CASIO fx-5200P Scientific Computer teardown

[jebem]

... and reassemble to fully working order.

I bought this machine from a private seller that enjoys Casio pocket computers.
This one was not working, so I got it for a few Euros.
Opening, dismantling, cleaning and readjusting battery contacts was enough to bring it back to life.

Despite the had life this specimen had, all the keys work flawlessly. This is a Casio know how. They designed these machines to last forever. I know the keys do not have mechanical feedback like in most HP models but it gets the job done.

This looks like another Casio scientific calculator but it is programmable in BASIC language as used in other Casio pocket computers of the same era (mid 80's)..

It has a limited RAM memory size of just 1KByte but it can be expanded with an additional 1KByte by inserting a RAM Expansion Pack in the back.




After inserting two new CR-2032 button cell batteries, a hardware reset is mandatory to initialize the system correctly.




On a machine without additional RAM Expansion Pack installed, typing DEFM and pressing EXE returns the default RAM memory partitioning after a hardware reset is:
- 26 variables
- 544 program steps


The RAM memory partitioning can be modified as required using the command DEFM.
Each program step takes one Byte of RAM.
Each variable takes 8 program steps consuming 8 Byte.

The above default partitioning results in a total of ((26 * 8) + 544) = 752 Byte.
Therefore the internal 1KByte RAM memory have (1024 - 752) = 272 Byte that are not available for the end user.
These 272 Byte would be equivalent to 34 internal variables (registers).

The RAM Expansion Pack would add up to 1024 program steps or 128 variables, depending of the user needs.

Going into programming mode (WRT / Mode 1), it shows 544 remaining program steps after a hardware reset.




The LCD display showing all annunciators




Power supply information in the back cover: 6VDC, 0.01Watt.
This translates to an average current consumption of 1.7mA.



(continued)

[jebem]

Machine teardown.

Started by undoing the four screws securing the alpha keyboard to the main body.




Then removed the aluminum back cover (not shown).
The two zebra connectors and the common ground spring coil interconnects the alpha keyboard to the main machine body.




And the ground spring coil for the back cover.
Both miniature metal spring coils were missing. I had to order them from a UK seller.



Undoing the screws securing the plastic frame gave access to the main PCA.



Lifting and rotating the main PCA reveals the keyboard assembly.
The LCD display is glued to the front cover and needs to be slightly heated to be removed without damaging the LCD unit.



Back aluminum cover and power switch among other parts.



(Continued)

[jebem]

The three flex connectors from the alpha keyboard.



Details of the zebra connectors to link the alpha keyboard to the main PCA.




The keyboard membranes have embedded hard plastic keys.



PCA back side.




PCB top side (main component side).
- Hitachi HD61747B16 - SoC processor (CPU, LCD & KB driver, ROM)
- Hitachi HD61914C- 1KByte static RAM

This fx-5200P uses the same chip set as found on other Casio machines, only differing on the ROM firmware. For instance:

- fx-4000P
SoC processor: Hitachi HD61747B38
Memory: HD61914 8kbit static RAM

- FC-200
SoC processor: HD61747B55
Memory: HD61914 1KByte static RAM.




General parts view.

[Chasfield]

Great tear down. The keyboard membrane construction is interesting. I am thinking that after losing sales with the squishy key tops of the Casio fx-7000G era, they thought they could do better by hardening the key tops yet stay cheap using one-piece membranes.

As a youth, my two biggest un-boxing let downs were the Texas TI57 and the Casio fx-7000G. The infra-red and ultra-violet of keyboard feel. In both cases I immediately felt that I had wasted my money.

[jebem]

(11-07-2016 04:34 PM)Chasfield Wrote:  As a youth, my two biggest un-boxing let downs were the Texas TI57 and the Casio fx-7000G. The infra-red and ultra-violet of keyboard feel. In both cases I immediately felt that I had wasted my money.

Thanks for your comments, Chasfield.
Interesting view on the keyboard operation!
The 57 gives a mechanical feedback as long as a strong force is used when pressing the keys (is this the reason why people says Texas is on the "dark" force? ) while the Casio's are absolutely feedback less and way to smooth requiring near zero force applied.

I guess HP was in the sweet spot taking the green spectrum region with their magnificent Classic series.

[Didier Lachieze]

Thanks for this detailed tear down. As always your pictures are beautiful.

(11-06-2016 06:58 PM)jebem Wrote:  And the ground spring coil for the back cover.
Both miniature metal spring coils were missing. I had to order them from a UK seller.

How did you find replacement parts ?

[jebem]

(11-11-2016 08:31 AM)Didier Lachieze Wrote:  

(11-06-2016 06:58 PM)jebem Wrote:  And the ground spring coil for the back cover.
Both miniature metal spring coils were missing. I had to order them from a UK seller.

How did you find replacement parts ?

Hi Didier, Thanks for your kind words.

I couldn't find any original parts supplier in the Internet and considering its vintage status I didn't even tried to call our local Casio distributor (despite the excellent service & spare parts they can offer here; once I needed a metal battery cover for one of my Casio fx 2.0 and they kindly and swiftly mailed the components for free).

So I had a look in the TAS UK and found a supplier for miniature metal coil springs here.
I made a mistake and ordered smaller diameter springs than ideal, but as these springs will not have mechanical movement because it are used to interconnect static metal components, it should be fine.

[damaltor]

Those are beautiful devices, i have quite a lot of the series. One of the follow-ups of your machine was the PB-220 ("Personal Computer"), a basic-capable calculator with a similar membrane keypad. as far as i remember it has 2k integrated ram, and accepts a single RP-8 extension which has another 8k for 10k total. the tokenized basic needs about one byte for most keywords.

the devices last more or less forever, the keys dont have the nice hp feeling but work precisely. the soft press makes number crunching fairly quick.

I made a usb-interface for the series a while ago, which pretends to be a cassette recorder and allows to SAVE and LOAD programs to the PC, so they can be edited on the computer instead of the tiny screen.

[jebem]

(11-14-2016 10:32 AM)damaltor Wrote:  I made a usb-interface for the series a while ago, which pretends to be a cassette recorder and allows to SAVE and LOAD programs to the PC, so they can be edited on the computer instead of the tiny screen.

Excellent ideia and great effort. Marcus as developped a lot of cassette utilities as well.

Is this project yours?

[damaltor]

Yes, it is. The interface is compatible to Marcus' tools, so you can download the programs from the calculator to the pc, decompile them with Marcus' tools, edit the program or write a new one, compile it back and then send it back to the calculator. Lots of pre-work was done by Piotr Piatek though, so i did not have to dig extremely deep to understand the casette interface.

[foroplus]

Wow. According to the dates printed on the chips and the back metal cover, your calculator is dated as of 1992, while mine is on 1985.
That means this machine was a bunch of years on the market.

[jebem]

(06-11-2019 07:31 PM)foroplus Wrote:  Wow. According to the dates printed on the chips and the back metal cover, your calculator is dated as of 1992, while mine is on 1985.
That means this machine was a bunch of years on the market.

That was the reason I didn't put a manufacturing date on this specimen as I usually do after dismantling my machines.

I am not sure about the Japanese date codes. They are not well known and documented imho.

Casio was well known for releasing new models every single year, meaning that each model would have relatively short production lifetime, maybe 2 or 3 years.

1985 to 1986 is the mentioned production years for this model in the usual internet portals, but no one knows for sure.
Probably not even Casio would know by now.

In my calculator I am sure the chips are the original ones, as there is no marks of resoldering anywhere.
So a lifespan of 7 years in production seems improbable for Casio, but we never know.

[foroplus]

(06-12-2019 07:37 AM)jebem Wrote:  

(06-11-2019 07:31 PM)foroplus Wrote:  Wow. According to the dates printed on the chips and the back metal cover, your calculator is dated as of 1992, while mine is on 1985.
That means this machine was a bunch of years on the market.

That was the reason I didn't put a manufacturing date on this specimen as I usually do after dismantling my machines.

I am not sure about the Japanese date codes. They are not well known and documented imho.

Casio was well known for releasing new models every single year, meaning that each model would have relatively short production lifetime, maybe 2 or 3 years.

1985 to 1986 is the mentioned production years for this model in the usual internet portals, but no one knows for sure.
Probably not even Casio would know by now.

In my calculator I am sure the chips are the original ones, as there is no marks of resoldering anywhere.
So a lifespan of 7 years in production seems improbable for Casio, but we never know.

Yes, it's very weird a lifespan of seven years (or who knows if more); but has to be legit since the date is on the very chips and no doubt about that.

Mine is
Both chips were made on November 85.

By the way, have you tried to overclock any of these calculators that have the processor HD61747/HD61913/HD61917?. They have an internal clock which speed can be modified depending on resistor soldered to pins 25 and 26.
Some calculators have a 68K Ohms resistors (this one), and others, like the PB100 a 56K.
Seems to be that if you solder another resistor in parallel with that one, the speed of the clock is increased an thus the performance of the machine (and so the energy draining). A 30K resistor will double the speed, and a 3,3K will triple it.

[jebem]

(06-12-2019 11:25 AM)foroplus Wrote:  By the way, have you tried to overclock any of these calculators that have the processor HD61747/HD61913/HD61917?. They have an internal clock which speed can be modified depending on resistor soldered to pins 25 and 26.
Some calculators have a 68K Ohms resistors (this one), and others, like the PB100 a 56K.
Seems to be that if you solder another resistor in parallel with that one, the speed of the clock is increased an thus the performance of the machine (and so the energy draining). A 30K resistor will double the speed, and a 3,3K will triple it.

Yes, the oscillator frequency was trimmed at the factory for specific frequencies and we could change it.
However I am not into overclocking things anymore these days, I would say that I am into under-clock machines to save energy
As a collector, I realized that I should leave the machines in the original state as much as possible. Same for my radio collection, I don't even replace defective caps anymore as that would result in a non original radio.

I discovered your nice web site here:
Foro Plus
Lots of manuals, tips, emulators and more.

[foroplus]

(06-13-2019 08:57 AM)jebem Wrote:  

(06-12-2019 11:25 AM)foroplus Wrote:  By the way, have you tried to overclock any of these calculators that have the processor HD61747/HD61913/HD61917?. They have an internal clock which speed can be modified depending on resistor soldered to pins 25 and 26.
Some calculators have a 68K Ohms resistors (this one), and others, like the PB100 a 56K.
Seems to be that if you solder another resistor in parallel with that one, the speed of the clock is increased an thus the performance of the machine (and so the energy draining). A 30K resistor will double the speed, and a 3,3K will triple it.

Yes, the oscillator frequency was trimmed at the factory for specific frequencies and we could change it.
However I am not into overclocking things anymore these days, I would say that I am into under-clock machines to save energy
As a collector, I realized that I should leave the machines in the original state as much as possible. Same for my radio collection, I don't even replace defective caps anymore as that would result in a non original radio.

I discovered your nice web site here:
Foro Plus
Lots of manuals, tips, emulators and more.

Yes José, you're right. These are the good criteria for a collector. And I share them.

But sometimes I feel the need to deep into the guts of these machines. Describing only the surface is not enough for me, and I'm saying that, knowing that I am only an amateur without the necessary tech knowledge. But you know, it's a hobby, something to spend the time.
I know that programming Z80 machine code of a TI-83 has no sense, but I like it. All these things, both hardware or software, relaxes me. They recall me my youthhood.

I am a fun of your teardowns. Never miss one.

[Eddie W. Shore]

Why would they create BASIC programming machines with such little memory? I can understand if we are dealing with keystroke.

[jebem]

(06-14-2019 05:21 PM)Eddie W. Shore Wrote:  Why would they create BASIC programming machines with such little memory? I can understand if we are dealing with keystroke.

That is true. Many Casio models of the 80's could have a memory expansion installed to mitigate that limitation.
Also some memory saving came from the way the commands and functions are stored, if I recall it correctly: each one just consumes 1 byte; a line number would take 2 bytes; variable names 1 byte per char; etc.
The big consumers are the variables containers with 8 byte each.

[Dave Britten]

(06-14-2019 09:16 PM)jebem Wrote:  

(06-14-2019 05:21 PM)Eddie W. Shore Wrote:  Why would they create BASIC programming machines with such little memory? I can understand if we are dealing with keystroke.

That is true. Many Casio models of the 80's could have a memory expansion installed to mitigate that limitation.
Also some memory saving came from the way the commands and functions are stored, if I recall it correctly: each one just consumes 1 byte; a line number would take 2 bytes; variable names 1 byte per char; etc.
The big consumers are the variables containers with 8 byte each.

Yeah, the Casios tokenize their BASIC code, but you still lose a couple bytes per line for the line number, and BASIC tends to be a little more "wordy" than the simpler calculator programming language. So 1 KB in the fx-5200P would probably get you about as far as the 422 bytes in the fx-7000G. I think 2 KB is a more realistic minimum where the extra overhead of BASIC starts to be worth the bother.

The PB-100 also has only 1 KB of memory, and can accept the same 1 KB add-on module as the fx-5200P.

And interestingly, the fx-720P has NO internal memory, meaning you can have multiple 2 KB or 4 KB battery-backed memory modules and swap between them for different sets of programs.

[Gamo]

Very nice Tear-down images.

I just acquired a Casio FC-100 (No Programmable feature)
This is a Financial Consultant calculator.

When I got this the display was all dark and very hard to read numbers.
After watching YouTube on how to fix dark spot on calculator display I
decided to fix this and all go well.

As shown in this attachment the NPV key is upside down and
the display shown all digits with number 8 to show all segment is working.

[attachment=7381]

Remark: The upside down NPV key was like that when I got it, forgot to put it
the right way when I fix the display. I just took time and carefully fix the display
that I totally forgot to change that NPV key.


Gamo

[jebem]

(06-15-2019 06:53 AM)Gamo Wrote:  Very nice Tear-down images.

I just acquired a Casio FC-100 (No Programmable feature)

Thanks for reading my tear downs posts.
I feel the need to explore the specific electronics implementations found in equipments, due to my background in that field.

I believe i saw that FC machine yesterday on facebook HP club here:
https://m.facebook.com/groups/1185231308...oup_browse
Is it yours?
I am sure i have a couple of them too somewhere in the house.
Nice machines from Casio, well built and packed with features.