15 Jul
2013
15 Jul
'13
12:14 a.m.
Just ran across an interesting oddity --
I ordered a few 7475s to replace the ones with corroded legs on my
Imlac's memory control boards; upon replacement I note a dead short
between 5V and GND (as evidenced by the 5V supply being drawn down to
1.5V or so on powerup). After pulling the chip out, it's pretty clear
that pin 5 (VCC) is shorted to pin 12 (GND), as are pins 4 and 13.
Verrry odd. I grab the other replacements I ordered and they test the
same way.
My curiosity getting the better of me, I break the top off the chip (not
entirely successfully). You can see the results here:
http://yahozna.dyndns.org/scratch/imlac/fake7475.jpg
The picture's not great, but you can see that pins 4,5,12, and 13 are
connected to asolid piece of metal! There does appear to be adie in the
chip, interestingly enough(not pictured). Defective chip? Other chip
relabeled to be a 7475? Who knows...
Unfortunately, I can't find the invoice for these items (I ordered these
a couple of months back), so I don't know where it came from; it was
probably either Jameco or Digi-Key...
Ugh, I wonder how many others of these chips I ordered are fakes...
- Josh
15 Jul
15 Jul
8:50 a.m.
On 07/15/2013 12:14 AM, Josh Dersch wrote:
Just ran across an interesting oddity --
I ordered a few 7475s to replace the ones with corroded legs on my
Imlac's memory control boards; upon replacement I note a dead short
between 5V and GND (as evidenced by the 5V supply being drawn down to
1.5V or so on powerup). After pulling the chip out, it's pretty clear
that pin 5 (VCC) is shorted to pin 12 (GND), as are pins 4 and 13.
Verrry odd. I grab the other replacements I ordered and they test the
same way.
My curiosity getting the better of me, I break the top off the chip
(not entirely successfully). You can see the results here:
http://yahozna.dyndns.org/scratch/imlac/fake7475.jpg
The picture's not great, but you can see that pins 4,5,12, and 13 are
connected to asolid piece of metal! There does appear to be adie in
the chip, interestingly enough(not pictured). Defective chip? Other
chip relabeled to be a 7475? Who knows...
Unfortunately, I can't find the invoice for these items (I ordered
these a couple of months back), so I don't know where it came from; it
was probably either Jameco or Digi-Key...
Ugh, I wonder how many others of these chips I ordered are fakes...
- Josh
Post a clear picture of the top of the ICs - Digi-Key only sells a few
brands (and only ALS or HC versions) and it would be fairly easy to see
if they were the source.
Jameco does show a 7475 @ $2.95 each...you should be able to get
confirmation from them if the date codes, etc. match.
Plus both Jameco and Digi-Key email invoices so (assuming you keep your
emails) you should have those and they wuld show order dates. Digi-key
keeps your order history and you can check that online, I'm not sure if
Jameco does too, but I imagine they do.
Counterfeit parts are a real pain!
John :-#(#
--
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
Call (604)872-5757 or Fax 872-2010 (Pinballs, Jukes, VideoGames)
www.flippers.com
"Old pinballers never die, they just flip out"
11:36 a.m.
On Monday, July 15, 2013 at 0:14, Josh Dersch wrote:
http://yahozna.dyndns.org/scratch/imlac/fake7475.jpg
The picture's not great, but you can see that pins 4,5,12, and 13 are
connected to asolid piece of metal!
That's a somewhat common arrangement for higher-current peripheral drivers
in DIPs. See, for example, the TI SN754410 Quadruple Half-H Driver, where
those pins are used for heat sinking.
Other chip relabeled to be a 7475?
That's my guess. TI had an SN75xxx series of interface parts. I wondered
if perhaps a 7574 was mislabeled as a 7475, but I can't find a reference to
the former.
Could you post a photo of the top of one of the bad parts that shows the
device marking?
-- Dave
6:36 p.m.
On 7/15/2013 11:36 AM, J. David Bryan wrote:
On Monday, July 15, 2013 at 0:14, Josh Dersch wrote:
I've put a picture up at:
http://yahozna.dyndns.org/scratch/imlac/fake7475-full.jpg
I was paranoid after you mentioned the "7574" thing that perhaps I had
gone insane (or become dyslexic) and had spent an hour last night
misreading a chip label, but no, it's clearly labeled as an SN7475N with
TI branding :).
- Josh
Other chip relabeled to be a 7475?
That's my guess. TI had an SN75xxx series of interface parts. I wondered
if perhaps a 7574 was mislabeled as a 7475, but I can't find a reference to
the former.
Could you post a photo of the top of one of the bad parts that shows the
device marking?
-- Dave
9:15 p.m.
On 07/15/2013 06:36 PM, Josh Dersch wrote:
I've put a picture up at:
http://yahozna.dyndns.org/scratch/imlac/fake7475-full.jpg
I was paranoid after you mentioned the "7574" thing that perhaps I had
gone insane (or become dyslexic) and had spent an hour last night
misreading a chip label, but no, it's clearly labeled as an SN7475N with
TI branding :).
Well, to my eye, it looks like no TI chip that I've ever seen. In my
plain-jane SN74xxx (not LS, S, etc.) collection from the 70s, the TI
packages that I have do not have pin 1 "dots", but the usual rectangular
relief on the pin 1 end. Your package looks more like a Motorola
bug--particularly the "fuzzy" pin 1 depression.
I believe you've got a forgery. On the clumsier forgeries, paint is
applied over the old marking and then the labeling. You might see if
that's the case with your chips.
--Chuck
16 Jul
16 Jul
2:02 a.m.
On Mon, 15 Jul 2013, Chuck Guzis wrote:
On 07/15/2013 06:36 PM, Josh Dersch wrote:
It looks like a counterfeit part to me as well. The spacing of the
font/printing is wrong and it isn't as sharp as the genuine TI chips I
have here that have the later style of printing. Is there a country of
origin molded into the bottom of the IC? All of the TI chips I just looked
at have such markings. The plating on the leads of the ICs I compared is
also very very smooth and almost polished-like. FWIW, my money is on
Jameco since DigiKey is very very cautious where they source
semiconductors from.
I've put a picture up at:
http://yahozna.dyndns.org/scratch/imlac/fake7475-full.jpg
I was paranoid after you mentioned the "7574" thing that perhaps I had
gone insane (or become dyslexic) and had spent an hour last night
misreading a chip label, but no, it's clearly labeled as an SN7475N with
TI branding :).
Well, to my eye, it looks like no TI chip that I've ever seen. In my
plain-jane SN74xxx (not LS, S, etc.) collection from the 70s, the TI packages
that I have do not have pin 1 "dots", but the usual rectangular relief on the
pin 1 end. Your package looks more like a Motorola bug--particularly the
"fuzzy" pin 1 depression.
I believe you've got a forgery. On the clumsier forgeries, paint is applied
over the old marking and then the labeling. You might see if that's the case
with your chips. 
7:54 a.m.
On Tue, Jul 16, 2013 at 12:15 AM, Chuck Guzis <cclist at sydex.com> wrote:
On 07/15/2013 06:36 PM, Josh Dersch wrote:
Well, to my eye, it looks like no TI chip that I've ever seen.
I agree. It just doesn't "look right".
Your package looks more like a Motorola
bug--particularly the "fuzzy" pin 1 depression.
I can believe it's a Motorola package. I remember them as having
somewhat rounded corners and a prominent seam around their
"equator". Also, ISTR TI parts having a short body and corner pins
that are right up against the edges of the package.
I believe you've got a forgery. On the clumsier
forgeries, paint is applied
over the old marking and then the labeling. You might see if that's the
case with your chips.
Indeed. In fact, you might be able to track down the (relatively)
small number of parts with those center pins bonded to each
other (certainly a much smaller pool than packages without
shorted pins).
I have a (PIC-based) chip tester that has a "discovery" feature, so
you can drop in a '00 or '02 or '14 or '74, etc., etc, and press
"test"
and it will stop on the first part that passes (that it recognizes).
I don't know if it handles any parts that happen to have those
internal connections on the middle pins, though.
-ethan
9:19 a.m.
I have a (PIC-based) chip tester that has a
"discovery" feature, so
you can drop in a '00 or '02 or '14 or '74, etc., etc, and press
"test"
and it will stop on the first part that passes (that it recognizes).
I don't know if it handles any parts that happen to have those
internal connections on the middle pins, though.
Can you give more info about this tester? ;o)
10 a.m.
Alexandre Souza <alexandre.tabajara at gmail.com> wrote:
No idea what kind of tester he has, but at least the old ALL-02 can do
exactly this kind of thing. You can even write own testing patterns for
chips unknown to the software (I did so).
http://matthieu.benoit.free.fr/145.htm
I have a
(PIC-based) chip tester that has a "discovery" feature, so
you can drop in a '00 or '02 or '14 or '74, etc., etc, and press
"test"
and it will stop on the first part that passes (that it recognizes).
I don't know if it handles any parts that happen to have those
internal connections on the middle pins, though.
Can you give more info about this tester? ;o)
10:40 a.m.
On Tue, Jul 16, 2013 at 12:19 PM, Alexandre Souza
<alexandre.tabajara at gmail.com> wrote:
I don't have it in front of me, but it _might_ be a TPT-100 "Handy
Tester". I can't find any data on it online (it's 20 years old) but
it's
similar to the Instek GUT-6600A (except that the TPT-100 can
test +5V DIP DRAMs from 64Kx1 through 1Mx4). There's an
older PIC microcontroller inside that reads the 5-6 buttons and
displays results on a 16x1 textual LCD. You insert the chip,
either navigate up/down through the list of supported part numbers
or press "TEST" to start at the bottom and work up the list
until there's a functional match. Great for identifying common
parts with obliterated numbers or OEM part numbers. It's not
perfect, but you can easily tell a 7400 from a 7474.
Long ago, I rigged up a DIP cable to a TC-16 test clip so I could
test all the ICs in simple (no interconnects) DEC M-series modules.
The power requirement of all 4-6 chips per board didn't crush
the tester, and I just had to reposition the clip, press "test",
look for "OK", then repeat.
We got it years ago because we had a 5150 (or compatible) PC
that we used with a $20K 68000 tracer/debugger that wouldn't
boot, and it was the least expensive instrument to test the DRAMs
one chip at a time (we identified 2 bad DRAMs and one bad
mux and had those parts inches away on the test bench).
I use it almost every month.
-ethan
I have a
(PIC-based) chip tester that has a "discovery" feature, so
you can drop in a '00 or '02 or '14 or '74, etc., etc, and press
"test"
and it will stop on the first part that passes (that it recognizes).
I don't know if it handles any parts that happen to have those
internal connections on the middle pins, though.
Can you give more info about this tester? ;o)
11:56 a.m.
On 07/16/2013 09:19 AM, Alexandre Souza wrote:
They still exist. I believe that even some inexpensive Chinese TOP
EPROM programmers have this mode. My vintage Superpro programmer has
this function. However, most such devices are limited to recognizing
simple combinatorial logic--my Superpro can't, for example, recognize a
74LS281.
--Chuck
I have a
(PIC-based) chip tester that has a "discovery" feature, so
you can drop in a '00 or '02 or '14 or '74, etc., etc, and press
"test"
and it will stop on the first part that passes (that it recognizes).
I don't know if it handles any parts that happen to have those
internal connections on the middle pins, though.
Can you give more info about this tester? ;o) 
10:40 a.m.
On Jul 15, 2013, at 9:15 PM, Chuck Guzis <cclist at sydex.com> wrote:
On 07/15/2013 06:36 PM, Josh Dersch wrote:
No date code, either, which is odd. I don't think that's very common.
-Seth
I've put a picture up at:
http://yahozna.dyndns.org/scratch/imlac/fake7475-full.jpg
Well, to my eye, it looks like no TI chip that I've ever seen. 
10:56 a.m.
I've put a picture up at:
http://yahozna.dyndns.org/scratch/imlac/fake7475-full.jpg
Well, that's marked as a TI 7475, but the packe doesn't look like a TI
one to me... Has it been relabelled?
It's an odd thign to fake. The 7475 is hardly a rare or valuable IC. But
who knows...
There certainly are companies who fake ICs. At one point I found a site
that advertised just about every IC I'd ever heard of, including HP
custom ones. I enquired about one the latter to be asked 'What package do
you want?' On the grounds that said IC only ever came in one package, I
suspected they were simply going to relabel any IC with the right number
of pins for me.
-tony
15 Jul
15 Jul
11:42 a.m.
Just ran across an interesting oddity --
I ordered a few 7475s to replace the ones with corroded legs on my
Imlac's memory control boards; upon replacement I note a dead short
between 5V and GND (as evidenced by the 5V supply being drawn down to
1.5V or so on powerup). After pulling the chip out, it's pretty clear
that pin 5 (VCC) is shorted to pin 12 (GND), as are pins 4 and 13.
Are you saying all 4 pins are shorted together?
There are ICs with the middle pins all commoned like that. Some
darlington driver type thigns have the common emitter (0V) connection
wired to the centre 4 pins, the ideal is that they can go to 'heat sink
traces' on the PCB.
Verrry odd. I grab the other replacements I ordered
and they test the
same way.
My curiosity getting the better of me, I break the top off the chip (not
entirely successfully). You can see the results here:
http://yahozna.dyndns.org/scratch/imlac/fake7475.jpg
Do you have a picture of the markings on an unbroken IC? Is it posible
that the '7475' is the date code or soemthing? Who claims to have made
the IC?
-tony
11:55 a.m.
On Jul 15, 2013, at 3:14 AM, Josh Dersch <derschjo at gmail.com> wrote:
Unfortunately, I can't find the invoice for these
items (I ordered these a couple of months back), so I don't know where it came from;
it was probably either Jameco or Digi-Key...
I'm not 100% sure about Jameco, but Digi-Key has excellent past invoice
retrieval capabilities. You can easily pull a PDF of the invoice from
at least the past year.
Might be worth checking that you didn't at least order the wrong part,
anyway. And, if you do find that you have mismarked/counterfeit parts,
I'm sure Digi-Key or Jameco would love to know about it, since they're
both pretty reputable companies.
- Dave
4178
days inactive
4178
days old
14 comments
11 participants
participants (11)
-
alexandre.tabajara＠gmail.com -
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
derschjo＠gmail.com -
ethan.dicks＠gmail.com -
fraveydank＠gmail.com -
jdbryan＠acm.org -
lehmann＠ans-netz.de -
lists＠loomcom.com -
pinball＠telus.net -
tothwolf＠concentric.net