What would your ideal HP calc look like now?

[Ángel Martin]

What a beauty this one is... subject of dreams, or an advance from the current 43 project? Not likely to ever materialize as a reality... but in the mean time the clear choice for me is the 41CL - aided by lots of next-gen math and utilities software. Nothing beats its flexibility and seemingly endless potential. It had to be said!

Cheers,
'AM

[debrouxl]

For making dream machines, being user-moddable and not paying any sort of attention to exam testing restrictions madness are musts

Quote:That seems to be pretty close to what people want (along with an SD slot, USB, BT, IR and an ADC/IO(*) interface :-) Possibly a higher res paper-white display. And of course a flashable µC with the source code, so that we can fix any silly implementation details (like having to add 'h' to numbers entered in HEX mode on the HP 35S).

Several folks at cncalc have recently started doing something along those lines, dubbed the "Arithmax E301". It's made from the outer enclosure + screen (IIUC) + keyboard of a Chinese-made Casio calculator clone, and from a brand-new PCB driven by a STM32F407VGT6 microcontroller based on a Cortex-M4 processor, offering a SD slot, an USB OTG controller, ADCs, DACs, timers, UARTs, USARTs, I2C, SPI, GPIOs, even an Ethernet controller. Obviously, the PCB doesn't expose all of those to the user, due to e.g. the size of the RJ45 female connector.
The device is at an early stage, but it clearly works, and has an early emulator. They ported the Eigenmath CAS engine, which was also ported to various other calculator models. I have no confirmation whether they intend to make this more than a learning project for themselves.

The raw power of that microcontroller beats the raw power of the vast majority of commercial calculator models: the TI-Z80 series, the TI-68k series, HP scientific calcs, Casio scientific calcs, most Casio graphic calcs, are outclassed by a STM32F407 microcontroller. The raw power of the announced $25 Firefox OS phone dwarfs the raw power of the STM32F407.
In this brave new world, hardware is cheap, hardware development cycles are short, and sooner or later, established hardware manufacturers need to adapt to that situation. Monopolies enforced by the education system never last forever in the face of market realities

[lmmt61]

But why complicating things this much? The calculator you are aiming at is an HP 48 GII. It is already here and with RPL, for me much easier than RPN.

[walter b]

Dream machines have a long tradition on this forum. Just to (re-)show you some of the vapourworx(TM) products of the last decade:

2005:

2006:

2007:

2008:

Early 2009:

We began the WP 34S project in December 2008. And we started layouting WP 43S in November 2012, and WP 31S in February 2014.

d:-)

[colinh]

Walter - the WP 34S project is extremely impressive. Even the manual is good :-)

But this WP 43S is just incredible! It looks even more gorgeous than the 45S (from the previous page). It's almost as uncluttered as the 42S (the 34S is a bit hard on the eyes, it has to be said :-)

What does "layouting" mean? Is there a software emulation of the 43S, or is it just a picture?


The wonderful thing about modern microcontrollers is that one can actually implement this sort of thing from scratch. A hobbyist could even mill the case from aluminium or use a Makerbot or whatever. Then one just needs to cannibalize some real calcs for the keys and display... *starts dreaming*.

[walter b]

Please note the pictures in my previous post were and are nothing more than just painted dreams (I had hoped the term 'vapourworx' being a sufficient hint). The respective model numbers had a meaning at the indicated time of publication - they are not, however, linked to any current project. Especially the 43S of 2009 is a very early antecessor of what we pursue with our present 43S project.

When I wrote about 'layouting', that requires having at least a calculator model of definite size, a fixed matrix of keys, and a display we know. Given those basic data, we can design a keyboard layout (sic!) - the user interface of the future calculator. The software design can be begun as soon as the electrical hardware is at least sufficiently known. To give you an idea about the time frame of such a project: Pauli and I began discussing in 2008 what later became the WP 34S - the first real WP 34S was created in 2011. Please look at http://www.hpmuseum.org/forum/forum-8.html for more information about our (HP-30b based) WP 31S and the (new) WP 43S, our currently active projects.

Hope this explains what may have been left unclear. Else continue asking (and please (!) use <Quote> so the forum software can put your post at the right place in the thread - else it may be hard to determine what you are referring to).

d:-)

[Garth Wilson]

We seem to go through this exercise from time to time, but it's always fun, so I'll jump in.

Echoing what Ángel said, the 41CL with time module (and lots of ROM module images) does not lack much to be my ideal calculator. I will add:

• RPN should be standard. Keystroke programming is nice for the short programs typically done on calculators, although other languages should be possible too (and assembly should be more practical than it is on the Nut), which means you need to be able to do text editing on source-code files. Myself, I'm done with algebraic languages.
• I want a reflective LCD screen that can be read in normal room light to bright sunlight. I don't want color, partly because of battery-life issues and partly because I've never seen a color screen that's easily readable in sunlight. I have no need for color anyway.
• I want standard AA or AAA batteries that give months if not years of battery life without replacing or recharging. Rechargeable is ok as long as it does not preclude the use of alkalines. Even when I was using my 41cx heavily every day with instrumentation interfaced through HPIL, a set of batteries still lasted a couple of months.
• It must be made to last for decades. A cell phone will probably be in the landfill in two years, but I still use my 28-year-old HP-41cx every day, and some of the programs that I commonly use have been in it continuously, without re-loading, for nearly 25 years. To really get good at using a particular machine requires a big investment in time that I am not willing to repeat frequently just because the next big thing comes along.
• In the 80's, everyone always wanted more speed and memory; but I think nearly anything would meet those needs today.
• SD card, but also megabytes of internal flash which is available today in tiny SPI ICs.

Quote:a) we like high quality keys and would be willing to pay a premium to get them.

Yes, and there needs to be good key de-bouncing in software too.

Quote:b) a portable/pocket calculator should fit comfortably in a pocket! Ie. be very robust but slim and small. Thus a Pioneer (Voyager) type format.

"Pocket" refers to the size, not the method of carrying. I will never put a valuable thing like this in a pocket. That makes a lot of lint get into the unit, risks dumping it out on the floor if you bend down to access a bottom file cabinet drawer, or, if in a pants pocket, risk breaking it when you sit down, or scratching it with coins or keys or something else you forgot you had in there. Never.

Quote:c) graphing capability doesn't seem to be high on anyone's list.

Right. A single-line screen is fine for me, but it ought to be dot-matrix (like the 71's, not the 41's) so all the special characters can be displayed correctly. Greek letters are used a lot in engineering. Case may change meaning, as with M (mega, or million) versus m (milli, or one-onethousandth). The display ought to be long enough to show both parts of a complex number (room for text to tell what it is, at the same time, would be an added bonus), but I personally have almost no use for graphing. I'll leave the graphics for CAD, web browsing, photo editing, etc..

Quote:d) expansion/interfacing capabilities are nice

For me, it more than just nice. The unit must be able to be used as a hand-held controller, to control equipment and take data on the workbench. HPIL was great. A laptop takes too much room on the workbench, and is not as practical to unplug and carry from the workbench to my desk and back. The laptop's battery life is pittiful, at a few hours max, and I constantly have to be thinking about it. Then it takes time to charge it up again too. I don't use my 41's ability to control instrumentation much anymore, but I still want the capability. USB does not qualify, for several reasons. A serial port would be great too, with user programs able to directly access it at the byte level, unlike the TI-8x calculators; ie, it can't be just for file transfer by underlying layers of software which you can't easily get past. I use the FSI164A (like to the HP82164A) HPIL-to-RS232 to get RS232 with my 41 and 71.

The Arithmax that debrouxl mentioned above has some attractive features in the I/O department. They are a fairly typical microcontroller inventory of onboard support modules. I've developed many commercial products using microcontrollers, and would say that using a microcontroller is probably a given for any new calculator; but when you say "Chinese-made" for keyboards and a lot of other things, right away it gives me visions of things that hardly last until the credit-card bill arrives.

Quote:And still, I'm wondering what people(**) use them for - ie. when/why they wouldn't use a computer.

• better portability
• instant turn-on
• battery life of months or even years without recharging
• no hasles with updating OSs and finding that something doesn't work anymore, at least not the way you liked, or that your OS is too old to run something new that you want
• greater stability of something that lasts decades instead of just a few years max, and not having to frequently transfer things to new computers

and I'm sure I can think of other reasons with more time.

[colinh]

(03-24-2014 08:15 PM)walter b Wrote:  Please note the pictures in my previous post were and are nothing more than just painted dreams (I had hoped the term 'vapourworx' being a sufficient hint). The respective model numbers had a meaning at the indicated time of publication - they are not, however, linked to any current project. Especially the 43S of 2009 is a very early antecessor of what we pursue with our present 43S project.

'Vapourworx' would have been clear, but for the fact that you've demonstrated that you can actually complete a project (namely, the 34S). Therefore I felt justified in considering these images as design studies / brainstormings.

I didn't know that there is a *real* 43S project coming along - and how far it's already got!

Quote:When I wrote about 'layouting', that requires having at least a calculator model of definite size, a fixed matrix of keys, and a display we know. Given those basic data, we can design a keyboard layout (sic!) - the user interface of the future calculator. The software design can be begun as soon as the electrical hardware is at least sufficiently known.

Yup, got it. But of course, the hardware design was only able to begin once the software/UI design was sufficiently known too :-)

Quote:Hope this explains what may have been left unclear. Else continue asking (and please (!) use <Quote> so the forum software can put your post at the right place in the thread - else it may be hard to determine what you are referring to).

All clear. I've read lots of posts by you and Pauli (and everyone else) in the meantime. It's fun to read. Looks like it's probably too late to make hardware suggestions - still, there's always the next project :-)

[walter b]

(03-25-2014 07:32 AM)colinh Wrote:  Looks like it's probably too late to make hardware suggestions ((for the 43S)) - still, there's always the next project :-)

It's never too late to try Please address your suggestions to Eric and Richard in that matter.

d:-)

[David Hayden]

I think there is room for 2 form factors. One is something that fits easily in a shirt pocket. You can hold and operate it with just one hand. This would be like the Woodstocks or the 30b/34s. The second is a graphing calc: larger display. To me the advantage of the large display is that it can show input forms for applications.

Fast, lots of memory, I/O.

I like RPN for entry and programming. The nearly flat learning curve of a keystroke programmable is ideal for students and quick calculation.

I'd rather see calcs that evolve over time instead of ones that are replaced every few years. I wonder if the 41 series would continue to sell if it had been updated over the years to the point of the 41CL now.

A really nice, intuitive interface. Detailed on-board help.

Clear, detailed manuals. These should be available in printed form. I wish HP would team with an on-demand printer to make manuals available.

Dave

[walter b]

(03-26-2014 08:45 PM)David Hayden Wrote:  One is something that fits easily in a shirt pocket. You can hold and operate it with just one hand. This would be like the Woodstocks or the 30b/34s.

I do love Woodstocks. Thus, I didn't put them in my shirt pockets and I also won't do that in future.

d;-)

[Frozen North]

An important part of the calculator experience is the user input interface. Those sturdy, double-molded, sculpted hinged keys with clear tactile feedback as seen in the HP-67 and its brethren surpass all alternatives, then and today. Having to use cheap, mushy keys is not for me. Likewise, using an emulator on a phone or tablet with no keys at all can be just as bad. I don't want to have to look at a keypad while usng it, just like I don't want to have to look at a keyboard. And I don't want to key an eye on the display to make sure that a press registered properly.

[Han]

I don't know what it is (perhaps because my first calculator was an HP48G) but I just really love the look of the Pioneer series. So I'd like my calculator to have that type of outer look and the inside and screen to be as powerful and nice as the best smartphones of today.

[walter b]

(03-27-2014 03:28 PM)Han Wrote:  I don't know what it is (perhaps because my first calculator was an HP48G) but I just really love the look of the Pioneer series. So I'd like my calculator to have that type of outer look and the inside and screen to be as powerful and nice as the best smartphones of today.

(Emphases added)

Hmmh, are you sure you want a Pioneer while you talk about a Charlemagne?

d:-?

[Manolo Sobrino]

Yes, you can have double-shot Cherry keys, but it won't be cheap. What if the keyboard alone ends up costing 50-100$? All the technology for your dream calculator and beyond is here, the problem is big companies believe they won't sell enough 250$ not-for-school calculators, and they're right.

The current situation is something like this:

HP: I don't think their priorities are there. Numbers say Personal Systems are not doing well and they're cutting 34,000 jobs this year. The fact that they are still producing new calculators is kind of a miracle.

TI is moving to Nspire... enough said. They have just one interesting scientific calculator, the 36X Pro, not programmable.

Casio: They really only have the FX-5800P, which is a humble mishmash of features from their golden age. The rest are school oriented or cheap meant for volume (yet I have a feeling that this is a company with enough resources and crazy enough to consider such a thing at some point, it wouldn't be RPN though... and it would have "plastic keys".)

Sharp: their only programmables are Japan only, and fairly modest.

Canon: they're doing interesting things on the low cost end of scientifics, with the oddity of an expensive and limited scientific meant for its looks.

Maybe a small company can develop such a calculator, there isn't even a kickstarter project for such a thing so I guess it's unlikely.

So unless Fluke comes up with a 5k$ calculator your only shot is basically the 43S from the usual suspects and then half of the board will complain that it isn't really an HP

(BTW, most, if not all, the relevant patents have expired, but did you know that you still have to pay Kinpo for an UNDO key in a calculator http://www.google.com/patents/US20030149710? You live and learn...)

[Raymond Del Tondo]

(03-27-2014 03:38 PM)walter b Wrote:  Hmmh, are you sure you want a Pioneer while you talk about a Charlemagne?

The HP 48 Series are also called "stretched Pioneers", and the overall design rules of the normal Pioneers also apply to the HP 48.

[Han]

(03-27-2014 03:38 PM)walter b Wrote:  

(03-27-2014 03:28 PM)Han Wrote:  I don't know what it is (perhaps because my first calculator was an HP48G) but I just really love the look of the Pioneer series. So I'd like my calculator to have that type of outer look and the inside and screen to be as powerful and nice as the best smartphones of today.

(Emphases added)

Hmmh, are you sure you want a Pioneer while you talk about a Charlemagne?

d:-?

Yeah, I'm pretty sure. Take a good look at Charlemagne and tell me if you see much of a difference between that and a Pioneer (other than screen size). Alcuin was just a recolorization of Charlemagne. So in terms of looks (as per the thread topic, in a literal sense), all Pioneers, Charlemagnes, Alcuins look like they all belong to the same familiy.

[Han]

(03-27-2014 03:45 PM)Manolo Sobrino Wrote:  (BTW, most, if not all, the relevant patents have expired, but did you know that you still have to pay Kinpo for an UNDO key in a calculator http://www.google.com/patents/US20030149710? You live and learn...)

Um, doesn't the HP48SX provide prior art? Or is the patent specifically for a single key labeled "UNDO" ?

[Manolo Sobrino]

(03-27-2014 04:39 PM)Han Wrote:  

(03-27-2014 03:45 PM)Manolo Sobrino Wrote:  (BTW, most, if not all, the relevant patents have expired, but did you know that you still have to pay Kinpo for an UNDO key in a calculator http://www.google.com/patents/US20030149710? You live and learn...)

Um, doesn't the HP48SX provide prior art? Or is the patent specifically for a single key labeled "UNDO" ?

Apparently it's the key (so it's different from shifted keys? patents are funny):

Quote:To achieve the object, there is provided a calculator capable of recovering cleared values, which comprises: an input unit including deletion keys for clearing currently displayed data or recorded data of the calculator, and an UNDO key for recovering the cleared data; an output unit for displaying the status of the calculator and operating results; an input/output buffer coupled to the input unit and the output unit and providing a display buffer for temporarily storing input and output data; a memory for storing program codes and providing memory space required for operating; an algebra logic processor coupled to the input/output buffer and the memory for providing arithmetic and logical operations; a stack register coupled to the algebra logic processor for temporarily storing previously cleared data by one of the deletion keys, wherein whenever the displayed data or the recorded data is cleared by pressing one of the deletion keys, the cleared data is pushed into the stack register; and a flag register coupled to the algebra logic processor, and being set when the displayed data or recorded data is cleared by pressing one of deletion keys, whereby, when the UNDO key is pressed and if the flag register indicates that the displayed data or recorded data was cleared by one of the deletion keys previously, a data entry is popped from the stack register and the flag register is cleared.

[walter b]

(03-27-2014 04:34 PM)Han Wrote:  Yeah, I'm pretty sure. Take a good look at Charlemagne and tell me if you see much of a difference between that and a Pioneer (other than screen size). Alcuin was just a recolorization of Charlemagne. So in terms of looks (as per the thread topic, in a literal sense), all Pioneers, Charlemagnes, Alcuins look like they all belong to the same familiy.

Thanks - I learn something new here every once and a while. So perfectly pocketable calcs and impossibly pocketable calcs belong to one product family? O tempora ... #-(

d#-(