Scientific Calculator Design - NEED HELP!!

Hello all,

I'm in the process of designing a scientific calculator with at least 10^99 resolution. I did some research on scientific calculator (SC) and found that SC's use Binary Coded Decimal (BCD) floating point arithmetic to obatin high precision.

Here are my questions and I hope someone out there can help me:

What is the best way implement a SC?
Are there BCD math routines available for microcontroller (PIC, H8, 6811, 8051)?
Has anyone designed something similar to a SC using an FPGA?
What route should I take for this project? Please suggest.

Thanks for everyone's help

I can't believe no one out there can help me with a scientific calculator design. Please any thoughts are welcome. I'm just in the research stage of the design.

Thanks

8051 Basic has Scientific commands.
At the past I intesrested by floating point numbers, trigonometric functions and I learned from 8051 Basic.

I disassembly the firmware of 8051 and I found all subroutines for arithmetic operations.

In this days you may find the ready source off 8051 and you can learn.

i think you need two processors for pic land
or infact any micros

one handles bcd the other operations
like display and addresses the calculator slave
via a psp port & a serial connection
this way you can break the psp into nibbles
and dump the calculated result accross a port in psp mode
dependant on the mode you set in serial

forinstance a basic stamp can have up to 5 programs reside
same can be done in a pic

http://www.geocities.com/andretuly/pv450addins.htm

i found this
it is a site with modifications
but containes other details and circuits also some links to other resorces

do you need a circuit of a proposed sc calc

or a keypad made in proteus for this purpose
if you do let me know and ill make you a keypad that works aok in proteus for a sc calc
get me a good picture of the front of one you want ill add also the keys names and any function letter

ill give you alsoa clear design to show the method

Thank Bunalmis and Monkey.

Bunalmis I have searched 8051 sites for math libs but the math libs are for 32-bit floating point. Single precision floating point has a 7-bit exponent, which aren't enough bits to represent 10^99. :? If you know where I can get 64-bit floating point or BCD math libs, please let me know.

Monkey I was thinking of using a CPLD to take care of the LCD display and the Keypad screening and uC for calculations only and I can read/write to the CPLD using PSP or I2C. Adding a CPLD will add to the cost though. Right now I'm leaning towards a uC with enough I/Os to take care of both LCD and keypad. Any suggestion are welcome.

Hardware in not so much the problem. I just don't know how to deal with 10^99 precision.

Thanks

Use PIC micros

Use pic microcontrollers (PIC16f877)beacuse of all / any of the following reasons :
- A very cheap programmer can be made with simple parts
- A lot of programmer designs available
- lots of I/Os 33
- 8kb eeprom
- lcd / keyboard interface easy to make
- technical support available from microchip

If you find the assembler source of 8051 Basic,
you can modify the floating point routine of 8051 Basic for 64 bit FP.
8051 Basic uses BCD arithmetic for all floating point operations.

The source for Basic-52 is here:
hxxp://www.programmersheaven.com/d/click.aspx?ID=F3736

have fun

http://www.datamath.org/Toshiba_IC.htm

found a list of chips
maybe of use