Bought 5 old window EPROMs from a vendor. Same Mfr, Part #, DC. Different IC!

[JoeN]

Did I mention it was an eBay vendor?     My ChipMax 2 says they are perfectly good, though.  Did ST really make two different dies and use them within the same date code?  The smaller one I think is just a shrink of the larger one, or is it the next size up in that EPROM series and they just tie an address line low and use half the chip to make use of a surplus of the larger sized die?  Or is this the usual Chinese eBay vendor game? 

Click link for a high-res version.

[bktemp]

Both EPROMs look suspicious:
The ST datasheet from 2000 says "NOT FOR NEW DESIGN". Therefore I doubt they were still in production in 2008, but maybe I am wrong, because you can find many pictures showing datecodes 07/08.
Either they are all fake (most likely), or somebody started production again. You can even find 27C32A with datecode 08 and 21V Vpp: 
http://www.digchip.hk/stock-M27C32A-20F1.html

I looked through my pile of old EPROMs and found similar looking ones: They have datecodes from the late 80s. The smaller die is maybe a 12V Vpp version (but there wasn't a 12V version for 2732, at least I couldn't find any datasheets from ST). Maybe your one with a smaller die is from the mid 90s.
So I would say both EPROMs are genuine SGS/ST (probabaly used ones), but have been relabeled to a more recent datecode.

Both EPROMs look similar, although the smaller one is a 2764 instead of 2732, therefore the die is a bit larger, but it also has the groove around the edge of the ceramic case.



By the way, does anybody know why there is a dot next to the die? I also have EPROMs made by Intel, AMD and TI having a similar device next to the actual die. It looks like a diode conected between substrate and Vss pin.

[Kjelt]

2732 in 2008 ? That would be strange , and the die size to the left is really a no go. Already in 1998 the dies of the 256 and 512 chips were smaller than those.
What they sometimes did is place larger eproms and not connect all address lines or connect those to ground and sell as smaller eprom.

[mikeselectricstuff]

EPROMs of the same part number changed dies over the years as processes shrank.
http://electricstuff.co.uk/shrinkingchip.html

[stj]

typical chinese.

i have a bunch of eproms with yellow print - dead givaway there.
they work, but the programmer is reading an unknown MFR i.d. out of them!!

it's just the price you pay if you want 10 for £10 instead of paying £7-14 for one.

[CJay]

typical chinese.

I have a bunch of eproms with yellow print - dead givaway there.
they work, but the programmer is reading an unknown MFR i.d. out of them!!

it's just the price you pay if you want 10 for £10 instead of paying £7-14 for one.

Not a guarantee you've not got a bunch of remarked chips but...

 I've got original EPROMs from years ago with yellow, blue and white markings, it's not necessarily an indication that they're fake on vintage parts and no, it's not age related yellowing or bluing 

MFR ID was not implemented until the EPROM market was 'mature' so it's entirely possible to find old chips that won't ID.

And for the OP, it's not unusual to find that sort of thing, when manufacturers moved to new processes in their fabs (or even opened up second fabs that ran alongside the first) it was, and I would assume still, common to have the same parts with different sized dies,

What you will probably find is the access speed for the smaller die is much faster even though it's marked up as the same part, the spec indicated in the part number is a guaranteed speed the chip will work at over its specified temperature and electrical range, so as long as it meets the datasheet requirements it can be 150nS, 250nS, whatever speed.
 

[stj]

lol, they are big cmos eproms - i.d. is fully implemented.

one other thing, is it just me - or does china only have eproms made by ST!!!! 

[wraper]

Been there. Exactly same ICs from one seller ordered with a 1 week difference, they also "were made in 2008". Exactly the same markings. One of those was just remarked ST but same die, dunno about speed grade. Second type was recognized by the die as AM2732D with 25V (or maybe 21,5) PGM  IIRC.

[CJay]

lol, they are big cmos eproms - i.d. is fully implemented.

one other thing, is it just me - or does china only have eproms made by ST!!!! 

Maybe they could only afford the silkscreen for ST?

 

[JoeN]

Well, they work.  It's only for a small personal project that is going nowhere except my desk.  So I appreciate that I got them cheap.  I just thought it was interesting to see two different dies from identically marked packages.