I collect here some notes about my (limited) experience about calculators manufacture date. Please,
feel free to write me and debate about my thesis.
First of all, let's define what is a calculator date: to me, the calculator date is the date of the final assembly,
not the date of the single components, not the date of the distribution
to dealers, not the date of the purchase in a shop.
Basically, we have two ways of establishing such date:
Decoding the date from the Integrated Circuit stamp
In this case we must ask ourselves:
what is the lag between the IC manufacture date and the calculator assembly? I take
for granted what assumed here by Emil Dudek (Vintage Technology maintainer):
the lag is between 3 and 7 weeks for his specific case study, so we may take an
weeks (rounded roughly to a month) to get the calculator date; of course this
method (precise as it can be) involves opening our beloved calculators,
and this is stuff for robust hearts. In any case, an
advice is necessary:
DON'T TAKE AS ABSOLUTE DATES RETRIEVED FROM IC STAMPS!
Here's what Emil himself wrote me in a February 17,
2007 email: "...my dating is based on the date of the main IC in the calculator.But these may have been hanging around for some years before being used, especially in the late 1970s, early 1980s. I do not know...".
Interpreting the Serial
Number (if it reports a date)
In this case we must ask ourselves: is this
date the calculator date? I think yes, provided it
doesn't conflict with the date retrieved from the first method,
since it prevails over the second. Another
advice is here necessary:
DON'T TAKE AS ABSOLUTE DATES RETRIEVED FROM SERIAL NUMBERS!
Serial Numbers are generally
established before manufacturing, and any modification in the
production line may alter the real meaning of the date hidden into the
Serial Number itself (if any). We must assume the calculator's date to be a range of days (even months) rather than a particular, specific date.
That said, we can finally assert that there's no certain data about a calculator date, or:
NO DATE IN DATA!
(so, one may wonder why I do write all this. Who knows? Who cares?)
Now, since IC stamps are treated here and here, we proceed to treat Serial Numbers. Here we go!
Hewlett-Packard calculators are among those that clearly report the manufacture date
in the Serial Number, through long-term demonstrated thesis (see, for
instance, the previous links or here).
The cut between the two formats is roughly 1996, but it's not neat, so you may meet
calculators made after 1996 reporting old pre-1996 Serial Numbers
or vice versa. In any case, their formats differ enough to understand what their interpretation must be.
The format of such Serial Number is:
YY is the year number (to be added to 1960); so, for instance 28 means 1960 + 28 = 1988.
is the week number; so, for instance, 46 means 46th week (about mid November).
C is the country code, where A is USA, B is Brazil, G is Germany, J is Japan,
S is Singapore, M is Malaysia (other countries I haven't heard of may be involved).
is the manufacture number that uniquely identifies a calculator (nobody knows if counting started from
1, 1000 or say 10000, so we cannot take this number to define the ith calculator).
My HP-11C has Serial Number 2846A86886, so I can state: model
HP-11C # 86886, produced around November 1988 in the USA.
My HP-20S has Serial Number 3215S00005, so I can state: model HP-20S # 5, produced around April 1992 in Singapore.
In the Italian market, USA pre-1996 calculators are much more common than others.
The format of such Serial Number is:
the country code, where US is USA, SG is Singapore, ID is
Indonesia, CN is China, MY is Malaysia; in calculators produced after
2004, a letter perhaps identifying the plant is associated,
as in CNA (meaning China, plant A).
Y is the year number, actually the last digit of the year; this forces us to know exactly the decade of the calculator (and the HPMuseum is a great help for this). So, for instance, 1 (one) means 2001, and 7 means both 1997 and 2007;
is the week number; so, for instance, 30 means 30th week (about end July)
is the manufacture number (again, we don't know if counting started
from 1, 1000, 10000 or else, so we cannot take this number to
define the ith calculator)
One of my HP-12Cs
has Serial Number CN13006020, so I can state: model HP-12C # 6020,
produced around July 2001 in China.
My HP-38G has Serial Number ID84901725, so I can state: model HP-38G # 1725, produced around December 1998 in Indonesia.
My HP-9G has Serial Number CNA62700068, so I can state: model
HP-9G # 68, produced around July 2006 in China, Plant A (I know it's
2006 because in 1996 the HP-9G was in the mind of God).
In the Italian market, there
aren't many USA post-1996 calculators,
while Indonesian and Chinese ones are much more common.
I was trying to decode Casio Serial Numbers when I had the brilliant
idea to write to Casio and let them explain. Unfortunately, after two
emails and half an year, here's what they answered: "...Thank
you for sending your e-mail regarding serial numbers of calculators.
Serial numbers of calculators are for the management of factories.
Unfortunately we are not allowed to inform individual customers of the
information as this is our company's rule....". That's all.
So, of all the different types of Casio Serial Numbers, I can only say:
a) information about them is not available from Casio
b) the Serial Numbers seem to hide quite nothing, or they seem too difficult to decode.
Anyway, for completeness' sake, they're analyzed as well.
All the following assumptions about Casio Serial Numbers are due to me and
me only. If there's any mistake, don't blame anyone but myself
(especially if you are a Casio employee), and don't hesitate to write me: I'll update this page.
This format belongs to 1970s machines. As far as I
know, there are
no clear assumptions that can be made upon this Serial Number.
It's probably only a unique number on its own. It
seems a variant
of the following cases, made up of 6 numbers only; here are some:
Casio 8E 198612 1970s (1973 purchase) Casio fx-1 206866 1972
No design that I can evince. And you?
Wide-7 old format
This format belongs to old 1960s and 1970s machines. Again,
no clear assumptions can be made upon this Serial Number. It's
probably only a unique number on its own. It's composed of 7 digits,
the first either a letter or a number, the others always
numbers. There's never a constant bound to the year, nor a number whose
value must be added to a constant (say 1960) to get the real year, nor
a digit in the same position reporting year's last digit.
list of some such Serial Numbers, associated with machines and years of introduction (mostly taken from Casio Decode System list):
Now, do you see a design, behind these numbers? Well, even taking in
account that introduction years are uncertain, I don't.
This format belongs to 1970s and 1980s machines. As far as I know, there
no clear assumptions that can be made upon this Serial Number. It's
probably only a unique number on its own and an extension/variant of the preceding format.
list of some such Serial Numbers, associated with machines and years of
introduction (partly taken from my collection and partly from Casio Decode System list):
As said, this latter may result in a particular case of the preceding Wide-7
The two classes might also be two different Serial Number formats. In
any case, I cannot see any design behind all this. Can you?
This format belongs to machines from 1980s to nowadays. Such
number is not really a Serial Number: this is
demonstrated by the fact that I have two Casio fx-180P, bought in the
same place, and having the same
number, 6B108A; and here's what Anton Thimet (Tony's Taschenrechner-Sammlung Calculator Collection) answered by email after my request of the Serial Number of his Casio fx-180P:
"I have the exact same number printed on my battery cover (6B108A)"; I bought my calculators in Italy, his was made for Germany
or France (basing on the manual), but the Serial Numbers are the same.
The first format we see, the 6-variant, is composed by 6 alphanumerics like this:
may be the year number, actually the last digit of the year; this forces us
to know exactly the decade of the calculator (The R/S site by V. Toth or the Voidware site are a great help for this). So, for instance, 2 means 1982, 1992 or 2002;
well, about this code place, I have a story for you. In collecting
serial numbers from Internet to set up a basis for study, I observed
that this M code rapresented 12 letters only, I though linked to the following list (be careful: it's wrong!): A January B February C March D April E May H June K July L August M September P October S November W December This seemed terse. But one day I bought a Casio MG-777,
with Serial Number 2F102A; F: the 13th month? So one thing is clear:
this definitely is NOT the month code. Probably it cannot
be disassociated from the following piece od code, constituting a whole identifying
(maybe in conjunction with M) is
some sort of manufacturer's code, being the last digit always a letter; it
may refer to a manufacturer or a country code, since both the fx-180Ps, the College fx-100 and the fx-3900Pv share the same identifying code, 108A.
My Casio fx-3600Pv
has Number 0D904A, so I can state: model Casio fx-3600Pv, # D904A,
produced in 1990; how do I know it's 1990 and not 1980 or 2000? Well, R/S has the answer.
My Casio PB-220 has Number 6D306A, so I can state:
model Casio PB-220, # D306A, produced in 1986; I remember
very well the PB series, and since it was during the 1980s, I know it's
Both my Casio fx-180P have identical Number 6B108A, so I can state for both: model Casio fx-180P, # D108A, produced in 1986; I know that the fx models with this chassis are from the 1980s, since I have used the Casio fx-100
College for all that decade, so I know it's 1986. Since I bought the
two calculators in the same shop, where they lay for twenty years, this
Another format, slightly different, is the following 8-variant which,
even if composed by 8 alphanumerics, can be associated with the
Year-6 format (my guess). Its only peculiarity is
that in the second position there's the letter "Q". This code is:
Y may be the year (as before);
Q is... who knows?
NNNNNN is... who knows?
My Casio fx-570MS
has Number 5Q10630A, so I can state: model Casio fx-570MS, # Q10630A,
produced in 2005; how do I know it's 2005 and not 1985 or 1995? Well, it's currently under production...
My Casio HR-8L has Number 8Q10606E, so I can state:
model Casio HR-8L, # Q10606E, produced in 1998; I
remember I bought this calculator in 1999, so this assumption holds.
My Casio fx-4800P
has Number 4Q11224A, so I can state:
model Casio fx-4800P, # Q11224A, produced in
2004; how do I know it's 2004? Well, it was introduced at the end
of the 1990s and was in production until few years ago.
Even if I'm quite sure my thesis about Y in the Year-6/8 format is realistic, this assumption needs confirmation!
Here are other brands discussions about dates hidden into Serial Numbers for the calculators I own.
Lexibook codes either miss or they don't contain any date code. So,
don't even try to look for them. Generally, instead, you should see the
year printed onto the calculator back, along with a Lot Number (useless for our purposes). Or you should try the IC way.
At present, I've heard about only one Tokatron calculator in the world (the one
I own), and this one lacks its Serial Number; so no assumption can be
Back to Main page