever) would be more costly than an 48GX and I doubt, that it would have the pretty professional look, that the 41 series or the current HP calculators have. Do you think anybody would like to buy a 41 for $300-400. I doubt, especially, when on eBay there are cheaper 41s with accessories (and eBay is a really expensive place for 41s these days).

Unfortunately the mechanical part is the most costly part of such a project (do you know, what a form for plastic parts would cost ?). There are no ready-made housings for portable devices. The ability to insert ROM extensions (which is a must in my point of view is especially expensive, when you consider there are additional mechanical parts.

Preproduction costs could easily reach abt. $10.000 or more (without considering anybodys time into this calculation). I'm not in the position for financing that (If I'd be, I'd be the first building a new series of high-class calculators). I would have to work full-time and so I'd have to be paid too, or I would starve and couldn't pay the rent, which would lower my productivity somehow.

Unfortunately, besides Charles. I'm the only one til now, who wanted to contribute (for free). Let's hope, there are more guys willing to and there's a broader discussion on this topic.

I've worked in a company which produced small-series items for the consumer market and the company had to live somehow too. The product surely was leading in the world (model railroad digital command control) and in technological leadership, but it was a product, which wasn't easy to sell considering the large number of competitors.

The product we want to design (produce ???) for free (or cheap, it's hardware !!!) would possibly have even HP as a cheaper competitor (the time of costly HP calculators is over, sadly).


I think we agree about the mechanical part being the costly "show stopper" here. The electronics is not trivial, but more managable. I was just trying to suggest a course that would re-use the mechanical portion of an existing calculator. The 41 being the most popular and widely used to date ( maybe? ). It doesn't necessarily have to behave like a 41. Or maybe it would have a personality module that plugs in to the port. Something like that.


I've been llosely looking around for such a device for some months now. Unfortunately I couldn't find anything til now. What would be the alternative? Buying another calculator (with a good keyboard - only HP's possible here) and implanting a new personality?

I think there wouldn't be any problem taking a 48GX, but that's the unit, which draws the most critics onto it (too large, ...). Besides that, I think the 41 is not that heavy (I have to check this now).

Besides that, my shirt pockets are HP-48 compatible somehow, even if it doesn't fit very nice :-)


Like you, I have been looking for another solution also. I am not as ambitious as others, although I admire and encourage them. My plan was to buy a HP 1000CX DOS based machine and load emulators for my favorite calcs. But I forgot about the screen resolution problem with these old PDA's. Most of the emulators/simulators won't work with this screen. I still think this solution might work. You need the software that simulates an HP RPN calc and remaps the keyboard. Keyboard overlays could be fabricated rather easily. I picked up a double-speed 8 meg machine rather cheaply on ebay - so memory and execution speed is no problem. But what a powerful calc it could be, if I could only manage the screen problem. Since I was already moving in this direction, I would choose this hardware to modify. This will not be everyone's solution since those old PDA's are not shirt pocket size. I would be willing to trade smaller size for greater power. I used an HP calc application on my 200 LX for almost everything I ever did on any of my other calculators. The only thing I missed was the I/O capability of my 41C. I would wish the calc software would take advantage of the RS-232 port on the hardware. If you want to take this argument to the extreme, HP has dropped this line of PDA's, maybe you can buy the design from them before they throw away all the documentation. That might be cheaper than starting from scratch. How would such a device be better than a 48 or 49 ? Depends on the software. The display is certainly better. The keyboard - depends on overlays and how everything maps. Couldn't be any worse though. I can't use the keyboards on the 48 or 49 . Too hard to read and too many shifted functions.


Look at StrongARM II at Intel Web-pages.


You guys are too much. Don't give up on this idea, market it or get together with the guys at Ashtech and develop a new handheld data logger to replace the 48gx. I don't want to be forced to use that $%&% 49! Check out www.rpls.com and then select "Ashtech discussion board" and give Bill some ideas.


The only thing I don't like is the format : I want it vertical and not clam-shelled. Everything else reminds me of an idea already described long ago. But it seems that we will never see that dream machine ...

Still, keep going. Who would have thought 10 years ago that we would see a thing like the HP 49G ?


Actually, i don't want to wait for ten years until now again... even for this dream calc.



No, clam shell and voyager style is the best ergonomic HP calculator I've ever used.

But maybo not to put too many functions on the keyboard : look at the 41C : few functions on the keyboard, all most often used and though a looooong catalog of other functions.


I understand. For the clamshell format I want all of the keys to be reachable by thumbs if you're holding it in both hands.

How's this, three form factors:

- like an oversized 15C with a big flip-top LCD, ethernet, IR, two card slots, etc.

- like a 42S with a two-line display and IR

- like a 48GX with a bigger display, IR, ethernet, and one card slot

The key issues are going to be THE KEYS, RPN and FIELD PROGRAMABILITY w/KEYSTROKES (mandatory) and RPL (optional). Once we get the core platform ironed out we'll see that we can actually pull of wrt to form factors.


I totally agree with you.

Please add: powerful financial solutions.


NOT bigger and heavier then the 42 please, It must be shirt-pocket-compatible.



Check out the top story on Slashdot.org as of the time of this posting:

A group in the Netherlands has published for specs for a StrongArm-based wearable/portable!

The Site - www.lart.tudelft.nl - is currently under the influence of the slashdot effect but I'm looking forward to seeing these specs.

This could be a big leg up...

Of course the real hard part's gonna be the buttons, not even HP can do them right any more ;-).

Possibly Related Threads...
Thread Author Replies Views Last Post
  Tricky XP 34 S Enter key Christophe Dubreuil 2 643 08-21-2012, 05:41 AM
Last Post: Christophe Dubreuil
  O.T. XP won't use all RAM designnut 7 1,001 12-27-2010, 02:23 PM
Last Post: designnut
  O.T. RAM uce in XP OS designnut 5 874 05-20-2010, 01:15 PM
Last Post: Norris
  Error message in Hpapine, HPGCC,XP Jan Kromhout 2 567 08-30-2009, 11:53 AM
Last Post: Jan Kromhout
  O.T. XP RAM utilization designnut 1 476 06-04-2009, 12:06 PM
Last Post: Dave Britten
  Start EMU48 instead of calc.exe in XP - possible? George Bailey (Bedford Falls) 7 1,210 04-09-2009, 03:33 PM
Last Post: Gordon Strickland
  Windows XP connection to 48SX Scott Larson 2 503 12-27-2007, 04:17 PM
Last Post: Ivan Nejgebauer
  41/71 to Windows/XP - OS/X file transfers via RS-232, who is doing it? Egan Ford 2 515 10-15-2007, 01:30 AM
Last Post: C. Dehnke
  EMU71/SERIAL1 and Windows XP Egan Ford 1 440 09-05-2007, 12:40 AM
Last Post: RonHudson(USA)
  Beaming Windows XP <-> 50G works Mike Ingle 1 454 04-11-2007, 04:03 PM
Last Post: Massimo Santin

Forum Jump: