2N5172 from eBay - real or fake?

Started by Onion Ring Modulator, May 29, 2022, 03:05:50 PM

Previous topic - Next topic

Onion Ring Modulator

These look really weird, but I've never had this particular part number before, so who knows? What do you guys think?








antonis

"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

idy

Those look like older transistors.

The dabs of color on the back imply that someone, somewhere in time, measured them and sorted them... before they ended up in the old lot/surplus bin.

Test them for gain. If they work the way you want, use them.

Rob Strand

#3
I believe the 2N5172's appears in wha pedals.  Can't remember which model off the top of my head, I'd would have to check my list.

The Fairchild 2N5172 datasheet indicates these come from "process 10".  BC547's also come from process 10.  The details graphs appear in the PN100/PN100A datasheet.
https://www.farnell.com/datasheets/243186.pdf

QuoteThe dabs of color on the back imply that someone, somewhere in time, measured them and sorted them... before they ended up in the old lot/surplus bin.
It was common for those paint markings to come from the manufacturer.  I suspect Fairchild and National Semiconductor.

I've pulled few examples,
part          mfg, year,  marks
2N3707      ?, 1967, 1xyellow
PN2222A    national semiconductor, 1980 (date code 003), grey green  ; grey could be white
2N5089     national semiconductor, 1984 (date code 429),  yellow red (violet?) grey
2N5172     national semiconductor, 1984 (date code 429),  yellow red (violet?) grey
2N5172     national semiconductor, 1984 (date code 439),  yellow silver violet

For the two bands the paint markings are next to each other, not at the top and bottom of the rear dot.

Interesting the 2N5089 and 2N5172 with same the date code have the same paint markings.   We might be tempted to say it's a secret date code, however, it might actually mean the factory or production line or something.

My 2N5172's have TO92 packages (simple and small style).  However early Fairchild did use the package shown in the pics at the top of the thread.  (IIRC) often the parts had gold leads with this package in the early days.

There are gain markings on *some very old devices*, perhaps late 60'.  Color markings are a little uncommon for US and EU transistors.  The (complete) datasheets did actually state colors were used for these parts.  An example in this thread and on the Boss Tone  transistor pics,
https://www.diystompboxes.com/smfforum/index.php?topic=48469.msg1150604#msg1150604



I dug up some Fairchild parts (PN4250 date code 308, PN3569 date code 309, PN4355 date code 309; perhaps 1983 more than 1973).  Both have the package as the part at the top of the thread but they have gold leads and there is no paint markings.   They don't have EBC markings at all and the F has the "scripted" Fairchild placed before the date code.

FWIW, National semiconductor took over Fairchild in 1987.   It's possible Fairchild parts after 1987 have the paint markings.  but no ...


I dug up another Fairchild example MPSA42, perhaps 1986 to 1989 (can't narrow it down).   The part is a simple TO-92 package, tinned leads, silvered face with black text.  A non-scripted F (fairchild) is before the part number (MPSA42).   The face/leads are marked EBC.   No paint marks but has 14 printed in the top.

Also a Fairchild 2N5401 marked exactly the same.  No idea of date but post 1990 likely.  No paint marks but has 20 printed in the top.


From all that I can't see the features of the part at the top of the thread matching the pattern of my examples:  Old style package with paint but Fairchild with large F in front of the p/n and EBC markings.  Also tinned leads.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Rob Strand

#4
Here's the same parts the op has,
https://reverb.com/item/40731105-fairchild-2n5172-silicon-npn-transistors-nos-bag-of-1000

The bag is stamped datacode 918.   Probably 1989, so National Semiconductor owned Fairchild.

Still not sure what to think.

Another angle is the "Assembled in Korea".   You would need to know when Fairchild made stuff in Korea and if they made transistors there.    I couldn't find much, possibly 1998, and in that case 918 might actually mean 1999.

Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

anotherjim

Not fakes. Those colour bands were common. I've never bothered to find out what they signify since they have part numbers. Sorted parts have a colour dot applied. Service sheets sometimes specify that dot with the part number although I may be quoting European practice.
However, old parts might be around that were factory goods inward inspection fails -  staff could take them home for little Johnny to play with. They may only be cosmetic fails. Resistors could be rejected because the lead bender got them off centre!
And does anyone remember the Archer assortment grab bags? Those parts didn't look good enough to be fakes!


Rob Strand

#6
QuoteAnd does anyone remember the Archer assortment grab bags? Those parts didn't look good enough to be fakes!
The logic IC's looked very weird.  They had a very light grey color and shiny plastic texture.  I don't remember any commercial chips like that.  IIRC the chip dates were 1968 to 1970 (maybe 1971) era - even though they were sold around 1978/1980.   I just found one and the manufacturers logo is kind of a square, it looks like a badly drawn elephant with rabbit's back legs  :o    Come to think of it they had miscellaneous paint marks on them - all different colors.  Perhaps they were floor scrapings that someone sorted into working batches.

Most of the voltage regulators were unmarked, you had to work out the voltage yourself.   Also the TO-220 regulators only gave out 500mA instead of 1A.  I know I've still got some of those around with the voltage hand marked on the back.


The logo was an old General Instrument logo!

Bottom of first column, second from bottom.
https://www.applefritter.com/content/ic-logos
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

duck_arse

I've bought a number of parts [in 100 lots] from a local supplier, mostly remainders/remnants, all dusty and old and obscure, from before the chinese fad fakers.  my $5/100? they look legit, silver inked markings, including the F, the ECB [actually, is the F2N part no indicative of the end base pinout?] markings, the colour bands [always good for sorting by], that dumb ugly variant TO92 package, and especially the plastic flashing along the moulding seams, and the thick tin plating.

at your risk, with salt pinch.
" I will say no more "

mozz

#8
Thinking the original 2N5172 was a package such as a 2n3391/92/93 TO-98-1. I've often seen a part number, made by different manufactures, to not be the exact same. It may meet spec but it isn't gonna be a twin. Such as 2N270, I have some, not RCA as was in the original fuzz, and they don't work properly.

What i am trying to say is, if you are trying to duplicate a pedal, and use original transistors, try to find out what make they were and search for those, or something made by that manufacturer in the same family.

"Color markings are a little uncommon for US and EU transistors." For US made parts, i've seen it a very lot of times. I guess you would have to had worked there to decode it and know the secret. It could have been gain, could be the machine they were made on, the shift, the building, the date code. Could have been the person who ordered the parts wanted color codes so the assemblers would have it easier and less mistakes.

GE or Fairchild?




  • SUPPORTER

ElectricDruid

I used to have lots of trannys with paint blobs like that that I'd pulled out of dead gear from the 70s and 80s.

I reckon they're old (who the hell bothers sorting transistors and putting paint on them nowadays?! In fact, who's ever done that in the last twenty years?!) but they're probably legit.

Phend

#10
Used this at times. For instance 2N5172 = ECG 289
Good ole days.

  • SUPPORTER+
Do you know what you're doing?

Rob Strand

#11
QuoteI reckon they're old (who the hell bothers sorting transistors and putting paint on them nowadays?! In fact, who's ever done that in the last twenty years?!) but they're probably legit.
In the old days I remember the parts with that package had gold leads. (my examples from early/mid 80s')

You can see a Fairchild device with the same package and tinned leads here (but can't see rear/paint)
https://www.radiomuseum.org/tubes/tube_fpn4250.html

The version of the part has no date code marked on the part.  And from the post with the bag pic you can see the date code was moved to the bag label.

FYI, found this re Korea:

https://eepower.com/news/fairchild-semiconductors-announces-its-south-korean-fab-has-made-10-billion-semiconductor-packages/#

"Fairchild Semiconductor International (South Portland, ME) announced that its plant in Puchon, South Korea, has manufactured 10 billion semiconductor packages since its acquisition in April 1999.Since the acquisition, the Korean plant set record revenues for its 26-year history and now exports half the semiconductors it produces. Worldwide sales of the plant's products increased 46 percent in the first quarter of this year, compared to the corresponding quarter last year. In response to the growing worldwide demand for its products, Fairchild dedicated $80.0 million this year to expand the wafer fabrication operation in Puchon and added at least 200 jobs."

That however refers to a Fab.  It's possible to have plants which package (ie. put silicon in packages) without making the actual silicon die.  Because of that it's still not clear if they are 1989 or 1999.

(My gut feeling is these are 1989 parts from when National Semiconductor owned Fairchild.)


Oddly enough a non-scripted F logo, tinned leads and two colored stripes, 1982.
https://www.ebay.com/itm/332411742733
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

anotherjim

Quote from: mozz on May 30, 2022, 03:19:17 PM
Could have been the person who ordered the parts wanted color codes so the assemblers would have it easier and less mistakes.

Could well be this reason. Pick & place machines back then were humans with minimal training.


duck_arse

Quote from: mozz on May 30, 2022, 03:19:17 PM

GE or Fairchild?



I'd say GE. that other awful package, and those thick round leads. also, didn't we argue in another thread that F2N and 2N were two distinct part numbers, at least as far as the pinout was concerned? P2N2222 or something?
" I will say no more "

Kevin Mitchell

  • SUPPORTER

ElectricDruid

Quote from: Kevin Mitchell on May 31, 2022, 11:10:52 AM
Quote from: ElectricDruid on May 30, 2022, 04:25:30 PMIn fact, who's ever done that in the last twenty years?!)
:icon_redface: :-X
What, really?!? Have you spent at least some of the last twenty years putting paint blobs on jellybean transistors?

I mean, I admire your patience if you have, because I certainly can't be ars3d. My general approach is that if it doesn't work with the random stuff that someone pulls of the shelf, I need to redesign it so it does. But I do recognise that other people have higher standards and *actually do* stuff like part matching!! ;)

In all honesty, I do in fact have one batch of old germanium OC71s that I measured and wrote little labels for showing their gain and leakage or whatever it was I thought I needed to know. So I've done it too, once or twice. Just without paint. :icon_redface:

Rob Strand

#16
QuoteWhat, really?!? Have you spent at least some of the last twenty years putting paint blobs on jellybean transistors?
When I was a kid I used to paint the tops of weird ass transistors so I knew they were NPN and PNP (following the blue or red dot scheme you find on old germaniums).

FWIW, the JFETs in old Maxon/Ibanez pedals have paint marks which represent the VGSoff parameter.   They must has sorted them in the factory.  (There's a table on the schematic in the Round button Maxon/Ibanez PT-909 thread.)
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Kevin Mitchell

Quote from: ElectricDruid on May 31, 2022, 04:40:17 PM
Quote from: Kevin Mitchell on May 31, 2022, 11:10:52 AM
Quote from: ElectricDruid on May 30, 2022, 04:25:30 PMIn fact, who's ever done that in the last twenty years?!)
:icon_redface: :-X
What, really?!? Have you spent at least some of the last twenty years putting paint blobs on jellybean transistors?
No lol. Just the passed three years and maybe only a couple hundred transistors have been marked by me. I got into the habit of measuring transistors & bagging them to save time down the road - mostly for folks who ask me if I can do it for them as they don't feel confident in doing it themselves. This is for VBE, HFE and sometimes both. Since tolerances are so tight nowadays for silicone devices I usually have a large pile of VBE matches that I then categorize for gain.

Is it necessary? Hell no. But normal people don't need to use mason jars to store their transistors :icon_lol:
Having them at the ready means I can offer these to friends & save future Kevin some work. Plus it's a nice nod to what we see in vintage synths.
  • SUPPORTER