Casio College fx-80 : which IC hides inside ?

[Mark]

Hello,
I own a Casio fx-80 (yellow LCD) and for a long time I've been wondering about the chip it uses : all my searches were unsuccessful, which is surprising for such a popular machine.
Of course I could disassemble my calculator, but since it involves detaching the display, I wouldn't risk that : I don't have any spare.
Could it be that someone already revealed what circuit it is ? That makes me think about Jose Mesquita and his very nice disassembly articles :-)

Thank you !
Marc

[damaltor]

Casio used mostly specialty ICs, made just for that exact calculator model. Most of the fx series do only have a single IC in them which contains Keyboard scanner, Display driver, RAM, ROM and CPU. I dont have an fx80 but lots of other casios of that time, and most of them have nothing interesting inside them - just that only IC which is different for every calculator model.

[Mark]

Hi,
You're right and I do know that these ICs are specific. It's simply some kind of curiosity : was it from Hitachi, NEC or Toshiba* ? The fx-80 is special because the IC markings are not visible without dismantling it, which can be risky for the display. On many other calculators, you simply need to open the machine. Should I have a spare one, especially not working, I'd already have it totally disassembled.

My purpose is that sometimes you actually can tell which processor or microcontroller is at work ! A machine I recently got (desktop 4-banger with printer) has a Mitsubishi IC from the M50740 family : its core belongs to the well-known 6502 family.

* : the many other Casios I have are mainly NEC and Hitachi-based, very few are powered by a Toshiba processor. In the more recent ones, the chip-on-board technology makes the identification a highly destructive process. Did you take a look at http://zeptobars.com/en/ ?

[damaltor]

Well, to get to know what ic is in your calc without opening it, you might try the "calculator forensics":
http://www.rskey.org/~mwsebastian/miscprj/forensics.htm
if you follow the mentioned steps and compare your results to the table, you might get a hint by looking at other calcs with the same forensic result value, and look which chips hide inside those.

[Mark]

Thanks, Damaltor :-) but... I already saw that, ironically, the result (8.99116141) is shared by 2 different processors : The Hitachi HD36130 and the -probably twins- NEC D1897/1898G !
Maybe we've got here the same CPU core+firmware made by 2 of the major Japanese chipmakers. I remember having seen 2 other ones (again one from Hitachi, the other one from NEC) with the very same instruction set -and again in Casio machines.

These chipmakers later merged with a 3rd one (Mitsubishi) to become Renesas. The connections between them seem to have old roots. It also tells us that these calculators were sold in such quantities that a second source was needed to follow the market : production yields weren't as high in the seventies.

Honestly, it's not a fundamental question, just curiosity for fun. Anyway, the historical scope keeps its mysteries, even if many are unveiled by studying the patents : many companies didn't care about keeping accurate records of their activities, 30-40 years ago the calculator market evolved at a break-neck pace. But it already acted like the software industry : chip architectures had a surprisingly long lifetime... our beloved hp's are there to illustrate that.

In a few words : I simply have liked to know if someone had actually dismantled his/her fx-80 in order to answer the question :-)

[jebem]

Hi Mark,

Unfortunately I don't own this model, so I can't tell what processor is using.

Two remarks:

In those old times Casio used to release several different models sharing the same processor and firmware, only changing the keyboard labeling and sometimes using a different display.

I'm away from home and do not have access to my personal database, but from the top of my head I remember one example of it because I had bought one fx-39 that had hand written labels in the arithmetic operators.

I tried them and it seems that the fx-39 can perform the same functions as found on other models like the fx-120 for instance.
I'm checking a few internet images for the fx-120, and I remember to check the R->P and P->R functions on the fx-39 and it worked fine.

So, using the same thinking for the fx-80, maybe the next model, the fx-81, shares the same processor.
Looking to the pictures of the fx-81, and assuming the fx-80 is similar, it seems that we just need to undo those 5 screws on the PCB to remove it.
The LCD display should come attached to the PCB.
From what I know from Casio machines of that era, they are very well built and easy to service, so I would be surprised to learn that it would be different on the fx-80.

But it is your machine, and your decision to not open it that I respect.

Cheers,
Jose

[Mark]

Hi Jose,

I didn't know of that fx-81 ; it's a white fx-80 from which statistical functions and polar - rectangular conversions have been removed. The opened machine looks exactly the same as my fx-80 which was a gift from a colleague ; surprisingly it's not as easy to find in France. The 5 components you can see are the same, the ceramic capacitor is connected to the clock oscillator running approximately at 100 KHz.

When I got that calculator, the case was cracked because of the high constraints from the too strong battery contacts. I chose to replace the batteries with a CR2032 cell (socketed). I unmounted the PCB to replace the 2 wires that were deeply corroded by battery leak, then I discovered that the IC is almost completely hidden under the display.

My concern about not disassembling the display is that undoing the clips holding the LCD in place may weaken them once twisted back. I already had such a bad experience of a broken clip : the metal was corroded enough that it didn't survive a gentle handling.

Now you've got the whole story :-)

Regards,
Marc

EDIT : new episode ! Having looked once again at rskey.org, I was surprised to see that the HD36130 (supposedly reporting the same forensic result as the fx-80) is found in the fx-29. I pulled my fx-29 from my storage and a few seconds later got this result : 14.462524. I misinterpreted the "sorted by result" table. (have a look at it, you'll understand why)

Now we can tell that the NEC chips D1897G and D1898G are very likely the ones to be found in the fx-80 AND the fx-81 (or reversed). That's for the theory of course.....

[jebem]

Thank you for the additional information, Marc.
I agree with you that it can be very risky to remove the LCD.

I'm looking now for one of these models in my local auction sites.
Apparently these are kind of rare models when compared with older models like the fx-29 or 39.
I am trying to not order from USA due to customs on top of s&h costs, but I understand that would be much easier to find a good specimen from there.

[Mark]

I wish you good luck for your hunt, Jose !

[matthiaspaul]

(06-02-2016 05:47 PM)jebem Wrote:  In those old times Casio used to release several different models sharing the same processor and firmware, only changing the keyboard labeling and sometimes using a different display.

I can confirm this. I own a modified Casio fx-180P with hyperbolic functions retrofitted.

Comparing the functions of my fx-180P calculator with a friend's Casio calculator of the same time (I don't recall the model number any more), I realized that although both calculators looked completely different and had a rather different keyboard arrangement, the functional assignments for each of the keys were the same. My fx-180P lacked a "hyp" key and the corresponding six functions, but otherwise the documented technical data and "programming" features were identical.
This observation made me open the cases and when both were found using the same processor, reverse-engineer the PCB traces between the controller and the keyboard matrix, and add two wires to emulate the "missing" key. Shorting these wires, my calculator actually performed the hyperbolic functions, so I retrofitted a push button in a corner of the inclined plane at the top of the case (the only place leaving enough room inside to add the button).

Greetings,

Matthias

[EDIT: http://mycalcdb.free.fr/main.php?l=0&id=390
Not my site, but it nicely shows the internals of a fx-180P including the keyboard, processor and the available space next to the battery case, where I retrofitted the button in my unit.]

[Thomas Radtke]

(06-28-2016 06:30 PM)matthiaspaul Wrote:  I own a modified Casio fx-180P with hyperbolic functions retrofitted.

Comparing the functions of my fx-180P calculator with a friend's Casio calculator of the same time (I don't recall the model number any more), I realized that although both calculators looked completely different and had a rather different keyboard arrangement, the functional assignments for each of the keys were the same.

Presumably the fx-3600p.

Sold my TI-25 to get the 180p back then to become one of few students with a programmable calculator .

[matthiaspaul]

(06-28-2016 06:30 PM)matthiaspaul Wrote:  Comparing the functions of my fx-180P calculator with a friend's Casio calculator of the same time (I don't recall the model number any more),

Unfortunately, I don't recall the model number any more, but it was a "flat" calculator model with silver keyboard labelling and black case running on coin-cells in a side-flip protective etui. I seem to remember that it featured a 4-digit model number, like in "fx-nnnnP".

In contrast to this, the fx-180P came in a rather bulky dark-brown case with slightly bronze-colored keyboard labels, and it was powered by two AA batteries. It came in a textured rubber sleeve.

Greetings,

Matthias

[matthiaspaul]

(06-28-2016 06:48 PM)Thomas Radtke Wrote:  Presumably the fx-3600p.

Yeah, quite possible. Thanks!

Quote:Sold my TI-25 to get the 180p back then to become one of few students with a programmable calculator .

;-) Well, the "programming" was rather limited. I haven't programmed it in decades, but IIRC it had only a few dozen steps and the only conditional jump supported was back to the start of the program? Nevertheless, I remember, that I implemented greatest common divisor and lowest common denominator programs for it. ;-)

Greetings,

Matthias

[Thomas Radtke]

(06-28-2016 07:08 PM)matthiaspaul Wrote:  ;-) Well, the "programming" was rather limited. I haven't programmed it in decades, but IIRC it had only a few dozen steps and the only conditional jump supported was back to the start of the program? Nevertheless, I remember, that I implemented greatest common divisor and lowest common denominator programs for it. ;-)

A lot could be done by placing setup stuff in P1 and a loop in P2. I programmed it to calculate prices for flowershop things in various sizes for my girlfriend in the 90th. A full blown rule of proportion solver .