Casio fx-3600PV/PA root finder

[Dave Britten]

I received a Casio fx-3600PA in the mail today, and it's definitely one the weakest scientific programmable calculators I own. 38 steps of program memory, 7 storage registers, and barely any flow control. Thanks to its two separate program areas, I was at least able to implement a simple root finder using the secant method, though it does require a certain amount of manual iteration control.

You'll need to do some extra prep when entering the program so that you don't run into a division-by-zero error while entering the steps - this model actually performs the calculations as the steps are being recorded! Just do all of the setup steps and program steps from top to bottom to get the program entered into program area P2.

Code:

Setup steps:
MODE EXP
1
Kin 3
Kin 4
2
Kin 5
P2

Program steps:
Kin 6
Kout 5
-
Kout 6
*
(
Kout 3
-
Kout 5
Kin 3
)
÷
(
Kout 4
-
Kout 6
Kin 4
)
=
Kin 5

This will leave you with 18 program steps to implement your function in P1.

Here's an example: -2x^2+13x-3

Code:

MODE EXP
P1

Min
x^2
*
2
+/-
+
1
3
*
MR
-
3
=

With P1 computing f(x), use this procedure to find the value of x such that f(x)=0:

1. Enter Guess 1
2. Kin 3, P1, Kin 4
3. Enter Guess 2
4. Kin 5, P1
5. P2 (next x value will be displayed)
6. P1 (next f(x) value will be displayed)
7. If display value is not equal to zero, repeat from step 5
8. Kout 5 to recall value of x

If you try the example above using 0 and 1 for initial guesses, you should arrive at x=0.239601355 after a few iterations.

Note: if the value of f(x) converges sufficiently close to zero, performing step 5 may cause a division-by-zero error due to the last two values of x being equal. If this happens, pres AC and Kout 5 to recall the last value of x. This can be seen if you try the example with initial guesses of 10 and 11, arriving at x=6.260398645.

[Hlib]

I bought a casio fx-3600pv 25 years ago. Already in the mid-90s, this calculator was not interesting to those users who had experience with normal calculators at that time.
<<... - this model actually performs the
calculations as the steps are being recorded!>>
These steps need to be typed blindly without displaying codes and a counter, running your finger over a piece of paper with the program text. Now I`m looking at a new generation of designers who are designing an office chair in SolidWork that falls apart after half a year of operation. But in those days we calculated very complex projects on such primitive calculators.

[Thomas Radtke]

And the legend lives on:

https://www.casio-schulrechner.de/de/pro.../fx3650p2/

Thanks for the program, nice to see someone still cares for this programming scheme. Hope students today have as much fun with it as I had back then with the fx-81p.

[Csaba Tizedes]

(12-23-2020 09:32 PM)Dave Britten Wrote:  Casio fx-3600PA (...) implement a simple root finder using the secant method (...) manual iteration control.

Hello Dave, the 50F has only 28 steps and that is good enough
You can find my 13 steps secant method HERE, posted months before on MoHPC. nJoy!

Csaba

[Pekis]

(12-25-2020 04:48 PM)Thomas Radtke Wrote:  And the legend lives on:

https://www.casio-schulrechner.de/de/pro.../fx3650p2/

Thanks for the program, nice to see someone still cares for this programming scheme. Hope students today have as much fun with it as I had back then with the fx-81p.

Hello, fun indeed, and for those who didn't see it already in a previous post, here is a "HMS+" program for the Casio fx-3650P II:

Code:


0->A:0->X
Lbl 1
?->X:Fix 0
((X>=0)-(X<=0))(250Abs(X)-Rnd(100Abs(X)-.5)-60Rnd(Abs(X)-.5))/90+A->A   
((A>=0)-(A<=0))(90Abs(A)+Rnd(60Abs(A)-.5)+100Rnd(Abs(A)-.5))/250->X
Norm 1:Goto 1

Notes:

• Rnd(X-.5) stands for INT(x) where x>0, because INT(x) is missing
• ((X>=0)-(X<=0)) stands for SGN(x), because SGN(x) is missing
• (250X-INT(100X)-60INT(X))/90 stands for HMS->DEC(x) where x>0, because HMS->DEC(x) cannot be used in programs
• (90X+INT(60X)+100INT(X))/250 stands for DEC->HMS(x) where x>0, because DEC->HMS(x) cannot be used in programs
• Input x (negative or positive) in HH.MMSS..... format
• Program displays accumulated result in HH.MMSS..... format, waiting for next x

[Dave Britten]

(12-25-2020 08:48 PM)Csaba Tizedes Wrote:  

(12-23-2020 09:32 PM)Dave Britten Wrote:  Casio fx-3600PA (...) implement a simple root finder using the secant method (...) manual iteration control.

Hello Dave, the 50F has only 28 steps and that is good enough
You can find my 13 steps secant method HERE, posted months before on MoHPC. nJoy!

Csaba

Very nice! I still need to track down a 50F (or the Radio Shack version) to experiment with.

[Csaba Tizedes]

(12-25-2020 04:48 PM)Thomas Radtke Wrote:  And the legend lives on: https://www.casio-schulrechner.de/de/pro.../fx3650p2/

The CASIO fx-50FH (50F Plus) has 680 steps, limited MODEs but improved programming function set compared to 3650P(II). Check the linked docs at the bottom, very useful!

A little quiz: If 50FH has only seven (A, B, C, D, X, Y and M) variables and I use all of them to store Cash Flows in my IRR solver, how is it do the calculation?!
Try to figure out, this little machine has hidden possibilities:





Csaba

[Csaba Tizedes]

(12-25-2020 09:13 PM)Dave Britten Wrote:  Very nice! I still need to track down a 50F (or the Radio Shack version) to experiment with.

Thanks, I'm sure on 3600PV the x<>Kn is also available.
Cs.

[Dave Britten]

(12-25-2020 09:21 PM)Csaba Tizedes Wrote:  

(12-25-2020 09:13 PM)Dave Britten Wrote:  Very nice! I still need to track down a 50F (or the Radio Shack version) to experiment with.

Thanks, I'm sure on 3600PV the x<>Kn is also available.
Cs.

Yup, it is, I'm just not enough of an AOS programming whiz to have used it.