Japanese date codes

[Franc Zabkar]

I was recently looking through a datasheet for a Sharp IR3M02 PWM
regulator and found the following information on date codes.

date code example 226A = 1st week of June, 1982

2 2 6 A
| | | |--- production week A,B,C,D,E (1st week)
| | |----- production month (1,2,3 ... 9,X,Y,Z)
| |------- denotes last digit of year (1982)
|--------- denotes the I.D. No.

I don't know if other Japanese manufacturers of that time used the
same coding, but I thought the information may be useful in any case.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Geoffrey S. Mendelson]

Franc Zabkar wrote:

I don't know if other Japanese manufacturers of that time used the
same coding, but I thought the information may be useful in any case.

Often they used the Japanese system of dating, starting from the date
of coronation of the current emperor.

Geoff.


--
Geoffrey S. Mendelson, Jerusalem, Israel gsm@mendelson.com N3OWJ/4X1GM

 

[N_Cook]

Franc Zabkar <fzabkar@iinternode.on.net> wrote in message
news:a0l1k417lnb6eh09kdrrqorrsemcomqbkb@4ax.com...

I was recently looking through a datasheet for a Sharp IR3M02 PWM
regulator and found the following information on date codes.

date code example 226A = 1st week of June, 1982

2 2 6 A
| | | |--- production week A,B,C,D,E (1st week)
| | |----- production month (1,2,3 ... 9,X,Y,Z)
| |------- denotes last digit of year (1982)
|--------- denotes the I.D. No.

I don't know if other Japanese manufacturers of that time used the
same coding, but I thought the information may be useful in any case.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

To get the decade you have to look at the quality/font/style of the
lettering

I will add yor info to my tips files

Some hints for determining the age of equipment not marked on serial
plate,dating old kit.
The best is date codes on ICs. These are usually 3 numbers say 712 distinct
from the IC type number which is usuallly letters then numbers. For 712 12
is
the number of weeks (ie march) into the year ending 7 . But is it 1977 or
1987 etc. The
70s printed numbers are usually more angular as though printed in the fonts
available to
an XY plotter. Also the IC series number is usually indicative of the era.
For Sanyo ICs a number like 4C1 probably means 1984 from the '4'.
Larger (power handling)
package/hybrid ICs of the 70s were a lot less standardised than later. Other
indicators dark brown paxolin board would suggest 60s rather than 70s as
would
carbon resistors rather than metal oxide (no bulges at the ends of the
barrel).
Also the sizes of 60s and 70s resistors tend to be larger than what is
actually
required for heat dissipating. E-line transistors come in the 80s .
Places to find dates are sometimes large
electrolytic capacitors, polycarbonate capacitors,large
transformers,loudspeakers ,relays,
microswitches,motors , some
small rectangular radial lead caps ,body of air vane tuning caps and some
small transistors
have a datecode. Clear wrapped ribbon cable tends to be 70s, grey covered
ribbon
cables tend to be 80s and thin tape type with foil conductors tends to be
90s.
PCB traces from 60s onwards tend to get less angular and more curved by the
90s ,double
sided traces on boards tends to be 80s onwards.
The year of introduction of certain types of connectors
gives a lower bound. eg Scart/Peritel would not be found
earlier than 1977 and those would be in French kit,
the earliest date for mini-din would seem to be 1985.
The pitch of IC pins of course gets less than .1 inch only 80s onwards and
surface
mount tends to be 90s onwards. LEDs started coming in early 70s.
For Technics / Panasonic ICs of the 1970a date codes probably take the
form a single number then in smaller script an underlined
HEX digit. ie 9 plus small underlined D would mean 1979 14/16 ths into the
year.
After year 2000, eg 5007 presumably week 07 of 52 in year 2005
JRC IC maker seems to use this convention of reversing post-millenium
dates - do
others ? Unless you find 2 or more IC date codes from
different manufactirers agreeing it is often more a feel
or balance of probabilities for a date.
After about 2000 pcb overlays often have the date as distinctive slanted
"7 segment display " font numerals of 2 number year then 2 number week
as a 4 digit code as on iCs.
Date marks on mouldings and PCBs
Moulding example
Plastic moulding with raised print 82 83 84 85 with a small moulded
pip over both 82 and 83.
Then a 1 to 12 'clock face' both with an arrow next the figure 8.
A centre pop dimple in the mould above
82 in 1982 and then another over 83 in 1983. Leading to
raised small conical pips on the plastic. Probably means
8/12 year or August 1983
PCB example
White silkscreen printed
panel on the overlay of a PCB, presumably with date coding.
A left panel and right panel consisting of
1 to 6 along top and 5 dots in each column so 6 x 5 dots
Right panel marked 7 to 12 and similar array of dots
but the third one down under the 7 was missing.
Probably third week of 5 weeks of July.
Central panel had marked like this
02
??
??
05
The queries were probably the remnant tops of numbers from scrubbed out
03 and 04 so probably like in the plastic case 03 is scrubbed
in 2003 and 04 scrubbed in 2004 but still leaving the 03 mark.
So probable screen printing date of mid July 2004


--
Diverse Devices, Southampton, England
electronic hints and repair briefs , schematics/manuals list on
http://home.graffiti.net/diverse:graffiti.net/

 

[Franc Zabkar]

On Thu, 11 Dec 2008 10:21:59 -0000, "N_Cook" <diverse@tcp.co.uk> put
finger to keyboard and composed:

Some hints for determining the age of equipment not marked on serial
plate,dating old kit.

Thanks for the tips.

Just as an exercise, I'm listing the codes on various ICs and
transistors that came out of an old Pye 3-in-1
tuner/cassette/turntable.

Sharp IR2E02 75YC 5 = 1985, Y = November, C = week 3
Toshiba TD6301AP 8720D 1987, week 20
Toshiba TC9147BP 8721H 1987, week 21
Toshiba TD6104P 8711K 1987, week 11
NEC uPC1225H 9241F - repair?
NEC uPC1225H 706D 1987, week 06 ?
NEC uPC1161C3 8713E 1987, week 13
NEC UPC1228HA 8728M 1987, week 28
Hitachi HA12045 6M4 6 = 1986, M = December, 4 = week 4
Rohm BA6124 706 170 1987, week/month 06 ?
Rohm BA3302 507 852 1985, week/month 07 ?
Rohm BA4558 701 535 1987, week/month 01 ?
Rohm BA4558 705 051 1987, week/month 05 ?
Sanyo 2SD313 E7E3 1987, E = May, week 3 ?
Sanyo 2SD313 E7F2 1987, F = June, week 2 ?

Here are more IC codes from a National VCR:

NEC uPD75108 8802EX705 1988, week 02 of year
NEC uPD4066 8734BN 1987, week 34
NEC uPC324C 8804C 1988, week 04
NEC uPC393C 8745C 1987, week 45
NEC 78M05 8802J 1988, week 02
Rohm BA15218 712 618A 1987, December (12th month) ?
Rohm BA15218 712 286A 1987, December ?
Rohm BA6246N 712 513 1987, December ?
Sanyo LA7308 8A5 1988, A = January, week 5 of month
Sanyo LA7098 7M3 1987, M = December, week 3 of month
Hitachi HA17805 7H4 1987, H = August, week 4 of month
Matsushita AN3215K 7D.5 (D is underscored)
Matsushita AN3321K 7D.5 (D is underscored)
Matsushita MN1451 41 7 (no underscore)
Matsushita MN1451 39 6 (9 is underscored)
Mitsubishi 5218 7Y02 1987, November 02 ?
Mitsubishi M5222L 7907 1987, September 07 ?
JRC 78M09A 7012 December 1987 ???

This page has some useful info (esp. for Hitachi):
http://www.datamath.org/Datecodes.htm

I can see the patterns for Sharp, NEC, Toshiba, Sanyo, and Hitachi,
and I can guess at the patterns for Rohm and Mitsubishi, but the rest
are a puzzle. I see that in your notes you have suggested that the JRC
year digits may be reversed, but this doesn't tally with the 78M09A
part. I see what you are suggesting for the Matsushita codes, but then
how do you account for the MN1451BVK parts? I tried Geoffrey's
suggestion regarding Hirohito's coronation year (1926?), but I
couldn't make it fit the data.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Geoffrey S. Mendelson]

Franc Zabkar wrote:

I can see the patterns for Sharp, NEC, Toshiba, Sanyo, and Hitachi,
and I can guess at the patterns for Rohm and Mitsubishi, but the rest
are a puzzle. I see that in your notes you have suggested that the JRC
year digits may be reversed, but this doesn't tally with the 78M09A
part. I see what you are suggesting for the Matsushita codes, but then
how do you account for the MN1451BVK parts? I tried Geoffrey's
suggestion regarding Hirohito's coronation year (1926?), but I
couldn't make it fit the data.

Hirohito was crowned Dec 25, 1926, but the new one was crowned 7 January 1989.

Some parts made in early 1988 were date coded with a year of 62 (they
start at year 1 I think), and after 25 December 1988 were dated year 63.
Since it is considered unforgivably rude to prepare for the death of the
Emperor, things made in early 1989 were still date coded 63 for a short
time. It's also rude to continue to use the old dating system, so some
companies stopped production until they could change the dates.

Very likley then things produced went from a year of (6)3 to 1.

Geoff



--
Geoffrey S. Mendelson, Jerusalem, Israel gsm@mendelson.com N3OWJ/4X1GM