Posts: 2
Threads: 1
Joined: Jan 1970
Hello there.
This is my first post. I use hp rpn calculators since long and have some of them (45, 21, 41cv, 41cx, 11c, 12c, 28s, 48sx, 23sII). I also have a 20s, not rpn, but never used it.
Im looking forward to buy the new 35s, but here where I live I cant't test it, have to order it per internet.
Can someone tell me how good is the 35s for working with 2d and 3d vectors, changhing them from rec to pol and so on?
And how good is it for solving linear equations systems and second degree equations?
Thank you very much for the help
Reto
Posts: 150
Threads: 6
Joined: Sep 2007
Get the HP 33s (!) pdf manual and read it. You'll get a pretty good impression what the 35s (!) can do, both (machines and manuals) are quite similar in many ways.
Go to hp.com and read the pdf training modules for the 35s (!) that I've linked you to about whatever interests you and find out more about that machine. Especially about vectors here as a pdf.
Vector support is limited, matrices are not supported at all, just a build in linear solver up to 3x3...
Edited: 9 Oct 2007, 1:49 a.m.
Posts: 151
Threads: 18
Joined: Aug 2007
I belive you will be disapointed with the vector type. It has not been utilized much by the machine itself: Only absolute value, vector arithmetic between vectors and vector/scalar is provided and cross product is missing.
I *do* find them usefull since I am programming and as such do create new vector operations in this: project.
But if you are not interested in programming with vectors you propably will be disapointed.
And the bad news does not stop: If you are going to be programming with vectors it is my experience that you will have big problems with the 'vector syntax bug'. The machine stop accepting vectors. I am almost positive this happens when one has been in programming mode and/or equation mode and return to 'user mode'. Strange magic gets you out of it. The only thing that saves this bug from render vectors useless on this machine is that it do not seem to happen when operating purely in 'user mode'.
In general see the museums HP35s bug
article.
Posts: 2
Threads: 1
Joined: Jan 1970
Thank you for your answer Meenzer.
I already have read the pdf about vectors on the hp website. Not very useful, can't understand why they publish it, it says nothing about cross product or polar form, gives only two examples of sum and dot product.
Now I looked at the 33 manual. Looks like the the calculator itself can not do very much with vectors. If I understand in right, you have to write programs yourself for polar form, cross product, rectangular to polar conversion and so on. Am I right?
On the 33 manual there is also a program to solve simultaneous equations. Can the 35s not solve such equations itself, I mean without writing a program?
Thank you very much
Reto
Posts: 735
Threads: 34
Joined: May 2007
Still what I consider a cool feature is the ability to use vectors within the solver. Not as variable you solve for but as input parameters for an equation.
Here's an example:
 Points A and B define a straight line.
 Point M is the center of a ball with radius R.
Now you can solve the following equation for T:
ABS(A + T*(B  A)  M) = R
IMHO that's pretty straight forward. And the same equation can be used for 2dimensional and 3dimensional vectors.
Posts: 1,755
Threads: 112
Joined: Jan 2005
Hi, Thomas:
Thomas posted (the underlining is mine):
"Still what I consider a cool feature is the ability to use vectors within the solver. Not as variable you solve for but as input parameters for an equation."
That's not exactly correct, there are cases where SOLVE can and will solve for a vector variable. For instance, in RUN mode key in this proofofconcept example:
EQN Tx5=[2,0,0]
SOLVE T > T = [0.4, 0, 0]
which is correct; more complicated examples are possible. The rule is that SOLVE will correctly solve for a 2D (or 3D) unknown vector if it has the 2nd (and 3rd) components equal to 0. This is covered in the User's Guide, by the way.
Similar restrictions apply to solving for complex numbers: this is solvable:
EQN 5*T=2i3
SOLVE T > T = 0.4i0.6
but this is not:
EQN 5*T*T=2i3
SOLVE T > BAD GUESS or NO ROOT FND
depending on the contents of variable T and stack register X.
Not tremendously useful, but it doesn't hurt to know about it either.
Best regards from V.
Edited: 9 Oct 2007, 5:38 a.m.
Posts: 1,755
Threads: 112
Joined: Jan 2005
Hi, reto:
reto posted:
" already have read the pdf about vectors on the hp website. Not very useful, can't understand why they publish it"
"If I understand in right, you have to write programs yourself for polar form, cross product, rectangular to polar conversion and so on. Am I right?"
"On the 33 manual there is also a program to solve simultaneous equations. Can the 35s not solve such equations itself, I mean without writing a program?"
Just the 2 and 3simultaneouslinearequations cases, and not very convenient to use at that.
As a side comment, this machine has:
 no support for matrices or linear systems (save 2 and 3 cases)
 minimal support for vectors
 minimal support for alpha capabilities
 passable support for fractions
 decent ergonomy and usability
 much better support for complex numbers (but still lacking in some important ways)
 good support for equations
 excellent programmability (also lacking some important features)
Best regards from V.
Posts: 4,587
Threads: 105
Joined: Jul 2005
Of the RPN calculators "available" today, a 42s would fulfill your requirements best. Please read about its features in this museum. Many members of this forum consider the 42s being the best RPN calc ever built (though it has some shortcomings, too).
The bad news: This calc is pretty expensive to get your hands on, because many people share the opinion about its value, and it is almost exclusively available on auction sites.
Based on reported sales success of the 35s, there may be another calc following it, possibly a 42ssuccessor, but so far this is pure speculation or wishful thinking, whatever you want to call it.
HTH, Walter
Posts: 70
Threads: 6
Joined: Oct 2007
I will be excellent, if a little slow, but that won't matter for relatively small numbers of equations. With banded symmetrical equations, I'd guess you could get up to 12 or 15 equations into it, including the halfbandwidth solver. Skyline storage or a frontal solver might do better.
It is also an excellent calculator, and I'm pleased to bits with it except for one problem. This one problem is, however, fairly catastrophic. Mine gets stuck in a loop from which I cannot break out, if my code makes it possible to do that by accident before I have finished debugging.
From your question, I would guess that, like me, you will be developing some complex software, some of which could also be quite long. Well, here's the thing. If your machine is like mine then you probably cannot develop advanced software like this. This is beacause, unless you can guarantee there will be no loops with bugs in them, you will not be able to debug your program without constantly losing the entirety of your work  yes, ALL OF IT  in both the current program and everything else you have also written.
The only way of brealing out of a loop, I have found, is to use a hardreset, which clears the entire memory. Unless I do this, I can't even switch the calculator off when it gets into its stuckinaloop modes.
John Wasilewski
