New Stuff
Guest book


For Sale
Hit or Miss
Metal Puzzles


Calculators: Dating Your Calculator
No, not going to the restaurant with it - finding out when it was made!  Specific to the 1970-1980

Trying to piece together the timeline of your calculator collection is a very important part of your hobby. Whether you collect a particular brand like Sinclair or HP, a particular era such as mid 1970s red LED, dating your calculator is full of problems and ambiguities.   Hopefully this page will help you pin down at least the rough date of availability.  Throughout this web site, dates are quoted as earliest actually available as best as I can gauge.  If I'm in doubt then I do not enter a date at all.  Before we start let's just remind ourselves of the issues:


Calculators were available for more than one year when issued
Excess stock of models were sold off to discount houses up to a few years later
Certain date ID marks can either be old component stock or there can be a large difference between manufacture date and street date
Certain models were revised for several years - but look identical
There will be variances on international availability of products


Basic Methods
Here are the main ways to date you calculator or at least verify the rough time frame of availability.  These are analysed in more detail in sections below.


Date ID marks for components such as ICs, displays and QA stickers
Dates shown in warranty, guarantee or receipt documents
Dates matched through press advertisements or catalogues
The type of component: such as the display type
The power supply requirements (9V, 6V, 3V etc)
The component count
The logic methodology (except legacy-RPN calculators) and bugs / artefacts
Size and keyboard technology
model name / number
Someone you know who bought this model originally swears by the date that they obtained it - highly unreliable ;-)

IC Date Codes (4 digit numeric)

The easiest way to start and date you calculator is to open it up and examine the main cpu chip.

Here you can see a typical Rockwell IC that has three numbers:
Row 1: A batch manufacturing code
Row 2: The IC identification number A1030PE
Row 3: The four digit date code

7423 is week 23 of 1974

Four digit date codes were very common but sometimes reversed.  Be careful with MOS ICs (found in Commodore calculators) as the numbers could be something like

MCS 7529 009 2676

Where the date code in this case is week 26 of 1976, the "75" number is the IC type.


Rockwell ICs had a large pin count (this one is 42 compared to the average 28) as they did not concatenate the keyboard matrix with the multiplexed display.  They also invented a strange staggered dual inline (DIL) pin approach which was not adopted by anyone else.  I can only think that they thought this made making the PCBs easier.

Notice too that this device has a plastic case with a metal plate sealed hole.  This was very expensive and generally outdated by the fully encapsulated plastic package.

Specific corner indicators like the blue resin blob (more usually a small indent) are there to indicate "Pin 1".

I generally use the IC date as the main date of the calculator.  In the case of there being more than one date coded item, I use the latest date.  This assumption can be open to some discussion over the time lag between date of manufacture of the IC, date of assembly, date of availability in the shops and finally, the date of sale.
However, I believe, at least in the height of the calculator boom years, that this lag is rather short.  I have a Rockwell 24RD-II that has an IC date code of week 45 of 1976 which has a warranty with a retailer's date stamp of December 1976.  This makes that maximum lag of seven weeks and a minimum of three: not very much!

GICO Sticker

It appears that GICO was a Japanese company that made and assembled keyboards.  If your calculator has one you will see GICO reference sculptured onto the PCB and if you are lucky there will be a readable QA sticker like this one.

The middle row is a date code whilst the lower row is either a batch number or test inspector ID.  Anyone with any information (especially what GICO stood for) would be most helpful. 

To interpret the Japanese date code the number is in the form:


and the year number is:

50 = 1975
51 = 1976
52 = 1977
53 = 1978 etc.

So this one is 29th March 1976.

Display Date Codes (2 digit alphanumeric)

VFD date

If your calculator has a Futaba Vacuum Fluorescent Display then you will see a date code on the underside - if accessible.  The first row number is the device ID: this one starts with a nine as it is a nine digit display unit.  The second row number is the two digit Japanese date code - in this case "6G"

From experience this gives a rougher date than the main IC as it appears that a wider stock was held: I have seen display date codes up to a year earlier than the main IC.

To interpret the Japanese two digit date code the number is in the form:

197Y.M where the month is:

A= January E= May J= September
B= February F= June K= October
C= March G= July L= November
D= April H= August M= December

So this one is July 1976.  Sharp may have used the "I"

IC/Transistor Date Codes (2/3/4 digit alphanumeric)

IC date

A lot of the Japanese companies (except NEC) used a similar date code system to the Futaba display above.  The core of the system is a two-digit alphanumeric year and month system.  However, as time progressed this went to a third digit and even a fourth.  

 The form could be:

3C 197{digit}.{month letter}
3E4 197{digit}.{month letter}.{week number}
6G14 197{digit}.{month letter}.{day number}
5K5 197{digit}.{month letter}.{batch number}

Which sometimes makes it guesswork which one it is, but the first two are the most important ones.

To interpret the Japanese two digit date use the table in the Futaba section:

So this one is March 1973

The lower transistor is a Toshiba T3086 with a date code of 3-G: July 1973.


The upper IC is a 16 pin DIL (dual in line) 0.3" width package with pin one denoted by an indent in the plastic.  This (along with the wider 0.6" forms) was generally adopted for the next 15 years.  It was superseded by surface-mount edge connection and pin-grid array systems.  In later calculators, and notably Novus (et al) they would dispense with the package altogether and mount the chip in resin on board.