hp33s

Model:	33S, calculator is based on 32SII

General:
	Name:			?
	Code-Name:		Swan
	Family:			?
	Logic:			RPN, algebraic
	Features:		?
	Firsts:			integer division, creased keyboard
	Introduction:
		Date:		2003-10-? (possibly 2004)
		Price:		$66
	Discontinuation:
		Date:		in production
		Price:		?
	Production-Run:		?

Display:
	Type:			LCD, bit-mapped
	Size:			2 lines, 14 chars, 5x7 dots = 14 x 90 pixels
	Number-Formats:		sign, 10 mantissa, ., exp sign, 3 exp
	Annunciators:		Left Shift
				Right Shift
				ALG
				RPN
				GRAD
				01234
				A..Z
				B (for busy)
				HEX
				OCT
				BIN
				HYP
				!
				batt
				<-
				->
				^
				V

Data:
	User-Visible:
		Smallest:	1E-499
		Largest:	9.999999999E499
		Signif.-Digits:	12
	Internal:
		Smallest:	1E-49999
		Largest:	9.999999999E499
		Signif.-Digits:	15
	Data-Types-and-Sizes:	real, 8 bytes?

Memory:
	Named-Registers:	X, Y, Z, T, last X, A-Z, i, /c,
				  n, \GSx, \GSy, \GSx^2, \GSy^2, \GSxy
				(all fixed allocation and do not affect
				program steps)
	Flags:			12
	Register-Usage:		N/A
	Numbered-Registers:	0
	Program-Steps:		10000
	Program-Editing:	insert
	Program-Display:	alpha
	User-RAM-Bytes:		32K
	Total-RAM-Bytes:	32K
	ROM-Bytes:		?
	Machine-State:		memory
	File-Types:		none

Physical:
	Technology-Used:	?
	Processor:		Sunplus SPLB31A, based on the 6502
	Chip-Count:		?
	Power-Source:		CR2032 x 2
	Continuous-Memory:	yes
	Expansion-Ports:	none
	I/O-Ports:		none
	Clock:			?

	Length:			158.0 mm
	Width:			83.0 mm
	Height:			16.1 mm
	Weight:			119 g
	Temperature-Range:
		Operating:	?
		Charging:	?
		Storage:	?

Keyboard:
	Switches:		none
	Shift-Keys:		Left Shift, green, above left
				Right Shift, red, above right
				Alpha, blue, right
	User-Defined-Keys:	6

Key-Arrangement::

**    **    **
    **  **
**    **    **
** ** ** ** **
** ** ** ** **
** ** ** ** **
** ** ** ** **
** ** ** ** **
** ** ** ** **
** ** ** ** **
** ** ** ** **

Key-Labels-Base-Keyboard::

	    ^
ENG			MODES
	<	>
SOLVE	    v		DISPLAY
 x		 x
e	LN	y	1/x	\GS+
	 2
Rv	x	\v/x	x\v/y	%
STO	RCL	SIN	COS	TAN
XEQ	x<>y	+/-	E	<-
R/S	7	8	9	\:-
<|	4	5	6	x
|>	1	2	3	-
C	0	.	ENTER	+

ON key in black below C
 b/c
a    key in black below .

Key-Labels-Left-Shift-Green-Above-Left::

	    []
<-ENG	[]	[]	CONST
[]	    []		[]
  x
10	LOG	ABS	x!	\GS-
	 3	
HYP	x	3\v/x	RPN	INT\:-
CMPLX	RND	ASIN	ACOS	ATAN
GTO	MEM	nCr	nPr	CLEAR
PRGM	v	^	ISG	x?y
[]	->\Gt,r	->HR	->DEG	BASE
[]	->kg	->degC	->cm	->l
[]	INPUT	FDISP	LASTx	LBL

Key-Labels-Right-Shift-Red-Above-Right::

	    []
[]	[]	[]	[]
[]	    []		[]
\.S	L.R.	x-,y-	S,\gs	SUMS
R^	IP	FP	ALG	Rmdr
EQN	INTG	SGN	\pi	%CHG
FN=	x<>	(	)	=
PSE	SEED	RAND	DSE	x?0
[]	->y,x	->HMS	->RAD	FLAGS
[]	->lb	->degF	->in	->gal
OFF	VIEW	/c	SHOW	RTN

Key-Labels-Alpha-Blue-Right::

	    []
[]	[]	[]	[]
[]	    []		[]
A	B	C	D	E
:	F	G	H	I
[]	J	K	L	M
[]	N	O	P	[]
SPACE	Q	R	S	[]
[]	T	U	V	[]
[]	W	X	Y	[]
[]	Z	i	(i)	[]

Programmable-Operations::

- modes: deg, rad, grad, ., ,
- display: fix, sci, eng, all

Program Memory

3\v/	cube root
FP	fractional part
IP	integer part
x^3	cube
\pi	12 digit pi on user input, 15 digit pi within programs

- flags: 0-11, SF, CF, FS?
- x?y x?: < > = <= >= !=
- show: press and hold SHOW to dipslay checksum value, program length
- mem: press MEM for the catalog of program labels
- XEQ, R/S: execute a program
- GTO, LBL, SOLVE, INPUT, VIEW, ISGn, DSE, FN=, RTN, PSE ke y
- (i) indirect address
- solving and integrating programs
- statistics programs:
	* curve fitting: straight line, exponential, logarithmic, power
	* normal and inverse-normal distributions
	* grouped standard deviations

Statistical Function

- one/two variable statitcis
- linear regression
- n, \GSx, \GSx^2, \GSy, \GSy^2, \GSxy

Base-n Function

- binary, octal, decimal, and hexadecimal number calculation and conversion
- complement calculation

Equations

- equalities: F(a)=G(b)
- assignments: a=G(b)
- expressions: G(b)
- SOLVE
- XEQ

System

- internal Napier precision
- 13-level partheses

Mathematical Function

- trignonometric functions, inverse-trigonometric, hyperbolics
- polar-rectangular coordinate conversion
- power, square root, square, cube root, cube, reciprocal
- logarithmic and exponential
- natural logarithmic and exponential
- factorial
- RND
- int \-:
- IP, FP, ABS, SGN
- permutation and combination
- RAND, seed

Fraction Mode

- /c function
- FDISP functin: fraction, decimal conversion
- most precise fraction
- factors of denominator
- fixed denominator

40 Physics constants

- nearly all the ones on the 48GX
- magnetic moment
- muon mass
- conductance quantum

scientific, engineering notation; unit conversions

Non-Programmable-Operations::

left shift, down moves to the bottom of the equation or program list
left shift, up moves to the top of the equation or program list
right shift, ALG: sets algebraic mode
left shift, RPN: sets RPN mode
0, right shift, SEED: stores a random value seed

Menus::

CONST
	c	g	G	Vm	Na	Rw?	e	me	...
		mp	mn	m\Gu	k	h	h-	\Gpo	...
		\Ga0	\GEo	R	F	\Gu	\Gu0	\GuB	...
		\GuN	\GUp	\GUe	\Gun	\Gu\Gu	r\Ge	Z0	...
		\Glc	\Glcn	\Glcp	\Ga	\Gt	t	atm	...
		\GY\Gr	C1	C2	Go

MODES
	1 DEG	2 RAD
	3 GRAD	4 .  5 ,

DISPLAY
	1 FIX	2 SCI
	3 ENG	4 ALL

SUMS
				    2	    2
	n	\GSx	\GSy	\Gsx	\Gsy	\Gsxy

HYP is an implied menu, converting the trigonometric functions into their
hyperbolic equivalents.

BASE
	1 DEC	2 HEX
	3 OCT	4 BIN

FLAGS
	1 SF	2 CF
	3 FS?

Bugs/ROM-Versions::

- 0 SEED (from DATAFILE V32N2P4)

- after exiting a self-text (C-y^x or C-1/x), the RAD/GRAD, 0-4, and
HEX/OCT/BIN annunciators are cleared, but the corresponding modes and
flags are not changed (reported by Ben Salinas)

- The ->HMS function does not always yield the correct result when the
input is a negative number.

- When doing ->y,x on a number where the X or Y value may be zero, the
result may be inaccurate.

- The Cn,r function will not display an overflow message but will
instead return an inaccurate answer.

- Antonio Maschio posted an entry to the HP Museum board about an
integer division bug.  When invoked by keyboard or from program, it
returns the correct results.  However, the results differ in EQN:

Operation  INT\:- Rmdr   IDIV(EQN)  RMDR(EQN) INT\:-(PRGM) RMDR(PRGM)   
 9 / 7      1     2         1          2         1          2   
 9 /-7     -2    -5        -1         -5        -2         -5
-9 /-7      1    -2         1         -2         1         -2
-9 / 7     -2     5        -1          5        -2          5

Full writeup at:

http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv016.cgi?read=97090#97090

- This machine went through a number of different screens.  In the
Fall of 2005, they introduced yet one more.

Notes::

Manufactured by Kinpo (http://www.kinpo.com.tw).

From USENET:

	From: Leif Harcke <lharcke@stanford.edu>
	Newsgroups: comp.sys.hp48
	Subject: Re: HP-33S Manual online
	Date: Sat, 08 Nov 2003 10:02:13 -0800
		...
	The manual is a one-to-one mapping from the 32sII manual, with an
	additional Appendix C - "ALG: Summary" on the algebraic mode.

	The 33s will have 
	  1) 31k user memory vs. 384 bytes in the 32sII
	  2) letter+4 program line labels (A0001 vs. A01)
	  3) Algebraic mode and algebraic programs
	  4) 2-line display (X and Y of stack, much more in algebraic mode)
	  5) cursor keys to navigate menus

	The programming model remains variables/labels A-Z and index i, so the
	increase in memory doesn't help much if you exhaust the label or
	variable space.  The algebraic mode may expand the user base for this
	type of programmable scientific calculator.  The complex number model
	remains the same as that in the 32sII, i.e. X contains real and Y
	contains imaginary.  It needs all 4 stack levels to add two complex
	numbers.  No 15c-like complex numbers.  The algebraic complex number
	mode looks interesting.

	-Leif

From "Summing up the HP-33S" by Jordi Hidalgo, DATAFILE:

- simple loops run about twice as fast as on the 32SII provided ISG
and DSE are NOT used

- solving and integration are 1.5 - 3 times as fast

- there's a reset hole in the back

- the length of an equation is now limited to 255 characters

- the trailing parenthesis (")") of two-argument functions is now
required

- on the other hand, +/- and - can ce used interchangably on the
second argument

- EXP(-X^2) now means EXP(-(X^2)), which fixes an inconsistency on the
32SII:

	25-5^2 evaluated to 0 while
	-5^2+25 evaluated to 50

- a colon and not space is used to separate arguments, but both the
colon and space can be used in messages

- calculations with "too big" numbers in binary modes are OK so long
as the result is in display range

- if a program is stopped while calculating an integral, the program
can't be resumed

- certain calculations do not produce the same results on both
machines (e.g., gamma of 9.29 or 8.29!), suggesting that the CPU is running
native rather than full emulation

- MEM and SHOW results differ, as do checksum calculations for
equations and programs.  Nibbles no longer come into play.

				Amount of Memory Used
Data			32SII				33S

Variables	8 bytes per non-zero value.	No bytes. Variables (including
		(No bytes for zero values).	i and the statistics registers)
						take up no user memory.
Statistics data	48 bytes maximum (8 bytes	Always allocated.
		for each non-zero register).

Numbers in	Integers 0 through 254: 1.5	15 bytes. No short form for
program lines	bytes. All other numbers: 9.5	frequently used values, but
						constants in program lines
						take 3 bytes, unless they
						are part of an equation.

Instructions in	1.5 bytes			3 bytes
program lines

Equations in	1.5 bytes + 1.5 for each	3 bytes + 1 byte for each
program lines	function + 9.5 or 1.5 for	character (255 maximum).
		each number. Each ( and each
		) uses 1.5 bytes, except
		the ( for prefix functions.

Numbers in	Integers 0 through 254: 1.5	Each entry in the equation
equations	bytes. All other numbers: 9.5	list takes 6 bytes + 1 byte
						for each character. (255
Operations in	1.5 bytes			maximum).
equations

- the algorithms for SOLVE and \S have been rewritten. VIEW and STOP
instructions in programs that are being solved or integrated are
executed only once, not each time the routine is called by SOLVE or
Integrate; in order to monitor their execution, intermediate steps can
be displayed by the sequence VIEW var PSE

- the new message INVALID VAR is displayed when attempting to solve an
equation for a non-existent variable.

- SOLVE now fails to find the pole in the example on page D-7 of the
user's manual (C-8 in the 32SII owner's manual)

A change in the LCD screen to make it more legible will be made in
late 2005.