The Museum of HP Calculators

HP Forum Archive 14

[ Return to Index | Top of Index ]

Bring Back The HP 15C
Message #1 Posted by Chris Woodhouse on 10 May 2004, 11:29 p.m.

I have my new web site up with a petition to bring back the HP 15C I give some of my thoughts on why I want HP to bring back the 15C on the site, so please check it out and if you think the 15C is worth an other go around sign the petition.

http://hp15c.org:8080

Note that if you are using IE you MUST include the http:// part or it will get confused by the :8080 part.

Chris W

      
Re: Bring Back The HP 15C
Message #2 Posted by Bill P on 11 May 2004, 1:26 a.m.,
in response to message #1 by Chris Woodhouse

I'd buy three if they cost the same each as a new 12c.

      
Re: Bring Back The HP 15C
Message #3 Posted by Karl Schneider on 11 May 2004, 1:46 a.m.,
in response to message #1 by Chris Woodhouse

Chris --

I agree with the cause, and would sign your petition after you post on your site how you plan to use the information the signatories provide (privacy, security, etc.) I would hope that it will go directly to HP when enough signatures are obtained.

Personally, I would like to see a re-issued 15C called a "15S" with a few very minor upgrades. Unfortunately, some functions would have to be sacrificed to allow AOS, and Eric Smith might be the only one who has the 15C ROM code. (There was a recent thread about the preceding.)

I've already asked NCEES via detailed correspondence to approve the 11C, 15C, and 20S for the PE/FE exams. If I can't convince them to take action, I'm considering an on-line petition (perhaps hosted on MoHPC, if Dave is willing) to help sway them.

-- Karl S.

            
And the 28S also, please!
Message #4 Posted by J. C. Boco on 11 May 2004, 7:48 a.m.,
in response to message #3 by Karl Schneider

Bring this sweet baby back into production, and also clear it for the tests.

            
Re: Bring Back The HP 15C
Message #5 Posted by hugh steers on 11 May 2004, 7:49 a.m.,
in response to message #3 by Karl Schneider

i agree. i would very much like a new 15c that was faster. that was really its only limitation.

            
Re: Bring Back The HP 15C
Message #6 Posted by Charles Oxford on 11 May 2004, 9:54 a.m.,
in response to message #3 by Karl Schneider

The NCEES only bans calcs with text editing or communicating capabilities. The 15c has neither and is therefore allowed.

                  
Re: Bring Back The HP 15C (NCEES Allowance)
Message #7 Posted by Jeff on 12 May 2004, 7:30 a.m.,
in response to message #6 by Charles Oxford

It is the stated intention of the NCEES to produce a list of approved calculators, rather than a list of banned calculators. Only those calculators on the approved list will be allowed in the tests. (See page 8 of this issue of the Licensure Exchange publication of the NCEES.) To keep things simple, they will likely not research every calculator ever made to determine if it meets their requirements. They will also not want to have a list of hundreds of calculators that would meet their requirements that they would then have to check each and every test taker's calculator(s) against. It may be possible to petition the NCEES to put the 15C on their "approved" list, but it should not be assumed that they will automatically allow it.

Edited: 12 May 2004, 7:43 a.m.

                        
Re: Bring Back The HP 15C (NCEES Allowance)
Message #8 Posted by Fubar Vikinghelmet on 13 May 2004, 1:44 p.m.,
in response to message #7 by Jeff

This is a sore issue with me...

I just downloaded and read the NCEES December 2003 document that's linked in your post. Here's the money quote (page 9, "Make proctors’ lives easier"):

"...On exam day, we had to confiscate 10 to 12 calculators, which gave us the opportunity to get acquainted with these models."

Doggone it, why didn't the NCEES people simply spend a weekend geeking-up on this and other websites if they wanted to "get acquainted" with whatever calculators? The knee-jerk response is "they don't have the time to..." and my multiple fist-in-the-face response is a) you certainly had time to think up the ban, b) you certainly had time to think up what constituted an unacceptable calculator, why not expand on that?, c) you came up with a rule that disqualified calclators whose ports were covered; did you even think about what those ports could do in a big room full of people?, d) you folks are engineers, right? that would imply more than a passing interest in calculating devices, eh?, e) etc.

I've had the "why don't you, Fubar, make up a list/document/screed of calculators and their characteristics and present it to the NCEES. And the fist-in-the-mouth response is: that is not my job - the NCEES people are either paid or have committed themselves to such issues; they should have been on top of this since day one and established protocols for calculator evolution vs. the tests.They're also unreasonable; who the heck am I in their pecking order? (a customer?! ha! we don't work that way!) After all, there is an experience element factored into one's being accepted to take the test in the first place! So if someone worked for four years using, say, an HP-42S, sorry, gotta change it for this test.

The bottom line is that the NCEES people are obeying the first rule of personnel management, i.e., punish everyone.

But, I've got my HP-33S and am happy, so the issue is moot - but the NCEES ought to be sacked nonetheless.

            
Re: Bring Back The HP 15C
Message #9 Posted by Chris Woodhouse on 11 May 2004, 10:04 a.m.,
in response to message #3 by Karl Schneider

I'm sorry I forgot to address privacy concerns on the web site. I will add some information to that effect shortly. I can tell you that all I plan on doing with this information is giving it to HP for the sole purpose of convincing them to bring back the 15C. I may decide to send a message to those who provide email addresses and inform them of any progress I make in that regard also.

Chris W

      
2007: 35th anniversary of the HP-35 . . .
Message #10 Posted by Paul Brogger on 11 May 2004, 11:50 a.m.,
in response to message #1 by Chris Woodhouse

. . . and the 25th anniversary of the HP-15C (according to the dates posted in their respective MoHPC articles).

If H-P were interested in calling attention to their heritage of invention, those anniversaries should be commemorated in some appropriate way. (But that's a pretty big "if"!)

            
33S is better
Message #11 Posted by David Ramsey on 11 May 2004, 12:55 p.m.,
in response to message #10 by Paul Brogger

I dunno, guys. I bow to no one in my admiration for the old HP RPN machines.

But the 33s is better than we have any right to expect. It's got a huge, very legible two line display. It's MUCH faster than the 15C. It's got a lot of constants built in, can display with 12 digit precision, and the clever menu system makes great use of the limit-resolution display required in a low-end machine. And after years in the wilderness we finally have an HP calc with a keyboard so good we're reduced to arguing about the nuances of key feel as compared to decades-old machines.

And don't forget algebraic mode. Yes, even having this as an option is a win.

If I were going to change things about the 33s, here's what I'd do:

* Fix the keyboard. Lose the chevron shape and use angled keys like the financial calcs so the mess of tiny legends could be cleaned up a bit.

* Enlarge the display's decimal point

* Make the manual spiral-bound (sigh)

* Restore the [Enter] key to its place of glory.

That's about it, really. I haven't set my two machines side by side and gone over all the functions, so there may be something critically important to you that the 15C does and the 33s does not.

I realize a lot of you really like the 15C form factor. Personally I never did...

                  
Re: 33S is better
Message #12 Posted by bill platt (les Estats Unis d'Amerique) on 11 May 2004, 1:24 p.m.,
in response to message #11 by David Ramsey

Hi David,

Good points, but I will put a different stress on it:

Quote:

(1.) Fix the keyboard. Lose the chevron shape and use angled keys like the financial calcs so the mess of tiny legends could be cleaned up a bit.

(2.) Enlarge the display's decimal point

(3.) Make the manual spiral-bound (sigh)

(4.) Restore the [Enter] key to its place of glory.


(1.) OK

(2.) OK!

(3.) Oh well, I like these too, but, well, oh well!

(4.) No. keep it in the new spot. It really is not so bad--and it matches the 17bii and the "space" on the 48.

I would stress that the more important issues are:

(A.) The Equation List should have been revised to allow editing---not backspacing to get to errors. With all the memory, this would have made a great tool--but who wants to enter a long Equation only to be unable to fix it! And it could have been implemented like the 17bii---so that the variables come up on line one when you use the equation in the solver. Instead, we have the same incomplete thought of the 32sii.

(B.) Solve call Integrate---a major 15C capability that is lost (sadly) in the 32 world...

(C.) Why not a real Algebraic History stack?

(D.) The Algebraic mode is interesting, but it is very poorly documented, and it is confusing to have two different "algebraic" methods on the same machine--the "equation list" method (all infix) and the "ALG" method (infix/postfix). Seems to me a collision rather than a union of logic systems. The 48 series "algebraic object" on a stack was a much more unified way of thinking. This 33s is sort of two unrelated machines in one: The 32sii+ blended with the HkinPo ALGIE.

(E.)(Complex) Matrix support built-in. There is more than enough memory. There is already a menu system. Eliminate the stupid key functions (cube root, x root y, down arrow end, up arrow end) and you have plenty of room! I believe they just don't want to do matrix on a non-graphing---but it would make a more complete system. Of course the documentation would really grow.....

I could be happy without matrices......but I know Valentin would disagree!

(F.) Why was "Horner's method" transcribed to Algebraic using registers?!! sheesh!

(There are more documentation issues but I won't go into them now).

But I do like it---it is growing on me....

regards,

Bill

Edited: 11 May 2004, 1:29 p.m.

                        
Re: 33S is better
Message #13 Posted by Gene on 11 May 2004, 2:27 p.m.,
in response to message #12 by bill platt (les Estats Unis d'Amerique)

The algebraic mode is documented to a degree in the 33S learning modules on HP's site.

Wlodek and I tried to do problems both ways, in algebraic and RPN.

If you haven't taken a look at them, you should go through the lists.

Gene

P.S. I like Algebraic's complex number abilities better than the RPN equivalent on the 33S.

                        
Re: 33S is better
Message #14 Posted by Raymond Del Tondo on 12 May 2004, 2:13 a.m.,
in response to message #12 by bill platt (les Estats Unis d'Amerique)

Response to Bill's answer to (4.): No, move the ENTER key back where it is in the 17BII (and all other serious HP calcs). The 17BII (not the 33S) is one of the best examples for an RPN&ALG machine. It has the right ENTER bar in the right position AND the equal ('=') key in a position where the '=' usually is in alg-only machines, and where it doesn't disturb RPN users. The 17BII (NOT the '+') was, in some respects, the best of both worlds, only lacking TRIG fcns.

When you wrote '...It really is not so bad...' it implies that it isn't that good, and since it's not that good, let's move it back where (nearly) all HP calc fans found it to be perfect;-)

This is not because of the 'good old times', it's just more ergonomic. There were various discussions about this topic in the past, beginning when ACO introduced the 49g, so I think there's no need to repeat it here.

Raymond

                  
Re: 33S is better....
Message #15 Posted by Karl Schneider on 12 May 2004, 2:36 a.m.,
in response to message #11 by David Ramsey

...than the 15C? In certain ways, definitely yes, but overall, I'd say not. In the near future, I plan to illustrate how KinHPo's 33S did not observe the fundamental design principles that the 11/15/32 models exhibited. HP really knew their stuff in the 1980's.

David posted,

Quote:
But the 33s is better than we have any right to expect.

All in all, it's worth having as a "32SII+" that is unconstrained by limited RAM.

Quote:
It's got a huge, very legible two line display.

It would have been a lot more legible if there were more interstital space between digits, if the decimal and comma were larger, if there were weren't so much "shadowing" of the pixels, if the display window weren't reflective, and if the annunciators were larger and not shadowed. None of these issues were problems with the one-line display of the 32SII.

Quote:
It's MUCH faster than the 15C.

Right on there. The 32SII is about 12x as fast as the 15C, and the 33S is faster than the 32SII.

Quote:
It's got a lot of constants built in, can display with 12 digit precision,

Another nice feature, but the 48G* models with large display had many physical constants with their names and units in SI and English.

Quote:
and the clever menu system makes great use of the limit-resolution display required in a low-end machine.

I thought the 32S/32SII menu system was more elegant -- "hit the key that is pointed to", instead of "hit the corresponding number" or "move the cursor and hit enter".

Quote:
And after years in the wilderness we finally have an HP calc with a keyboard so good we're reduced to arguing about the nuances of key feel as compared to decades-old machines.

If I hit the "8" key off-center on the upper-right corner of my new 33S, I can often induce a missed entry, even as the key clicks. Yesterday, I induced 50 consecutive misses! (Can anyone else do that on theirs?) That is just unheard-of on the old HP's.

Quote:
I haven't set my two machines side by side and gone over all the functions, so there may be something critically important to you that the 15C does and the 33s does not.

TheMoHPC website offers a "calculator comparison" feature. The main advantages of 15C functionality are its complete complex-number support, its matrix functions, and the ability to use INTEG inside SOLVE, or vice-versa.

However, the 15C also is tremendously better in clarity of display and keyboard, and of logical grouping of functions.

Quote:
I realize a lot of you really like the 15C form factor. Personally I never did...

The horizontal layout is better if the calc is placed on a desk near the user. If the user must hold the calc for any reason, the vertical layout is better.

-- Karl S.

                        
Re: 33S is better....
Message #16 Posted by Veli-Pekka Nousiainen on 12 May 2004, 2:51 a.m.,
in response to message #15 by Karl Schneider

"If I hit the "8" key off-center on the upper-right corner of my new 33S, I can often induce a missed entry, even as the key clicks. Yesterday, I induced 50 consecutive misses! (Can anyone else do that on theirs?) That is just unheard-of on the old HP's."

This does not happen with my 33s. Your unit is defective (or your fingers are) Sent it back...

[V. <)P. <)N.]

                        
Re: 33S is better....
Message #17 Posted by Arnaud Amiel on 12 May 2004, 6:10 a.m.,
in response to message #15 by Karl Schneider

Quote:
If I hit the "8" key off-center on the upper-right corner of my new 33S, I can often induce a missed entry, even as the key clicks. Yesterday, I induced 50 consecutive misses! (Can anyone else do that on theirs?) That is just unheard-of on the old HP's.

On one of my 45s I can do that as much as I want with the [1/x] key and its original owner told me it was the case when the calc was new as well.

Arnaud

                        
33S Keyboard Misses
Message #18 Posted by Wayne Stephens on 12 May 2004, 8:15 a.m.,
in response to message #15 by Karl Schneider

Nope, I can't induce a "miss" anywhere on the keyboard unless I fail to hit the key altogether (and I guess that doesn't really count as a miss). As near as I can tell, the keys are pretty near perfect.

I really tried to dislike the 33S, with its flashy "non-HP" look and its chevron shaped keyboard, but I just couldn't do it. It just works too well.

My only real complaint (and remember, this is from someone who has just turned 40 and has the commmensurate "failing eyesight") is the tiny decimal point; and in a previous posting (2 months or so ago) I mentioned that the decimal becomes much easier to see if you adjust the display so that it is as light as you can possibly stand. It definitely works (at least for my "old" eyes).

Take care.

Wayne

                        
Re: 33S is better. No it's not :)
Message #19 Posted by Chris Woodhouse on 13 May 2004, 12:18 p.m.,
in response to message #15 by Karl Schneider

Quote:
The horizontal layout is better if the calc is placed on a desk near the user. If the user must hold the calc for any reason, the vertical layout is better.

Vertical is ONLY better if you must hold the calculator with one hand and hit the keys with the other hand. The right way to do it is with a horizontal layout. Hold it with both hands, and press the keys with both thumbs.

Chris W http://hp15c.org:8080

P.S. I hope to move the site to a standard port (drop the :8080) in a few weeks, and I have fixed some problems with posting data with certain extended characters.

                        
Re: 33S is better....
Message #20 Posted by bill platt on 13 May 2004, 2:23 p.m.,
in response to message #15 by Karl Schneider

Hi Karl,

Quote:
The horizontal layout is better if the calc is placed on a desk near the user. If the user must hold the calc for any reason, the vertical layout is better.


I actually find that it is not quite so straightforward. For me, the 10c (voyager) series is actually *easier* to one-hand than many vertical format machines. The reason is the balance point. On the voyager, you place the machine resting on your fingers, and you push the buttons with your thumb. The machine is turned about 20 degrees horizontally, but this does not bother me. I can comfortably reach any key on the board without feeling like I am going to drop it.

On the other hand many vertical machines, including the new 33s, have a heavy screen all the way at the top, which means it is tricky to maintain balance when pressing the lowest row of buttons with your thumb. And since this is where (on), 0, (.), Enter and + are, you find this quite a pain sometimes.

Separate issue:

You (Karl) posted, and Raul Lion reminded me, of a simple and ingenious way to use Indirection in the 15c to make it possible to solve for *any* variable in an RPN equation, using the solver----which makes it VERY versatile!

Best Regards,

Bill

Edited: 13 May 2004, 2:26 p.m.

                              
Re: 33S is better....
Message #21 Posted by Veli-Pekka Nousiainen on 13 May 2004, 2:45 p.m.,
in response to message #20 by bill platt

"simple and ingenious way to use Indirection in the 15c to make it possible to solve for *any* variable in an RPN equation, using the solver"

I read that too - BUT I did not get it Could someone here walk me through that code how it works? {VPN} AXL

                                    
The "Karl's solver"
Message #22 Posted by Raul Lion on 13 May 2004, 3:40 p.m.,
in response to message #21 by Veli-Pekka Nousiainen

Here you have the Karl's post (and all the thread)

regards

                                          
The "Karl's solver" - HELP VPN
Message #23 Posted by Veli-Pekka Nousiainen on 13 May 2004, 5:56 p.m.,
in response to message #22 by Raul Lion

Yes - I read the following post AND I don't get it!

http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv011.cgi?read=29015

That's why I need someone to walk me through it... {VPN} AXL

                                                
Re: The "Karl's solver" - HELP VPN
Message #24 Posted by bill platt (les Estats Unis d'Amerique) on 13 May 2004, 6:34 p.m.,
in response to message #23 by Veli-Pekka Nousiainen

Hi Veli-Pekka,

OK, here's the deal: You have to tell the machine which variable is going to be solved for, before running solve. You do this by STOring the register address in to <I> before calling f<solve> <address> (It is exactly backwards of the 32sii model, where you 1st tell the machine which function is to be solved, and then when you press solve, the machine asks for the variable to be solved for).

You also must tell the machine what the declared variables values are first, before you do the f<solve> <address> call

[quote] from Karl: Use the I register in a general program. For example, to solve or integrate a function with variables x, y, and z:

(x) STO 1; (y) STO 2; (z) STO 3 {here you are storing the delcared values to the appropriate registers--this can happen before or after you make the program--here it is shown before}

001 LBL "A"   {in program mode, make an program address}
002 STO (i)  {this is the command which makes it all work}
(use RCL 1, RCL 2, and RCL 3 each time a variable is used)
nnn RTN

Key in n STO I (n = 1,2 or 3) and STO each of the parameters (fixed variables) as above prior to executing SOLVE or INTEGRATE on the program "A".

Any numbered registers may be used; the 15C won't use your choices for its own purposes.

That having been said, the 32Sii and 42S *are* much faster number-crunchers. While the 32Sii allows equations to be solved and integrated, RPN programs run faster.

[/quote]

OK, so like S=1/2at^2 + vi*t goes like this: make 1/2at^2+vi*t -S = 0

t STO 1
a STO 2
vi STO 3
S STO 4 

LBL A
STO  i
RCL 1
<ENTER>
<ENTER>
<ENTER>   {Horner's method.....not really needed entirely}
RCL 2
*
2
/
RCL 3
+
*
RCL 4
-
RTN

So to solve for S, start by storing all the acceleration, time intitial velocity etc, then 4 STO I, then f<solve><A> of course you can also store an initial guess in register 4 and etc.

To solve for another variable, just STO its register address in (I) and proceed with f<solve>A

I don't have my 15c in front of me but it should work.

Have fun!

Best regards,

Bill

Edited: 13 May 2004, 6:53 p.m.

                                                      
The "Karl" says 'thanks'...
Message #25 Posted by Karl Schneider on 14 May 2004, 2:56 a.m.,
in response to message #24 by bill platt (les Estats Unis d'Amerique)

... to Raul and Bill for acknowledging and explaining my fairly-straightforward technique for using SOLVE and INTEG for different variables of multi-variable programs without modifying code. (I admit that I could have explained it in a more-detailed fashion.)

This year, I plan to write several articles:

1. Practical uses of SOLVE and INTEG functions on RPN HP calc's (15C; 32S; 32SII; 33S; 34C; HP-41C/Advantage; 42S)

  • Area between two curves
  • Solving for one limit of integration
  • Using SOLVE and INTEG for multii-variable functions

2. Benchmarking of built-in Romberg's Method INTEG function on HP calculators

  • Short discussion of Romberg's Method vs. other common methods
  • Results and speed for one specific problem on all RPN models equipped with the method (and a few RPL models and other calc's or software)

Edited: 15 May 2004, 2:06 a.m. after one or more responses were posted

                                                            
Re: The "Karl" says 'thanks'...
Message #26 Posted by Ángel Martin on 14 May 2004, 8:01 a.m.,
in response to message #25 by Karl Schneider

Slightly off-topic, but also related to SOLVE.-

I wrote a short routine to calculate matrix (real) Eigenvalues, using SOLVE to find the roots of its definition equation.

It combines the matrix functions of the advantage (to calculate the matrix determinant on each iteration) with SOLVE. A bit of brute force, but works like a champ.

Here's the RPN listing, where the matrix is assumed to be created in XMem (labeled "AA"). The routine "MIDN" makes the current matrix into an identity matrix.

Let me know if you'd like further details.-

LBL "EIGEN"
"LOW'VAL=?"
PROMPT
"HI'VAL=?"
PROMPT
"*VP"
SOLVE
TONE 4
"VP="
ARCL X
PROMPT
RTN
LBL "*VP"
STO 00 -'stores current iteration value "IDN"
XROM "MIDN" -'makes matrix IDN an identity Matrix
RCL 00
"X"
MAT* -'multiplies [IDN]* X
"IDN,AA,IDN"
MAT- -'obtains [AA]-X*[IDN]
"IDN"
MDET -'calculates its determinant
RTN

                                                                  
E-values on the HP-42 w/Advantage
Message #27 Posted by Karl Schneider on 15 May 2004, 4:46 a.m.,
in response to message #26 by Ángel Martin

Angel --

I entered the eigenvalue program, except for the "MIDN" command, which does not seem to exist on the Advantage ROM. Am I using the wrong one?

-- Karl S.

                                                                        
Re: E-values on the HP-41 w/Advantage
Message #28 Posted by Angel Martin on 15 May 2004, 10:53 a.m.,
in response to message #27 by Karl Schneider

Karl, rather than posting it here I think it'll be much more efficient if I sent you the rom image (SandMath-III). In this way you can run the programs on V41, and of course copy them without issues.

There are two programs that use the Advantage SOLVE and INTEG that you may be interested in taking a look at. One is the above mentioned EIGEN (and yes you're right, MIDN isn't in the Advantage but a sub-routine within the SandMath). The other is a brute-force Fourier coefficients using SOLVE.-

In general there are many examples using INTEG to solve math problems, (Bessel functions, Beta function...) but they're normally slower than other techniques. Besides, they take another global label (for the function to integrate or solve), which means a precious FAT entry in a ROM!.

So send me an e-mail at "ovolacteo@onlinehome.de" and I'll reply with the rom image.

Best, AM


[ Return to Index | Top of Index ]

Go back to the main exhibit hall