HP 33S help



I just bought a HP 33s in preparation for the FE Exam. Upon receiving the calculator, I noticed something does not seem right with the calculator.

When I press a number followed by enter the calculator populates the X and Y stack with that numbers. For example, if I want to press "2" , "enter", "1", "enter". the display would show 1 on both the X (botton) and Y (top) stack.

I have gone through the book to see if I have a setting wrong, but have found nothing. In one of the chapters it talks about populating the stack, and I follow the exact example in the manual, and my calculator will not replicate the example.

Any suggestions?


Hi, Clay;

this is completely correct. Everytime you press [ENTER] while in RPN mode, whatever (number) is in the X register is copied in Y register. In fact, the stack contents are "bumpped up" one register, meaning T-contents are lost, Z-contents are pushed up to T, Y-contents are pushed up to Z and X-contents are copied in Y.

If you want to perform any calculation involving a two-number function (+, ×, -, ÷, y^x, and some others) first key in Y-contents, press [ENTER] only once, key in second number and press the key (sequence, if applicable) to perform the desired calculation. If it is a one number function, like square root or reciprocal (1/x), simply key in the number and press the key (sequence, if applicable).

This is probably the main (only?) difficulty found by those wh have not used RPN calculators before: master the [ENTER] key usage. What else is somehow easy.

Remember: you can use the HP33S in algebraic mode as well, and keystrokes must be changed in this case.

Have you found yourself in the need to ask more questions, feel free!

Best regards.

Luiz (Brazil)

Edited: 22 Jan 2005, 3:41 p.m.


Thanks for the quick response.

I might not have explained the problem very well.

With my HP 48GX, if I press:

1 - enter
2 - enter
3 - enter
4 - enter

I have the following in my stack

T-stack = 1
Z-stack = 2
Y-stack = 3
X-stack = 4

If I do the same with my HP 33S

1 - enter
2 - enter
3 - enter
4 - enter

I have the following in my stack

T-stack = 2
Z-stack = 3
Y-stack = 4
X-stack = 4

For some reason, when I put a number into the X stack, it also forces the number into the y stack. So any number I press followed by enter, it goes into my X and Y stack.

Am I explaining myself very well?




ENTER works somewhat differently on these two machines.

ENTER on the HP-33S works like it did on traditional RPN calculators dating all the way back to the original HP-35 in the early '70s. On these calculators, numbers are entered directly into the X register, so ENTER duplicates X into Y. Immediately keying in another number overwrites X.

On the HP-48 and similar, the purpose of ENTER is to pop stuff from the edit line into the X register (i.e. Level 1). It takes a second ENTER to copy that number into the Y register (Level 2) and immediately keying in another number does NOT overwrite X.

I hope this helps.



Hi, Fred;

sorry posting after you... It took me some more time to write what I wanted to.

Best regards.

Luiz (Brazil)



I don't mind at all. Besides, your explanation went further than mine did.

I'm much more comfortable with traditional RPN--30 years of experience will do that--and I'm still a rookie on the 48. In fact, I probably wouldn't have responded had I not been in the middle of writing my first significant RPL program and spending lots of time with the Advanced User's Reference Manual.

I have a simple hydrology (storm runoff) program I wrote several years ago for the 42S. The 42S version is 1211 bytes. The new 48 version (actually two programs--one creates data files for different situations and the other does the calculating--plus one subroutine for building a screen full of results) totals 2066 bytes. Actually, the calculation is pretty trivial. The bulk of both versions is required to make the I/O friendly and handle two sets of parallel calculations (three different storm levels and two different methods of estimating the intensity of the storm). The 48 program works as of about 9:00 p.m. tonight. Now all I have to do is look for ways to streamline it and I want to slightly modify one small feature.

After this venture, I have decided that RPL is very powerful but can be a pain in the rear end.



Hi, Fred;

just to add that I felt somehow scared with the HP28S after using an HP41C and an HP15C for about five years. Anyway, I had my first contact with C language a couple of months before seeing an HP28S and I remember thinking that it would be great having a calculator that could be programmed in a structured language. I like RPN and the keystroke programming, but I feel more comfortable with structures than with GoTo and GoSub (or call). You see, an RPN based calculator could also have a structured programming language, but I guess that it would oversize program power. I also believe that the procedural paradigm chosen for RPN clculators might have been the only choice at that time (70's and 80's). I also see no problem on having a program language with GoTo and GoSug, Label and Return in an RPL-based calculator (in fact, their assembly code uses Goto, Gosub and Return), but that would make things difficult.

You see, I take 'programming' and 'keystroke usage' as different aspects. There are calculators that cannot even be programmed, others that use the same entry system with different program paradigms (mostly algebriac). Till now I did not take some time to force myself using and understanding how to program the HP49G while in algebraic mode. I read the manual and felt myself not comfortable with the ideas. But I must know what algebraic programming in an RPL-based calculator is like prior to have an opinion about this.

Best regards.

Luiz (Brazil)

Edited: 23 Jan 2005, 4:21 p.m.


@Fred and Luiz:

Interesting points of view. I have to admit that - after decades of RPN interface and RPN programming - I felt most unconfortable when I bought my first RPL model (48SX).

The strictly postfix and "stack consuming" paradigma used in RPL is a good thing for complex programs. I did realize this after a very short time.

But what is powerful (and maybe better) for greater solution libraries must not be necessarily the best user interface for daily problems!

Coming from the RPN world I really dislike the RPL approach for ad hoc calculations; what I dislike more is that the 48/49 series doesn't give me a switch to RPN mode for daily business.

An exeample: That's not postfix logic, but for my part the better human interface.

function: store the content of X into variable/register A
RPN: STO A (not strictly postfix, but like "we speak")
RPL: 'A' STO (more keystrokes, and where is my X gone for further calculations?)

Not to mention the daily programming stuff. A little program with 2 subroutines and a loop is keyed in without manuals and big thinking. In RPN. Without manual, because of the simple functions interface.

Back to the initial question, regarding the ENTER/DUP keystroke for ad hoc calculations:
RPN: obj ENTER --> obj obj (stack lift off)
RPL: inp ENTER --> obj or obj ENTER --> obj obj
in RPN you always know what ENTER does
in RPL it dipends if you have a hanging input or not.

That's why I carried many years a RPN model for daily use (additionally to the 48)! But in a few week I hope I need it anymore, because I'm writing a userRPL program (RPNcalc) that does exactly what I need: A RPN interface for RPL calculators.

My 0,02 Euros



Thanks for your comments. I agree with everything you said. When I got my 48 I too was surprised at how hungry it was, consuming all my numbers!

If you get your program working, I would be interested in giving it a try if you don't mind passing it out.




...if my daily work permits (with 32SII, for the moment, hehehe...) I hope to finish my program next month.

I will post a new thread for a Beta and - after working on feedbacks - upload it to hpcalc.org.



"RPN you always know what ENTER does in RPL it dipends if you have a hanging input or not."

no, No, NO! Both requires advanced knowledge about the ENTER operation. IMVHO RPL ENTER is easier (unless Vectored).




Thanks for your reply. For me, at my age (46) and with my current responsibilities and activities, I simply don't have the time to program my calculators like I did even 10 years ago. Thus, I will probably never really get comfortable with RPL, let alone master it, but my 30+ years with RPN means I can pretty much do it in my sleep (which, BTW, I don't get enough of!).



Any RPL/Algebraic program will work in Algebraic/RPL mode.



I dont think so. The logic is different. That is if you truly write an RPL program using all of its power. I have not tried that,but why would I want to. If I wanted to program in algebraic, I have gotten a TI.


Think what you want, but I tell the truth.
Vectored ENTER is not supported after 1.19-3
but otherwise ALG/RPL programs do work in RPL/ALG mode
Try it out yourself!
You may take a peek at an RPL program in ALG mode using the Filer.
You will then see
RPL> << .... >>

Hi, Clay;

the problem is that you are taking an RPL-based calculator (HP48GGX) and comparing with an RPN-based calculator (HP33S). You see, the HP48GX has an input line (also called command line), where you can key in any user-valid object. Try this in your HP48GX:

You see that only the number 1 is in the stack level 1, right? Press [ENTER] again and you'll see a copy of it in level 2 as well. OK, OK, you'll tell me that a DUP command is performed when there is no "active" input line and you press [ENTER] in the HP48GX, right? And this is true! Press [ENTER] with nothing in the stack and a DUP error message appears in the HP48GX.

The HP33S has no input line, so pressing ENTER performs the ENTER function, also programmable (there is no [ENTER] command or function in the (user)RPL structure; in the HP48GX, ENTER is an operation and cannot be part of an object). Everythime you key in a number in the HP33S it is already being placed in the X-register. To key in an object in the HP48GX (or any other RPL-based calculator) you should key it in the input line first and then press [ENTER] to 'push it up' into the stack.

That's why a copy of any number already in the X-register is sent to the Y-register when you press [ENTER] in the HP33S. In fact, in an RPN-based calculator, [ENTER] performs three actions:

1 - ends (finishes) number entering (when keyed in by the keyboard)

2 - "pushes" stack contents up and copies X-contents into Y-register

3 - disables stack auto-lift so the next entry (either with the keyboard or with a RCL) overwrites the contents of the X-register

The third action becomes concise when you remember that a 'safe' copy of the X-contents is already available in the Y-register. The only loss is T-contents.

I think this sentence of yours makes sense, now:

For some reason, when I put a number into the X stack, it also forces the number into the y stack. So any number I press followed by enter, it goes into my X and Y stack.

Please, feel free asking again. For now, I'd suggest having a brief with the HP33S User's Manual, chapter 2, pages 2-5 to 2-7. If you think you need a bit more, chapter 2 is only 14 pages long.

Best regards.

Luiz (Brazil)

Edited: 22 Jan 2005, 7:13 p.m.


Thanks guys,

I am by far NOT a calculator expert, but I have become pretty comfortable with my HP 48 GX, and although HP 33S is nice, it is defiantly not a 48GX. Yep, I hate change.

Thanks so much for the information. I truly thought my calculator was not functioning properly. After you guys responded, I went back through Chapter 2 and everything worked as you guys suggested. I guess I was assuming that things to work as they did with the 48 model, and miss understood chapter 2 the first time.

Anyway, I am on track now thanks to you guys.

Have a great weekend.



You can get the 33S to behave a bit more like the 48 series by a program:


Run XEQ A to get into this program and then use the R/S key as a substitute ENTER key. This will avoid the duplication of the X into Y.

Put an ENTER function after the R/S to get the R/S key to behave like the ENTER Key but in a different location


Remember that this is not change but a return to the basic RPN calculators. The change was the 28s and eventualy the 48 and 49. Those idiots in NCEES forced HP to go back to the Old Style calculators in order to take the exams. Soon they will want examines to use slide rules, logarithm tables and measuring tapes.


I need help with the polarities of new batteries placement.
There are no polarities shown in the battery compartment, nor in the manual.
rjyak --


Hi, RJYak;

the polarity of all Spices (HP31E, 32E, 33E/C, 34C, 37E and 38E/C) is as follows.

By facing the back of the calculator, batteries compartment in the upper side, you should see the contact tabs in the left side, OK? The upmost contact is the negative (closer to the AC adapter square hole). The other one is the positive.

When I use standard rechargeable batteries (not the pack supplied with the calculator), I add a small, thin metal stripe in the right side of the compartment when inserting the batteries that must be enough to "close the circuit".

Hope this helps.

Luiz (Brazil)

Edited: 30 Jan 2005, 2:01 a.m.

Possibly Related Threads...
Thread Author Replies Views Last Post
  Integration Times "Old" 33s vs "New" 33s John Smitherman 21 3,498 12-14-2005, 12:04 AM
Last Post: Karl Schneider

Forum Jump: