The Museum of HP Calculators

HP Forum Archive 13

[ Return to Index | Top of Index ]

The superb HP-15C (adding to archived post)
Message #1 Posted by Karl Schneider on 15 June 2003, 3:01 a.m.

Valentin Albillo wrote nice short essay in the Forum (3 June, 2003) extolling the HP-15C. I've been meaning to expand on parts of it as a reply, but it's already gone to the archives. Here's a link to his posting:

Voyagers rule ! [LONG]

I bought my 15C in late 1983, and it met my needs through three university-degree programs. Only in the last 15 months did I fill out my HP collection with 12 other models. The 15C is still probably the one I'd take to an exam, if I were allowed to bring only one calculator. It's that capable and useable.

Now, if the HP-11C is amazing, the HP-15C is simply unbelievable, the best pure calculator in the world, hands down. It has everything the HP-11C has, and much, much more.

An excellent HP Journal article about the development of the HP-15C, written by Dr. William Kahan in 1983, is available on the MoHPC CD-ROM/DVD-ROM set, and Dr. Kahan's website. If "pure calculator" implies non-alphanumeric, I'd also agree with the "best in the world" assessment. Otherwise, I'd give the nod to the 42S, while conceding that the 15C has some advantages -- e.g., more-legible display, quick keyboard access to all functions.

Further, everything was so well (thought through) in its design .. You felt there were no arbitrary limitations at all .. All features were integrated and complemented each other perfectly, most served for several apparently unrelated purposes.

Exactly! The 15C is a masterpiece of thorough, meticulous engineering, not only in the construction, but the up-front design of algorithms and functions. The engineers packed an impressive amount of functionality onto a keyboard and display that lacked the expandability of named functions and menus, and yet made it all intuitively logical. The detailed, error-free manual is thicker and heavier than the calculator, but I believe that most users, after initially learning how to program and to use the advanced functions, need not consult the manual (or the supplementary Advanced Functions Handbook) in order to perform tasks using the 15C.

I also agree about the lack of "arbitrary limitations" -- something I see in the direct-successor 32S/Sii and the 20S. The 32Sii has less user RAM, and lacks the matrix functions and complete, integrated complex-domain functionality that were specifically provided in the 15C as a design objective. The 20S provides certain functionality of the 15C as "loadable" programs that require usage-instructions and wipe out the contents of user RAM. Both products seem "stripped down" in that respect.

The 15C, by contrast, is "right-sized" -- it has enough resouces and built-in functionality to tackle any problem you'd reasonably want to try on a handheld calculator. Practically anything it can't handle would be better attempted on a PC with the appropriate software. Of course, the 41CV/CX and 42S are better for extensive programming than the 15C, which wasn't really indended for those purposes.

The dual-purpose keys of the 15C were ingeniously intuitive. For example:
-- nPk and nCk keys [statistical permutation and combination; or, "partitioned" and "complex" matrices]
-- I and (i) keys [indirect addressing of registers/labels; or, imaginary (complex) numbers]
-- 1/x [reciprocal or matrix-inversion]

There's never been anything like the HP-15C, there never was before, and there never will be in the future...

Alas, probably so. Modern high-end calculators compute faster and have more fancy capabilities, but I'd doubt that any are as thoroughly well-executed as the 15C. It's more than a still-useful tool; it belongs in a "Hall Of Fame" of consumer electronics.

Re: The superb HP-15C (adding to archived post)
Message #2 Posted by David Ramsey on 15 June 2003, 12:22 p.m.,
in response to message #1 by Karl Schneider

The orthagonality of the functions in HP calculators was something I always admired. This carried through to their other products. For example, the HP-01 calculator watch is the only calculator watch I know of that actually integrated the time and calculation functions: you could, for example, get a real-time readout of the cost of a long distance phone call after inputting the per-minute cost.

In the HP palmtops (1xx, 2xx), the built-in Lotus 1-2-3 spreadsheet's cells could link to an external routine written in the "Solver", so that the contents of a cell were determined by algorithms you implemented.

I recently started seriously playing with the 82161A digital cassette drive. As I expected, it can easily load and save programs. But it can also load and save a range of registers, as well as a custom keyboard layout on the 41.

They really don't make them like this any more.

We can only hope that the rumored successor to the 32SII isn't as "improved" as the HP-12CP is.

Re: The superb HP-15C (adding to archived post)
Message #3 Posted by David Smith on 15 June 2003, 4:09 p.m.,
in response to message #2 by David Ramsey

Too bad the HP-01 does not have trig functions... you could then take the sine of the times.

Ouch! arrgh! Gag!
Message #4 Posted by Norm on 17 June 2003, 4:33 a.m.,
in response to message #3 by David Smith

The sine of the times ! Ouch! arrgh! Gag!

Well, with trigonometry on such a watch, you'd take it with you whenever you COSINE a note at the bank !! har har har

That is quite amazing, to have it do things related to math on a time-basis.

SOON I WISH TO BRING UP CALCULATOR WATCHES. There is a unit being made presently that can be acquired on eBay. For $40 it can make U every happy. I wear mine daily. It is called Casio CA-602 . No, it is not RPN but other than that its a very sensible little package !! :o)

The seller I have bought from is "IVSHC", and you can trust him. He ships from Hong Kong. Another guy was also reliable, "loose-logic" from Japan, BUT I think he ran out.

Re: Ouch! arrgh! Gag!
Message #5 Posted by Michael F. Coyle on 17 June 2003, 11:48 a.m.,
in response to message #4 by Norm

Hi Norm,

Try not to let this thread go off on a tangent.

As for the calculator watches, secant you shall find.

- Michael

gasp! gak! blak! (thud)
Message #6 Posted by Norm on 17 June 2003, 1:50 p.m.,
in response to message #5 by Michael F. Coyle

fell over dead of a heart attack on that last one !!!

Re: The superb HP-15C (adding to archived post)
Message #7 Posted by Euclid on 24 June 2003, 3:28 p.m.,
in response to message #3 by David Smith

Time, RPN, now I get it... revolutions per.. uhhhh nanosecond!

the 9g fights back (a bit)
Message #8 Posted by hugh steers on 24 June 2003, 4:19 p.m.,
in response to message #7 by Euclid

for amusement, i decided to see if i could make the 9g do something useful. no really. and to this end; is it possible to make up for the woeful lack of numerical integration by programming it to do so.

in order to make its ability comparable to the 15c, i wanted a romberg method and for added piquancy, replicate the sampling redistribution performed which avoids evaluation at endpoints and problems arising from fixed width sampling. ideally to beat the 15c.

but does the 9g have any trousers?

originally, i wanted to supply solve and integrate subroutines. however, the 9g has so little memory, only 400 bytes for basic (can you believe it?), that it is impossible to write anything very useful. you can fit in a secant solver and simpsons rule and still have memory for your own function, but not a lot more. i have a very nice ridders solver, but it wont fit with romberg at the same time.

you can see the finished romberg method below. sampling redistribution is incorporated, so expect similar limitations to the 15c. problems mostly arise from infinite limits because the method requires limit scaling to unity then multiplies by b-a (the interval) and b-a is large, consider: int(x*exp(-x),0,inf)==1. 9g gives 0.99999988 after 2.5 mins using 228 ~ inf. see the 15c advanced manual for this example.

all in, it ought to flatten the 15c, but its so slow that it only just beats it. im seeing something like 2 evals/sec, which is more than the 15c but then the 15c is 20 years old and has the algorithm in rom.

more cpu and memory please.

INPUT A,B,E H=2;D=0;M=0;J=1;B=B-A FOR (K=0;K<9;++K) { L=D;F=H/2-1;C=0 FOR (I=0;I<J;++I) { D=1-FF;X=((F+DF/2)*B+B)/2+A GOSUB PROG 9;C=C+DY;F=F+H;} F=4;D=M;M=(M+HC)/2 FOR (I=0;I<=K;++I) { C=N[I];N[I]=(F*M[I]-D)/(F-1);D=C;F=4F } D=M[I] IF (ABS(D-L) < 16E*ABS(D)) THEN { GOTO 1; } J=2J;H=H/2 } Lbl 1: S=3DB/4;PRINT S END

write function in prog9, for example:

Y=Xexp(-X) END

all the variables are used except for Z :-)

[ Return to Index | Top of Index ]

Go back to the main exhibit hall