5 Feb
2012
5 Feb
'12
1:59 a.m.
I have a come into an opportunity which people here might help with.
The opportunity is a 75 to 100# box of Eproms. So far w/o any major
skimming yielded 2732's 2764's 27128's 27512's and the 1mb 27 series
(don't recall the # right now)>
I also found a couple of parts that traced back to an HP inhouse 27128
part from AMD.
These were recently pulled from a warehouse full of printer material,
and are were just packed loose in a 12" x 12" x 12" box for gold scrap.
Since there was not manpower or time to sort, this what I got. There
are other boxes of NOS which I will deal with later.
Any ideas on what to do, what the chances are that they will work? I
don't have the means or time to do testing, though recent discussion
suggest there might be some cheap eprom programmer which might do that.
I have never blown up a prom from normal handling, w/o antistat but with
the shear number of parts here, the odds are that some are blown. I am
handling these now with antistat, but the prior situation is what it is.
Should I sort them by part, list them with a "if it doesn't work I'll
send you another", sell them as untested pull lots (I will sort and tube
them by type a and PN as my contribution to this project), or some other
way?
I'm guessing 10000 or more parts, don't know for now. also there could
be other parts at lower layers, for now w/o pulling them and risk
physical or electrical damage all I see are eproms and some junk pals
which will go back into the stew for gold recycle.
thanks
Jim
5 Feb
5 Feb
10:51 a.m.
On Feb 5, 2012, at 4:59 AM, jim s wrote:
I have a come into an opportunity which people here
might help with. The opportunity is a 75 to 100# box of Eproms. So far w/o any major
skimming yielded 2732's 2764's 27128's 27512's and the 1mb 27 series
(don't recall the # right now)>
I could certainly use a few of those.
Any ideas on what to do, what the chances are that
they will work? I don't have the means or time to do testing, though recent
discussion suggest there might be some cheap eprom programmer which might do that. I have
never blown up a prom from normal handling, w/o antistat but with the shear number of
parts here, the odds are that some are blown. I am handling these now with antistat, but
the prior situation is what it is.
I dunno, the 27xx parts (especially the non-C parts) are pretty tough. I've never
killed one, and I've put them through a lot of abuse (including plugging the power
supply in backwards on a ROM board; I could actually see the die glowing underneath the
sticker).
Should I sort them by part, list them with a "if
it doesn't work I'll send you another", sell them as untested pull lots (I
will sort and tube them by type a and PN as my contribution to this project), or some
other way?
I'd buy a tube or two as an untested tube lot, if the price is right. Mixed or
otherwise. I've got uses for all the sizes you mentioned.
A cheap EPROM programmer should do the trick for testing them, but it would take ages (if
memory serves, the UV erase time for those is at least a few minutes). Otherwise, just
programming in a testable pattern (say, a PRBS) should be sufficient to see if
anything's obviously blown, though individual bit errors would probably be something
you'd want to leave up to the buyer.
I'm guessing 10000 or more parts, don't know
for now. also there could be other parts at lower layers, for now w/o pulling them and
risk physical or electrical damage all I see are eproms and some junk pals which will go
back into the stew for gold recycle.
Yeah, the PALs are OTP, so no use there. I'm interested in hearing what else you
find; feel free to ping me off list (I suspect there will be many more here along the same
lines).
- Dave
10:57 a.m.
On 5 Feb 2012 at 1:59, jim s wrote:
Any ideas on what to do, what the chances are that
they will work?
The chances are good that they'll work--EPROMs are pretty rugged.
Sorting them out is an interesting problem. Obviously, the ones with
missing leads, or cracked packages/windows are junk--so the first
thing would be a cosmetic sort--and you can sort by type in the same
pass. So now you have several piles--junk, and then by type.
Next would be to sort the types out--try reading one. Since they're
used, if they come up with some data, you can put those in the
"probably good" pile for each type. If they come up FFs, they might
be erased or simply bad. If they come up 00 or all the same pattern,
but not FF, those are probably bad.
Finally, erase the lot (unless you want to save contents), erase and
attempt a program and verify. Those that pass would be the "good"
ones.
Alternatively, you have to balance what your time is worth against
what you expect to make. You might just bag bunches up in anti-
static bags and do the old Poly-Paks thing--some mignt be good, some
not; it's up to the buyer to determine which. Before you do that, I
think it would be a fine idea to pull a couple of sample bags and see
what the yield is.
The sure-fire way to kill most EPROMs is to install them upside-down
and let electricity take its course.
But as I said, UVEPROMs are pretty rugged things.
--Chuck
11:47 a.m.
Next would be to sort the types out--try reading one.
Since they're
used, if they come up with some data, you can put those in the
"probably good" pile for each type. If they come up FFs, they might
be erased or simply bad. If they come up 00 or all the same pattern,
but not FF, those are probably bad.
A 'santiy test' I use when I'm dumping the contents of a ROM or PROM for
backup or simialr purposes goes like this :
1) Make usre every (used) data line is high sometimes nad low someimes.
Finding bout 00 and FF in the dump is a good start. Acutally, I try to
make sure that one data line is not exactly the same as another in all
cases.
2) Make usre all (used) address lines do something .Make sure there's an
even address such that the immediately-following odd address has
differnet contents (meaning A0 is significant), THen make usre there's an
address with A1 low such that the same address but with A1 high has
different contents, and so on/ Finally make sure the high and low halves
of the IC are differnet ;-).
The problem with these tests for unknown ICs is that there;s no way of
knowing which address and data lines were used in the original
application. For example, if the EPROM was a character generator for a
7-wire printhead, it would be reasoanble for one of the data liens ot be
either 0 or 1 all the time, it wouldn't have been used in the original
application. Of if an EPROM was used as a replaemetn for a smaller drvice
(say a 27128 used in place of a 2764), it would not be uncommon to
program the same data into the high and low halves of the IC.
But if the IC contains data (not all FFs) and passes those tests, it's
likely to be good. Most failures I've foudn relate to the circuitry
conencted to the pisn, so that, for example, an address input appears to
the stuch high or low.
-tony
2:36 p.m.
On Sun, Feb 5, 2012 at 1:57 PM, Chuck Guzis <cclist at sydex.com> wrote:
The sure-fire way to kill most EPROMs is to install
them upside-down
and let electricity take its course.
I've seen them survive that (though I don't recommend doing it on purpose).
But as I said, UVEPROMs are pretty rugged things.
Indeed.
-ethan
3:55 p.m.
Many thanks for the suggestions. I've got a lot of evenings to pass
sorting the box. I'll let you all know how things go.
Back in the day, I once tested a 2732 as Chuck suggests and got an Intel
symbol tattoo on the end of my right index finger for it. It was
readable for quite a time before my fingerprint came back. If I were of
the bad guy persuasion, it would have been interesting to see the
fingerprint card. Maybe burn in a Signetics symbol on another finger,
and AMD and Harris on the rest. Not recommended.
Jim
On 2/5/2012 2:36 PM, Ethan Dicks wrote:
On Sun, Feb 5, 2012 at 1:57 PM, Chuck Guzis<cclist
at sydex.com> wrote:
The sure-fire way to kill most EPROMs is to
install them upside-down
and let electricity take its course.
I've seen them survive that (though I
don't recommend doing it on purpose).
But as I said, UVEPROMs are pretty rugged things.
Indeed.
-ethan
5:14 p.m.
On 5 Feb 2012 at 15:55, jim s wrote:
Back in the day, I once tested a 2732 as Chuck
suggests and got an
Intel symbol tattoo on the end of my right index finger for it. It
was readable for quite a time before my fingerprint came back. If I
were of the bad guy persuasion, it would have been interesting to see
the fingerprint card. Maybe burn in a Signetics symbol on another
finger, and AMD and Harris on the rest. Not recommended.
You forgot Seeq, Fujitsu,.... :) I wonder how many fingers you'd
have left.
Given what you time is worth, I'd probably sell these by the pound if
a few samples turn out to be good.
--Chuck
7:30 p.m.
--- On Sun, 2/5/12, jim s <jws at jwsss.com> wrote:
I have a come into an opportunity
which people here might help with.? The opportunity is
a 75 to 100# box of Eproms.?
Any ideas on what to do, what the chances are that
they will
work?
They will work. EPROMs are very hard to damage. Despite all the precautions and warnings,
most of this stuff is quite resilient. I have a box-o-EPROMs like that, scavenged pulls
from a recycler. I think I've come across two or three bad ones out of hundreds. And
there's no telling if they even worked before they were tossed into the pile.
I don't have the means or time to do testing,
though recent discussion suggest there might be some cheap
eprom programmer which might do that.? I have never
blown up a prom from normal handling, w/o antistat but with
the shear number of parts here, the odds are that some are
blown.? I am handling these now with antistat, but the
prior situation is what it is.
Don't bother testing them. I'll bet they work. As a frequent purchaser/user of
scavenged, piled in box EPROMs, I wouldn't worry. If you want to ship small quantities
without them getting bent, just make little stacks, interleaving the legs, invert the one
on the bottom, and tie them together with a rubber band. But, I've even been shipped a
box of ROMs just crammed in there, and they were mostly OK. Yeah, you have to sit there
and straighten legs, and a couple will get trashed, but again, these things are hardy.
Should I sort them by part, list them with a "if
it doesn't
work I'll send you another", sell them as untested pull lots
(I will sort and tube them by type a and PN as my
contribution to this project), or some other way?
If you want, you can sort them out by part number, but I'd just say send them out by
the pound as mixed lots. Much easier that way, and most of us hobbyists use a lot of
different types of ROMs. Maybe just go by "small flat rate box-full" or
something.
-Ian
10:08 p.m.
In article <1328499019.104.YahooMailClassic at web121602.mail.ne1.yahoo.com>,
Mr Ian Primus <ian_primus at yahoo.com> writes:
If you want, you can sort them out by part number, but
I'd just say
send them out by the pound as mixed lots.
I'd go for that.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 version available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
4655
days inactive
4656
days old
8 comments
7 participants
participants (7)
-
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
ethan.dicks＠gmail.com -
fraveydank＠gmail.com -
ian_primus＠yahoo.com -
jws＠jwsss.com -
legalize＠xmission.com