HP Forums

Full Version: Sneek Preview: Anniversary Edition
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

Here it is...a fantastic 4-function Voyager...


4F? Must be milspec.

Regards,
Howard

I would buy it :)

List it on eBay, see if anyone bids!

Dave.

I really like it, provided that it has a reasonable price and fits comfortably in my shirt pocket.

It seems to me that it would be easy enough to manufacture. After all, shouldn't it be simpler and cheaper to design and manufacture a calculator with an RPN user interface than with an algebraic user interface?

But it seems to me the a "35th Anniversary Edition" model ought to have at least the "scientific" capabilities of the HP-35, and be "shirt pocket-sized".

Regards,
James

Very very beautiful and tiny.

I'dd add a big enter horizontal key (where now are the enter and + keys), shifting the four operations one row up, and using the three upper left keys for square root, percentage and reciprocal.

With these add-ons, this could be my most used calculator, since I find myself using these keys the 99% of the time (for trigonometry and exponentiation I could use another calculator, after all).

-- Antonio

P.S. could you update the picture as I suggested? Just for fun!

Edit: I did it for you.

Here's my result:



... and of course, if a key could be destined to a shift function, a lot of alternate operations could be added! But let's keep it **very** simple...

-- Antonio

Edited: 21 May 2007, 11:28 a.m.

Hmmm... So is there a flimsy dip-switch on the back to toggle RPN mode from the factory-set algebraic entry?

Please excuse me while I charge at those giants with my trusty slide rule!

Kinda awkward, not having X<>Y, Rdown, CLX...

Hi!

Quote:
Kinda awkward, not having X<>Y, Rdown, CLX...

As for x<>y and Rdown I couldn't care less, but without CLX I would be lost instantly...


And anyway, a real 35th anniverary edition calculator must have one main feature in common with the 35year old hp-35: the LED-display!

Like my Sinclair scientific here, that has got a CLX-Key, but no decimal point instead:


Greetings, Max

Quote:
4F? Must be milspec.

Regards,
Howard


Should be HP-4B as in 'four banger'!

Very nice Photoshop work ;)

Regards,

Nelson

I'd buy that one, if it weren't too expensive.

Regards,
Howard

But I'd exchange the X<>Y and CLx key positions. Having Clx next to the + key invites disaster, particularly with the lack of lastX. X<>Y is inherently reversible.

Regards,
Howard

Well, you're right.

What to say? If I have time, tomorrow, I'll fix this bug.

In any case, a four-op calculator in pure RPN mode, with a four-levels stack, with 10 or 12 digits (not more) and with that beautiful interface like the one suggested by Chuck, would probably cost around 15-20 dollars, and be the entry-level calculator with RPN input for thousands of algebraic users that cannot even conceive this usage (I'm not speaking of those who use algebraic calculators and know RPN and have chosen the former - I have already done such a mistake that arose a holy war some time ago).

For what about me, I'd buy 5, two for me and one for each member of my family. And maybe others for all my friends and parents for birthdays.

I like the idea.

Chuck, you made a start. Hopefully, HP will gather your suggestion. Who knows? Who cares?

-- Antonio

But for a very simple RPN non-programmable 4-banger calculator, wouldn't just 2 stack registers suffice? After all, each of the available operations always takes 2 arguments and returns 1 result.

Okay, even with just x and y registers, x<->y and CLx could be useful.

What I might add would be a "memory" register, with M+ and MR keys to use it, a +/- (CHS) key, and maybe a 1/x key, and even "dollar store" "4-bangers" usually seem to include a square root key.

I've never seen a percent key as being particularly useful on a calculator, but it seems that most seem to expect it.

How about an On/Off key to conserve battery power?

Yes, I really think that HP ought to market a low-cost entry-level RPN calculator to introduce algebraic users (and youngsters using a calculator for the first time) to the ease and simplicity of using RPN models.

Regards,
James

Thanks all for taking part in my fun. It's also quite interesting that if we include all the wants and desires of everyone we end up very close to the 11C or 15C. I personally hope HP keeps the size similar to the 15C. I'd love a 15C with 1gig of memory and a jack to dump programs to and from a computer. But a 4+ banger would be pretty cool, too. :) Cheers.

Quote:
I'd buy that one, if it weren't too expensive.

So would I -- several, probably.

Several of the old 4 function RPNs like this one saved valuable key-al estate by using the + and - keys as positive and negative enter keys. They gained a change sign key in the process of loosing an enter key. This one could use the added two keys for a square root and what the hell, let's live dangerously, an ON key.

It's getting closer! I almost busted out my wallet when I saw the revised version. :-)

thanks,
bruce



Well, the "C" key is pointless, as one could use the on/off switch to clear 8)

Here's an **ugly** attemp to figure out the calculator after your suggestions.

It's a collection of cut&copy, so the yellow of shifted functions is not the same, but just imagine it is.

What about it?

Of course, in my mind:

The OFF should be automatic after 5 minutes, the ON happens when you hit any key; the RESET could be a back pencil hole (as in old SHARP calculators), or the batteries removal (there's little chance that this calculator hangs, anyway).

STO 0..9 are ten available memories

STO +/-/÷/x 0..9 work as accumulators (e.g. STO+6 add reg. X to memory 6)

f RCL 0..9 recall ten available memories

RCL +/-/÷/x 0..9 work as accumulators (e.g. RCL+6 add memory 6 to reg. X)

f 1..9 could set the FIX

f 0 could reset FIX

scientific format starts automatically when number doesn't fit the 10/12 digits (if you need to input 1E-25 probably you need a scientific calculator...)

the greek pi is a gift (one could calculate circle areas, right?)

well, that's all. Any need of a manual, for such a calculator?

-- Antonio

P.S. I invite you to find some substitutes for % and pi, since they are really the lesser useful keys of this proposal.

I also invite you to riconsider this: if the available memories were, say, 0..6, there would be three shifted function to add. What do you think could be added to enhance this calculator, without losing the point it must be a **simple** calculator?

Edited: 22 May 2007, 7:25 a.m.

Quote:
Here's an **ugly** attemp to figure out the calculator after your suggestions.

It's a collection of cut&copy, so the yellow of shifted functions is not the same, but just imagine it is.

What about it?


Of course, in my mind:

The OFF should be automatic after 5 minutes, the ON happens when you hit any key; the RESET could be a back pencil hole (as in old SHARP calculators), or the batteries removal (there's little chance that this calculator hangs, anyway).

STO 0..9 are ten available memories

STO +/-/÷/x 0..9 work as accumulators (e.g. STO+6 add reg. X to memory 6)

f RCL 0..9 recall ten available memories

RCL +/-/÷/x 0..9 work as accumulators (e.g. RCL+6 add memory 6 to reg. X)

f 1..9 could set the FIX

f 0 could reset FIX

scientific format starts automatically when number doesn't fit the 10/12 digits (if you need to input 1E-25 probably you need a scientific calculator...)

the greek pi is a gift (one could calculate circle areas, right?)

well, that's all. Any need of a manual, for such a calculator?

-- Antonio

P.S. I invite you to find some substitutes for % and pi, since they are really the lesser useful keys of this proposal.

I also invite you to riconsider this: if the available memories were, say, 0..6, there would be three shifted function to add. What do you think could be added to enhance this calculator, without losing the point it must be a **simple** calculator?


Forget "f 1..9" being for FIX, the shift function on all those keys can be used for so many things, heck you could turn it into a very usable scientific.
But if it had to stay "non-scientific" I'd add EXP, +ENG, -ENG, X^2, R->P, P->R, and simple base conversion. Would it be asking too much to get a Log key as well?
Oh, and my favorite key that should be on every calc, a Parallel key. Only my Casio FX-61F has that.

Software and labels are cheap!

Dave.

What is a Parallel key ? Is this 1/(1/X + 1/Y) ?

I feel the stack should be 4 levels deep, not variable size (yuck RPL !).

Quote:
What is a Parallel key ? Is this 1/(1/X + 1/Y) ?


Yep, that's it. Incredibly handy for electronics guys like me.
I only know one calculator that ever had it.

Dave.

Hello!

Quote:
What do you think could be added to enhance this calculator, without losing the point it must be a **simple** calculator?

Before talking about adding something, I think there is one item that can easily be removed: The "CHS"-key (or better key label). This can be replaced by pressing the decimal point twice as on some old "Interton" calculators! (But I only found that out by reading the manual, so for a real "simple" calculator it might not be the right thing to do).

To add, I would consider some user-defineable constant-keys (say f-"C1", f-"C2" and f-"C3") that multiply the contents of the x-register with a constant without pressing any further key: Very useful for currency and units conversions!

Saluti, Max

And something else too: A ":" key that allows you to enter time values as hh:mm[:ss] for simple time calculations (working hours spent on a project, or flying time for the not-so-small (see below :-) ) fraction of aviators amongst hp calculator enthusiasts). I have a little four-banger Casio calculator that can do exactly that and that probably is the calculator I need most often...

And to repeat myself I still think that a 35years anniversary hp calculator must have LEDs of some kind in its display. Be it a dot-matrix OLED display or a backlit (by red LEDs) inverse LCD.


Edited: 22 May 2007, 9:36 a.m.

I think you missed the point Chuck established. The calculator must be simple.

X^2 can be calculated by

<value>
ENTER
x

and the Parallel key is something too far for simple calculators. The same for LOG and +/- ENG. For those you need a scientific calculator.

for EXP: yes, it may be useful, but then something to manage the SCI format should be added. Since I said that scientific format should activate automatically, well, I think it's enough.

for the CHS: some guy in an older message suggested its presence, and I agree. I find it useful, like the 1/x key and the fix format (since calculating money or centimeters decimals without having to round by mind is useful).

for time calculation: use decimal format (10:30 as 10.5 and so on)


The feeling of such a calculator should be: keep it simple. Chuck proposed an even simpler calculator, probably not that usable, and the effort of mine reached a (of course as a 'scherzo', if you know the musical notation strings) more usable an not that much complicated calculator.

I guess as HP should do.

-- Antonio

Hello!

Quote:
for time calculation: use decimal format (10:30 as 10.5 and so on)

If the world was so easy that there are only half hours, then we would need no calculator at all :-)
HP itself has shown long ago that a time function can be very useful indeed: With the hp-01 watch that could perform such calculations...

Saluti, Max

BTW: This is my most useful little Casio time calculator (nearly as difficult to find than the hp-01!):

Ok, Max,

you won.

I thought over your proposal a lot, and probably the time function could be a very interesting add-on even for this tiny calculator.

So let's figure that HH.MMSS (or ->HMS à la HP) could be put in the place of greek pi: after all, digiting 3.1415 is not that difficult, and if somebody doesn't know the pi value at the fourth decimal, probably he doesn't even need it.

Now, let's figure out the help page:

digit the hour in decimals (so for instance 10:25:56 becomes 10.2556) and hit f ->HMS 
(the display shows the decimal version of the hour, that is 10,1520160000).
Digit another hour in the same way and then operate a + or a minus....

and now? There would need another conversion key like ->HR on the result to restore the hour format. How does your Casio work? I don't see any other time key. We could think to use ->HR in place of the %. This could go.

What about it?

-- Antonio

Y'know, if they put together a desktop model (something like the HP-97) with LEDs, they could sell that like one of those retro, "Sharper Image" products, and probably make a bundle. It wouldn't even need a printer or magnetic card I/O -- just a big, handsome LED display, with maybe clock & alarm functions to give it a reason to be ON all the time. It would probably be easier nowadays to accurately reproduce the HP-97's tactile feel (springy, adding-machine-like) than that of, say, the HP-15C.

But, of course, they're supposedly going to be celebrating the anniversary of their pocket scientific.

If you introduce ATAN and SIN as replacements for % and PI you get functional trigonometric capabilities:

  PI being 4 ATAN(1)

COS(x) being SIN(PI/2 - x)
TAN(x) being SIN(x) / COS(x)

ASIN(x) is ATAN(x/SQRT(1-x*x))
ACOS(x) is PI/2 - ASIN(x)


Things have to be in radians so we can get PI back.


- Pauli

Hello!

Quote:
How does your Casio work? I don't see any other time key.

With the Casio, the "HMS"-key is really a mode-toggle: You start your time calculation by pressing "HMS" and from then on, it stays in time mode (until you press "AC") displaying a "-" sign as hour separator and " and ' to separate minutes and seconds (as in the picture above).

To enter values, you digit HHH (yes it can do up to 999 hours, other than some spreadsheets that wont go beyond 24 which is quite useless when you sum-up working hours!), then "HMS" then MM then "HMS" then SS. Of course, minutes and seconds are optional. And it is smart enough to know that when you dont press "HMS" first thing after "AC", you mean "%" instead :-)

Greetings, Max

Nice calculator. I'd switch STO and RCL (that is, have STO be the shifted function and RCL the unshifted one), since RCL will be used at least as much as STO, and often more.

Aah! So it toggles its mode.

Well, I believe this is too much Casio-oriented. HP ->HR and ->HMS are more common on a (even hypothetical) HP calculator.

Don't you agree?

-- Antonio

You can remove the decimal point key as old Sinclair machines used to do. They only had an EE (enter exponent) key.

Also, there was one shift key only but pressing it twice would give a second shift state (IIRC).

But simplicity is probably better than saving one or two keys.

Ok. This is my last effort.

I considered (hopefully) all the requests of this thread, some have been discarded and some kept (for what about the STO/RCL inversion, well, it can be done).

Here's my final result.

It could a very handy calculator for any common calculation, even professional (if you know your formulae and don't need solving capabilities).

Apart from the display (LED or diods), I remark that if HP took this proposal into consideration (which is *always* an idea of Chuck), the keys should be the ones of the old Voyager of Pioneer series, not the ugly keys of the current HP-9S/9G or HP-33S models. They shouldn't look too back in time: the keys of the HP-12C 25th anniversary I got are the best keys of all the voyager models I own.

Saluti.

-- Antonio

... and let another supercalculator be conceived at the 40th anniversary!

Edited: 23 May 2007, 9:18 a.m.

Right! I agree on all the lines!

-- Antonio

Half of the point of an RPN calc, though, in my opinion... is to have the multi-level stack, so you can queue up numbers.

Now, your "average" "4"-banger has the following functions:

Add/Subtract/Multiply/Divide

Memory (can be implemented the same way as the scientifics, with an S level)

Square Root

Percent

That's what I think an RPN "4"-banger should have, to be competitive. Now, of course, as has been pointed out... why not make it a modern day HP-35?

Quote:
Half of the point of an RPN calc, though, in my opinion... is to have the multi-level stack, so you can queue up numbers.

Now, your "average" "4"-banger has the following functions:

Add/Subtract/Multiply/Divide

Memory (can be implemented the same way as the scientifics, with an S level)

Square Root

Percent

That's what I think an RPN "4"-banger should have, to be competitive. Now, of course, as has been pointed out... why not make it a modern day HP-35?


Exactly. What's the point in making something so deliberately "simple" that's it almost no good for anything. The form factor is its draw card, why not add some real scientific functionality to it?
Then it goes from being useful only for adding up your grocery bill, to something which could do 99% of everyday engineering and scientific tasks. All for the sake of adding a few labels to the keys and some extra lines of code.

Dave.

Quote:
The form factor is its draw card, why not add some real scientific functionality to it? Then it goes from being useful only for adding up your grocery bill, to something which could do 99% of everyday engineering and scientific tasks.

Now you are next to OpenRPN, aren't you? Or what shall be the difference? Waiting for enlightenment ...

This has been an exercise in minimalist design. I doubt anyone here expects something like this to actually be built. But it was interesting to start from something radically simple, and then add only those features people deemed essential in a basic "four-banger" calculator. I really would buy something like what we came up with - if only for the novelty. But that doesn't mean it would be a viable product.

Regards,
Howard

Quote:

Now you are next to OpenRPN, aren't you? Or what shall be the difference? Waiting for enlightenment ...


I don't know too much about OpenRPN, so correct me if I'm wrong, but I get the impression OpenRPN is a "do everything" project?
All I want is *small* and *basic* scientific calculator. By small, I mean smaller than a Voyager.

The proposed "4 banger" seems a perfect size for me, but deliberately crippling it with only 4 functions seems incredibly silly, even as a "what if" argument. Buy hey, if everyone wants to play a theoretical "4 banger" game then by all means continue!

Perhaps it could have a swappable face plate and a switch on the back for "4 banger" or "Scientific"?

Dave.

Quote:
This has been an exercise in minimalist design. I doubt anyone here expects something like this to actually be built. But it was interesting to start from something radically simple, and then add only those features people deemed essential in a basic "four-banger" calculator. I really would buy something like what we came up with - if only for the novelty. But that doesn't mean it would be a viable product.

Regards,
Howard


Ok, if you want to talk minimalist design, then there are two keys too many, minimum. The STO and X-Y keys are redundant when you have a function key and all those normal keys. Put all the alternate functions as shift functions and you have saved those two keys.

Then again if you want to get radically minimalist, in theory you could do it all with just 2 keys total. Would suck to use, but that would be minimalist. In fact you could probably do it with just one key, but that would suck an order of magnitude more :->

A minimalist RPN calc would also only have a two level stack, not four IMHO. But of course 4 is much better, and would better deserve the "RPN" title.

Of course, a 4 function RPN calc would never sell in todays market, ever. But that exact same calc with scientific functions *would* most likely sell.

I can design a real one in a few days, if someone can somehow make the buttons and housing!

Dave.

I have always thought if HP could make a basic RPN calc, with tactile keys, and include the basic scientific functions of say, a $9 TI-30, with a few TVM functions built-in, you would really have something. I would easily pay twice that for an HP with those features. :-)

I don't need hundreds of functions to make it a great daily user.

[edited in light of subsequent comment by DaveJ and my response to same]

If we're talking improved 4-bangers, how about this as an attempt.
Think the bottom five rows of a spice model with a double width ENTER key on a row of slightly narrower keys.

My non-graphical layout is thus:

     ENTER  CHS X<>Y  CLx
7 8 9 -
4 5 6 +
1 2 3 *
0 . shift /

and the shifted functions to the same template:

     LASTx  EEX  Rv   CLreg
1/x sqrt y^x HMS+
% %chg %T HMS-
STO LN EXP ->HMS
RCL pi shift ->H

Shift-2 and shift-3 aren't defined. Any good suggestions?
Likewise, pi isn't a heap of use and could easily be replaced by something else.

I'd assume that STO/RCL include arithmetic and that there are at least 10 registers: 0 .. 9 & maybe .0 .. .9, although the latter is overkill.

Thoughts anyone?

- Pauli


Edited: 25 May 2007, 5:56 a.m. after one or more responses were posted

Quote:
I have always thought if HP could make a basic RPN calc, with tactile keys, and include the basic scientific functions of say, a $9 TI-30, with a few TVM functions built-in, you would really have something. I would easily pay twice that for an HP with those features. :-)

I don't need hundreds of functions to make it a great daily user.


It is the glaringly obvious model missing from their line-up.
They have cheap-n-cheerful algebraics for the school market, they have the 33S as a high end programmable, and the powerful graphic machines, but no normal RPN scientific.

Incidently, when was their last basic scientific model produced?

Interestingly, in the real world of engineering the only calcs I ever see are basic scientifics, no engineer I know at any company I've ever worked for uses a graphic or other high end programmable on a daily basis. If they do have them I never see them on the desks.

Dave.

Quote:
If we're talking improved 4-bangers, how about this as an attempt.
Think the bottom five rows of a spice model with a double width ENTER key on a row of slightly narrower keys.

My non-graphical layout is thus:

     ENTER  CHS X<>Y  CLx
- 7 8 9
+ 4 5 6
* 1 2 3
/ 0 . shift


Thoughts anyone?

- Pauli


I'd have the ENTER key on the bottom right, and the +-*/ keys on the right column.

As a right hander it just feels more natural to me to bring my hand back right to perform the last key operation (+/-*/). With the current arrangement I'd have to move my hand left to push the final button and then bring it all the way back to the naturally resting right side.

I go to the gym to do my lateral exercises, so have no need to perform extra here :->

Dave.

Quote:


Ok, if you want to talk minimalist design, then there are two keys too many, minimum. The STO and X-Y keys are redundant when you have a function key and all those normal keys. Put all the alternate functions as shift functions and you have saved those two keys.


Fair enough.

Quote:
Then again if you want to get radically minimalist, in theory you could do it all with just 2 keys total. Would suck to use, but that would be minimalist. In fact you could probably do it with just one key, but that would suck an order of magnitude more :->

That ignores the starting point we were given, and is absurd anyway. (Your smiley indicates agreement with that last judgment, I believe. 8)

Quote:
Of course, a 4 function RPN calc would never sell in todays market, ever. But that exact same calc with scientific functions *would* most likely sell.

I would buy one each of both. I'm a somewhat narrow demographic, however. 8)

Quote:
I can design a real one in a few days, if someone can somehow make the buttons and housing!

That exact hardware seems to be a bugaboo for every hobbyist and low budget professional project I know about. It's possible to do PCBs relatively cheaply, but housing it all in something that doesn't feel flimsy, and with keys that don't make a mockery of the expectations of the average consumer, let alone HP vintage calculator fans apparently takes too much money. (Talk to Eric Smith for a more informed opinion about this than I can offer.)

Regards,
Howard

Speaking of Eric, cases and keys, I mentioned here sometime back that I thought an HP-97 shell, display and keys with new "guts" would be a killer hobbyist project. (I happen to have two spare 97s with hardware problems unrelated to the components in question.) Eric, you have your DIY calculator board. Have you given any thought to housing it inside classic HP calculator shells like the 97? What challenges would you have to overcome to interface the keyboard and display?

Regards,
Howard

Quote:
That exact hardware seems to be a bugaboo for every hobbyist and low budget professional project I know about. It's possible to do PCBs relatively cheaply, but housing it all in something that doesn't feel flimsy, and with keys that don't make a mockery of the expectations of the average consumer, let alone HP vintage calculator fans apparently takes too much money.

So does that mean OpenRPN will never materialise in hardware form?

Dave.

No, "do everything" is for the big companies to play with. Everything from the 49g on has been grudgingly accepted at best by engineers and scientists at best. To satisfy virtually everyone in this community requires two basic forms of calculators: voyagers and pioneers. Generally speaking, people around here prefer one of the two layouts. Update the electronics to modern levels, keep the source code in the public domain, over-engineer everything, make outstanding documentation, and package the whole thing for $100 or less.

No objection to shifting the digits to the left and moving the arithmetic operators to the right side. I don't like the idea of moving the enter key to the bottom, looks ugly :-) Also makes the keyboard unlike anything I've ever seen...

For the two missing operations, LN and E^x seem fairly natural. We'd then probably want to ditch the R^ and replace it with EEX. The E^x and y^x we're going to want exponents.

- Pauli

From my understanding, the OpenRPN project really stalled over a lack of a hardware engineer to design the electronics. The software was/is coming along nicely (I did commit a fair chunk of source code so I am biassed) and the physical design looked quite good. Some of the manufacturing issues had been investigated but I don't have any idea to what extent etc.


- Pauli

I'm a right-hander, too, but I like the arithmetic keys on the left, along with ENTER. I like to look at the keyboard as I am punching away, and this arrangement lets me see what I am doing.

And, appropriately, since this is the "Anniversary Edition," it then also mimics the original '35 key layout!

Dave,

I have had quite the opposite experience. Most nearly every engineer that has a calc on his desk at my work appears to have an HP graphing model. The 48g/x dominates with some 49g and also a gold 49g+ (must've been from the good batch). There's even (cringe) a few Ti-89s'.

Myself, I find myself using my 33s alot, since it is small and I carry it to the shop alot. But, I also keep the 48 on my desk (because I turn out needing it only when I leave it at home).

/ECL

Paul,

Quote:
My non-graphical layout is thus:

     ENTER  CHS X<>Y  CLx
7 8 9 -
4 5 6 +
1 2 3 *
0 . shift /

and the shifted functions to the same template:

     LASTx  EEX  Rv   CLreg
1/x sqrt y^x HMS+
% %chg %T HMS-
STO LN EXP ->HMS
RCL pi shift ->H

Some proposals:

  1. Rearrange the arithmetic operations like in Pioneers (/ * - + from top to bottom). Reason: more logical sequence.
  2. Swap the x<>y and the CHS key. Reason: stack operations are next to each other.
  3. Use +/- instead of CHS. Reason: better worldwide understanding.
  4. Put HMS+ on + and HMS- on -. Reason: more logical.
  5. Put y^x on * and 1/x on /. Reasons: more logical and 1/x next to y^x for roots.
  6. ...
Became more than I first thought. For sake of clarity, I propose to arrange the shifted functions like this:

        LASTx  Rv  EEX CLreg
% %chg %T 1/x
sqrt LN e^x y^x
STO >HR >HMS HMS-
RCL pi shift HMS+

BTW, I do not see the need for %. IMHO, for scientific people, trig functions are far more useful than % (=E-2). So an alternative may look like this:

        LASTx  Rv  EEX CLreg
SIN COS TAN 1/x
ARC LN e^x y^x
STO >HR >HMS HMS-
RCL sqrt shift HMS+

It will even revive the ARC operation as in HP35 d:-) Of course you will reach ARC SIN with 1 shift only: Shift 4 7

Best regards, Walter

(Edited for error correction and to add the unshifted pattern proposal for sake of completeness:

     ENTER  X<>Y  +/-  CLx
7 8 9 /
4 5 6 *
1 2 3 -
0 . shift +
)


Edited: 26 May 2007, 5:15 p.m.

Quote:
Dave,

I have had quite the opposite experience. Most nearly every engineer that has a calc on his desk at my work appears to have an HP graphing model. The 48g/x dominates with some 49g and also a gold 49g+ (must've been from the good batch). There's even (cringe) a few Ti-89s'.

Myself, I find myself using my 33s alot, since it is small and I carry it to the shop alot. But, I also keep the 48 on my desk (because I turn out needing it only when I leave it at home).

/ECL


Interesting...
What type of engineers are you talking about?
I'm an electronics design engineer in Australia and have worked at primarily electronics (including defense) engineering companies with a large amount of mechanical, acoustics and software engineers. It's the same across all 4 disciplines.

I've heard a lot of guys talk about having owned a graphing calc when they were at Uni, but they don't use it any more as it's simply not needed for everyday calcs. Casio, Sharp and TI algebraics get about 90% of the desks, with the rest being oddball units like pocket computers, 4 bangers, or the occasional old HP.

I've had a similar experience while visiting parent companies in Singapore, UK, and Germany.

Age of the owner doesn't seem to be a major factor, although the old HP's and such are owned by the older crowd.

Dave.

Dave,

I'm an aerostructures guy, although we run the gamut here. There's mechanical designers, analysts, aerodynamicists, electrical too.

I did forget to mention that there's a daily use 15c here too.

The 48/49/50 series' pack a lot of capability, however I don't think that is to their discount. The problem as I see it is that they aren't as powerful as the PC. Memory, and speed included.

If you could carry your PC's functionality around in a 50g case, and USB it to a monitor when you get to your desk, would you turn it down? I just don't see much point to becoming a power user of the 50g, when I can't import/export files from my analysis package at work (and Excel, etc.)

I know this is a bit beyond the friendly chat we were having. Just some thoughts.

Quote:
Dave,

I'm an aerostructures guy, although we run the gamut here. There's mechanical designers, analysts, aerodynamicists, electrical too.


Presumably in the US?
Beats me why there is such a big difference!?
Perhaps your fields are a bit more theoretical that need custom programs to model stuff?

Quote:
I did forget to mention that there's a daily use 15c here too.

The 48/49/50 series' pack a lot of capability, however I don't think that is to their discount. The problem as I see it is that they aren't as powerful as the PC. Memory, and speed included.

If you could carry your PC's functionality around in a 50g case, and USB it to a monitor when you get to your desk, would you turn it down?


I believe that would be called a PDA?, and yes, I would turn it down as I don't need it. Frankly I have never really needed anything more than a basic non-programmable scientific calc for almost all of my work, and that's over 20 years in electronics design.

When I have had to do something a bit more complex, excel, Matlab, or other PC tools like a custom BASIC/C program get used as I find they are a much better interface.

The mechanical engineers seem to do everything on their cad packages, with FEA modelling and all that stuff.

Acoustics guys use Matlab all the time.

Production engineers use Excel.

But on the desks it's always the same when I look (and I do take note), I see almost nothing but non-programmable scientific calcs used by all these disciplines.

Quote:
I just don't see much point to becoming a power user of the 50g, when I can't import/export files from my analysis package at work (and Excel, etc.)

I know this is a bit beyond the friendly chat we were having. Just some thoughts.


Amazing difference isn't it?
Anyone else care to share their experiences?

Dave.

I disagree. The electronics is pretty easy, really. Even without an EE it's easy to buy a development board and get started on the programming.

OpenRPN's biggest challenge was the case and keyboard, and I don't believe that a solution had been found.

No, the case and keyboard are not a big deal. Rather than a development board, I would like to have developers working with at least prototype hardware.

-Hugh

Quote:
I disagree. The electronics is pretty easy, really. Even without an EE it's easy to buy a development board and get started on the programming.

Getting started without hardware isn't problem, making a prototype is.

I have *fix running on one of our SnapGear ARM based devices, it works just fine. That is close enough until real prototype hardware is available. My day job involves writing and porting software to embedded devices, I'm not in the slightest bit worried about this phase :-)


- Pauli

As usual Walter is way ahead of me in keyboard layouts. I like all the changes.

I too had thought to drop the % functions. I thought of replacing them with gamma/x!, permutations and combinations.

I did think about including the trigonometric functions but then it kind of escallated into wanting degrees and radians, conversions of same, PI etc :-(

So the modified shifted layout excluing trig functions and including the ones I mentioned above is:

        LASTx  Rv  EEX CLreg
x! Pxy Cxy 1/x
sqrt LN e^x y^x
STO >HR >HMS HMS-
RCL PI shift HMS+


- Pauli

Working for the instrument company formerly know as HP, it seems I have a three tier approach to calc use:

I primarily use a HP15C for:

Quick rule of thumb calculations - during initial design and discussions.

Simple arithmetic when creating new layout geometries.

Any from first principles paper work.

I still occasionally dig out my HP49 for:

Conversions : dBm <-> Vpp , RL<->VSWR, dB+ etc.

Pad design: T and PI pads, min-loss pads. T-couplers, Y-Delta etc.

Using legacy code I wrote on a HP48SX in the early '90s.

Anything that more involved, and it's the PC based design tools that get used:

ADS - for circuit and system simulation

Eagleware - for filters

MATLAB for DSP

Analogs PLL tool.

Lots of C++ for instrument control

R - Sweave - Latex for result plotting and reporting ( avoiding Excel and Word as much as possible, except when dealing with managers and production).

Lately, I've resurrected some 2-port network analysis code similar to LNAP on the HP41 and have a HP49 version and soon a HP15 version. This is to quickly see the response of LC filters I find in others schematics ( in less time than firing up ADS ).

I guess my ideal calculator for today's use would have a HP15 style complex stack and basic alpha features from the 41 ( bringing back the 41's extra segment LCD) . Adding basic text prompting and a XEQ key for alpha labeled functions - all in a small voyager styled case.

I should add there seems to be only a few engineers still using HP calculators.
Of those who got a HP49 from a batch a few years ago, I can think of only one other who made the effort to program it

Those who did not use RPN HPs in the 80's/90's have never gotten into the habit at all, so tend to use basic TI calcs.

Currently, our IT procurement system does not even have a standard way to buy calculators.

Edited: 27 May 2007, 4:39 a.m.

Quote:
I should add there seems to be only a few engineers still using HP calculators.
Of those who got a HP49 from a batch a few years ago, I can think of only one other who made the effort to program it

Those who did not use RPN HPs in the 80's/90's have never gotten into the habit at all, so tend to use basic TI calcs.


I think you've hit the nail on the head there, most people just couldn't be bothered with programing a calculator, and it snow-balls from there:
If you don't program it all the time then you forget how to do it and then it's just too much trouble. Next thing you know you have this wizz-bang calculator which is more optimised for programming and other complex graphics stuff than for the basic calcs you do 99% of the time. Next thing you know you toss it in the draw and replace it with a $20 non-programmable scientific for everyday use.

I suspect that's why I virtually never see programmable or graphics calcs on engineers desks, yet many will say they have owned one at one stage.

I find I do the same thing myself. I have a HP28S which I really like, but it's just useless for everyday use, so I've gone through several sets of batteries just having it sitting in the draw gathering dust.

Dave.

OK.

The one major exception to your rule is field surveyors. We have always bought and used RPN (or RPL), and HP in particular. Something on the order of 99% of us have and use an hp programmable daily. The majority use canned programs and most of the rest, like me, just write additions to or throw out unused portions of commercially available packages. Still; the furthest you will usually see a surveyor go from the basic 41/42/48/49/50/32/33 is a 200lx or one of those TDS Rangers, and they both have an onboard RPN calculator too.

One thing we look for in a new calc is a logical progression of use and function, so we don't have to start learning from scratch - and just "toss it in the drawer and replace it with a $20 non-programmable scientific for everyday use". About 20 years ago Ted Kerber gave my union's apprenticeship a bunch of his HP41 programs - when he was still selling them commercially. I've used his D'zign software on my 41, 42 & 48 and i see others using the software on the 33 & 49/50. The program runs similar enough across the different platforms so the old ones don't always end up "gathering dust" either. Well, not figuratively. In this work; everything gathers dust.

Hello!

Quote:
Amazing difference isn't it?
Anyone else care to share their experiences?

I have been working as an engineer in Germany and Italy (for 3 years in the early 90ies) during the last 25-or-so years in the fields of satellite design, radar remote sensing, mobile network planning/wawe propagation and CAD/CAM software development. My colleagues were either aerospace engineers like myself, mechanical engineers, electrical engineers, mathematicians, physicists, geographers or IT-people.



I have never ever seen anyone use a graphing calculator (actually, the only grahphing calculators I ever saw are the ones in my collection). Not one. During the last 10 years, I have not seen anybody use a scientific calculator of any brand. Before that, they used mostly cheap Casio or Ti calculators that they had already used in school and university before. I probably have seen less than 5 hp RPN calcualtors beeing used at work!


Greetings, Max

This one is my fav.

Quote:
Here it is...a fantastic 4-function Voyager...


You start with this one, leave it up to this forum and you end up with a 150 keys calc with f, g h, i and j shift keys.

Good luck !

Edited: 27 May 2007, 2:52 p.m.

In my last office, with 17 engineers, we had quite a few HP:

Two used 48 series (onea 48s, the other a 48gx)
three used 11c and 12c
four used the 20s algebraic
two used the HP41
one used the 17bii
And then there was me: I used every HP!

In my current office, the chief engineer uses a 32sii, and one of the project managers (30 years old) uses a 48Gx and the top finance guy uses a 12c. There are about 10 engineers.

Edited: 27 May 2007, 4:29 p.m.

Quote:
You start with this one, leave it up to this forum and you end up with a 150 keys calc with f, g h, i and j shift keys.

You are right. Engineers are dominant in this forum. Thus, the very, very slight trend to "over-engineering" you observed is no real surprise.

Speaking for me personally, I may *perhaps* buy an RPN 4-banger, but I will *most happily* buy an RPN scientific, even a primitive one like one of the above proposals.

Edited: 27 May 2007, 5:35 p.m.

Come on why only five shift keys? That only elucidates 55 functions from the remaining eleven keys. The "optimal" arrangement has eight shift keys producing 64 functions from the remaining 8 keys. For useabilitys sake, I'd consider six shift keys for 60 functions leaving the digits all unshifted.

:-)

- Pauli


Edited: 27 May 2007, 7:21 p.m.

Hahaha, that's why I leave keyboard layouts and menu systems to him on OpenRPN. Walter is downright scary with it.

Quote:
LN and E^x seem fairly natural

Yes, but LOG and 10^x are more common.

... or use multiple shifted function, like f g [key] or f g h i [key]!

-- Antonio

(IT'S A JOKE!)


Edited: 28 May 2007, 5:07 a.m.

Touche!

At least somebody read my message and noticed :-)

- Pauli

I'm thinking of the Commodore N61 (with 61 keys), anyone has another idea ?

Of course BASIC pocket computers had a full Qwerty, but did not reach the N61 level however.

Second prize goes to the desktop with the most keys - but then machines like the 98x0 come into play...

Hello!

Quote:
I'm thinking of the Commodore N61 (with 61 keys), anyone has another idea ?

The Ti Voyage 200 has 80 keys, if my quick count is correct. Since it is programmable in a Basic-like programming language, it may not count according to your definition, however!

Of the 2 (I have one of each :-) ), the N61 is much cooler, of course...


Greetings, Max

... and it's both ergonomic and quite elegant, IMHO.

Best regards from V.