Spotting Fake Parts

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Joe., luketeaford, lisa, Kent

User avatar
Willzyx
Common Wiggler
Posts: 126
Joined: Mon Jul 18, 2011 12:58 am
Location: Taiwan

Spotting Fake Parts

Post by Willzyx » Sun Sep 18, 2011 10:48 pm

Its rare and doesn't happen to offen but sooner or latter you will get some fake parts. Living out here in Taiwan I do see a lot of sketchy parts in the local shops which is why I buy all of my parts from Mouser or trusted sources. But there is always that rare part that can force you onto eBay. Well here is what a fake IC looks like. You need a magnifying glass to see it.



Here are some picts of a fake SSM2044-
Look at the to circular indentations. There should be no texture in them and they should be the same hight.

Image

Here is picture of how a real IC look.Image

If you flip it around a fake part. You will notice how the circle dose not have a perfect lip and there is texture in it as well. The text will not be crisp.
Image


For comparison here is what a real TL074 looks like.

Image

Some parts are way easier to spot: :woah: Image

User avatar
decaying.sine
Broken <--> Fixed
Posts: 5398
Joined: Wed Jun 03, 2009 10:05 pm
Location: New Haven, CT USA

Post by decaying.sine » Mon Sep 19, 2011 12:41 am

Those are really great examples.

Do you know if the circle imperfections are due to sanding?
Brian
"I must create a system or be enslaved by another mans; I will not reason and compare: my business is to create." William Blake

"Vactrols ringing, Dude." "Thank you Donny"

User avatar
Muff Wiggler
wait, what?
Posts: 7815
Joined: Fri Nov 24, 2006 12:19 pm

Post by Muff Wiggler » Mon Sep 19, 2011 12:55 am

awesome thread :hail:

thanks! any more tips are always greatly appreciated!

User avatar
rosch
roadkill
Posts: 802
Joined: Thu Dec 23, 2010 12:44 pm
Location: doucheland

Post by rosch » Mon Sep 19, 2011 1:15 am

omg, there are even fake trimmers now

User avatar
andrewF
Super Deluxe Wiggler
Posts: 4042
Joined: Wed Apr 15, 2009 7:36 pm
Location: WA
Contact:

Post by andrewF » Mon Sep 19, 2011 1:46 am

does the trimmer actually work??

User avatar
Willzyx
Common Wiggler
Posts: 126
Joined: Mon Jul 18, 2011 12:58 am
Location: Taiwan

Post by Willzyx » Mon Sep 19, 2011 2:13 am

decaying.sine wrote:Those are really great examples.

Do you know if the circle imperfections are due to sanding?
I think they coat them with a type of texture paint and stamp it when its still wet. I want to file off the new paint to see what the chip really is.
rosch wrote:omg, there are even fake trimmers now
I have heard of fake resistors. :waah:
andrewF wrote:does the trimmer actually work??
I am going to find out here in a little while, but not keeping my hopes up.

User avatar
Willzyx
Common Wiggler
Posts: 126
Joined: Mon Jul 18, 2011 12:58 am
Location: Taiwan

Post by Willzyx » Mon Sep 19, 2011 2:21 am

Is there a good resource on date codes? I am not too sure if the 0025 makes any since ether.

I am no expert in this stuff so if anyone knows any other tricks let us know.

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Mon Sep 19, 2011 2:35 am

I'm confused, are the "Bochen" brand fake? They are available all over and seem quite indistinguishable in operation from any other trimmer of that form factor...

Blair

User avatar
Willzyx
Common Wiggler
Posts: 126
Joined: Mon Jul 18, 2011 12:58 am
Location: Taiwan

Post by Willzyx » Mon Sep 19, 2011 2:43 am

ringstone wrote:I'm confused, are the "Bochen" brand fake? They are available all over and seem quite indistinguishable in operation from any other trimmer of that form factor...

Blair
They are not fake but they are really trying to look like bourns trimers, maybe knock offs are a better term.

User avatar
gwaidan
Super Deluxe Wiggler
Posts: 1939
Joined: Sat Mar 07, 2009 7:26 pm
Location: Sinney, Orstraya
Contact:

Post by gwaidan » Mon Sep 19, 2011 2:56 am

The SSM2044 chips on eBay are pretty much all knockoffs, however most seem to work fine-most of them are datecoded 8730. Emu did recreate the SSM2044 when they reissued the SP1200 in the 90s so its possible that the schematic was leaked (or "retained") by a Chinese foundry. I've seen the "8730" chip in the photo of the Doepfer SSM filter module on their site.

I've seen Suntan brand trimmers with the 3296 designation-as it's a number there's nothing Bourns can do to protect it as IP so these are no more dodgy than Alpha sealed rotary switches copying Lorlin.

jupiter8
Wiggling with Experience
Posts: 359
Joined: Sat Jul 31, 2010 4:53 pm

Post by jupiter8 » Mon Sep 19, 2011 3:55 am

Willzyx wrote:
rosch wrote:omg, there are even fake trimmers now
I have heard of fake resistors. :waah:
That's nothing,they're counterfeiting eggs FFS.

User avatar
Paradigm X
Super Deluxe Wiggler
Posts: 2056
Joined: Tue Nov 02, 2010 5:43 am

Post by Paradigm X » Mon Sep 19, 2011 4:25 am

I am constantly worried about this, i have no idea. Maybe a 'good seller' thread would be helpful?

I bought some 2sa733ps recently whcih came stamped ksa but appeared to have the pinout of the 2sas, for a noob its really confusing and frustrating, not really money wise but time faffing around. Also got some pt2399s which i have not got round to trying yet.

Thanks

User avatar
Tim Stinchcombe
Uncommon Wiggler
Posts: 1420
Joined: Fri Mar 13, 2009 1:55 pm
Location: Cheltenham, Glos, UK

Post by Tim Stinchcombe » Mon Sep 19, 2011 6:27 am

The process is known as 'blacktopping', and Willzyx's first photo is one of the easier ways to spot it. Counterfeiters grind off the original marking, then apply a new coating (the blacktopping), and then re-etch markings into the device. I presume the coating is done by hand, and if the person doing it isn't very practiced, the molding holes can get filled, giving the tell-tale signs seen in the photo. Sometimes you can see the 'join' between the blacktopping and the original colour of the package sides, again depending on how well it has been done.

At work we went from being aware that counterfeiting exists to seeing three different examples in the space of just a few months. The first were FPGA chips, a large quad flat pack (QFP) package, which showed classic signs of blacktopping - most worked, so we can only assume they were old non-RoHS chips that had been re-date coded so they could be fobbed off as newer lead-free devices. The blacktopping also felt much, much rougher to the touch than the genuine article.

The second were EPROM chips, which we think were of smaller memory size than they should have been (and probably a different make) - these were confirmed as being 'bad' by the fact that we found three different type-faces for the etching on the chips, all with the same batch code (which seemed highly unlikely if the chips had been genuine)!

The third lot were some sort of CMOS logic chips, in a fairly common 14-pin SOIC package, again some showed signs of blacktopping, and since they simply didn't work, heavens knows what the chips actually were!

Apart from the annoyance of the cost (we had to scrap several hundred of the FPGAs, at over £5/chip), the biggest frustration is knowing that the counterfeiters simple get away with it scott-free, as there is no possible way to trace back how or where these chips entered the supply chain (most were bought on the 'grey market') - all one can do is learn how to spot them, and to remain vigilant. Another annoyance is that even though it is a massive problem for the industry, the sharing of data of chips known to have been counterfeited (chip type, date codes, batch nos. etc.) seems to be only available to those in the supply business, and not the rest of us, i.e. people buying the chips who are going to lose out if we get stung.

A few minutes Googling will generally bring up many hits of sites showing photos of bad chips and how to spot them.

Tim

User avatar
Psychlist1972
Veteran Wiggler
Posts: 508
Joined: Sat Oct 13, 2012 11:55 am
Location: Near Annapolis, Maryland, USA

Post by Psychlist1972 » Fri Jan 04, 2013 2:21 am

From looking at these photos, the XR2206 chips I bought some time ago from ebay (seller polida2008 "Polida Electrical Limited") are almost certainly fakes.

I haven't had a chance to test them yet. I just pulled them out after stumbling across this thread.

But they have textured ejection pin marks as well as low-res laser etched markings.

Image

Image

Image

The seller is still selling XR 2206 chips. Photos uploaded in case they help anyone else.

Pete
Pete Brown
@pete_brown | SoundCloud | YouTube | Flickr

[I work in Windows at Microsoft, but participate here in a personal capacity. My opinions are my own.]

User avatar
negativspace
Manhattan Analog
Posts: 6576
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Fri Jan 04, 2013 2:47 am

I have a large (50+) batch of CA3080 that are plainly blacktopped but are 100% good. I tested every one vs. a known sample pulled from a Doepfer module as well as one from a batch of pulls I bought years ago. Makes me curious... why fake working chips? No ROHS shenanigans with those, so... :despair:

User avatar
Psychlist1972
Veteran Wiggler
Posts: 508
Joined: Sat Oct 13, 2012 11:55 am
Location: Near Annapolis, Maryland, USA

Post by Psychlist1972 » Fri Jan 04, 2013 2:48 am

negativspace wrote:I have a large (50+) batch of CA3080 that are plainly blacktopped but are 100% good. I tested every one vs. a known sample pulled from a Doepfer module as well as one from a batch of pulls I bought years ago. Makes me curious... why fake working chips? No ROHS shenanigans with those, so... :despair:
Probably factory rejects that failed one quality test or another. I'd be curious to know if they're 100% in spec, or if they just work.

Pete
Pete Brown
@pete_brown | SoundCloud | YouTube | Flickr

[I work in Windows at Microsoft, but participate here in a personal capacity. My opinions are my own.]

User avatar
negativspace
Manhattan Analog
Posts: 6576
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Fri Jan 04, 2013 2:59 am

Well, all i have is in-circuit observation but they all performed indistinguishably from their genuine counterparts. I used a 258J oscillator as a testbed but have since used them in dozens of assorted modules to the same effect. Maybe typical synth circuits just don't put them through enough stress to bring out the bad behavior?

I did get 10 a couple years back that were bunk and i kept one for reference. It failed in the 258j completely.

User avatar
mOBiTh
Super Deluxe Wiggler
Posts: 1992
Joined: Sun May 20, 2012 2:20 pm

Post by mOBiTh » Fri Jan 04, 2013 6:35 am

Perhaps at very high frequencies their behaviour is poor?

what about if they were incorrectly marked at the production stage, so therefore needed remarking?

Luap
Super Deluxe Wiggler
Posts: 1126
Joined: Fri Nov 30, 2012 6:41 pm
Location: North'ish London, UK

Post by Luap » Fri Jan 04, 2013 12:15 pm

Another give away is laser etching on parts that would have been produced long before laser etching even existed.

djs
Super Deluxe Wiggler
Posts: 1174
Joined: Wed Sep 01, 2010 8:36 am
Location: South Beloit, IL

Post by djs » Fri Jan 04, 2013 1:17 pm

I found this link very interesting and useful:
http://www.aeri.com/counterfeits.html

I got some 2n3904's in a shipment from an ebay seller where the datecode was the exact same on two parts in the same shipment, but the pinouts differ :despair: However, they tested out to be the right type of transistor, so I just marked the bag with a note "check pinout!", and will test each one again before use.

note- the seller was thaishine- Usually i buy only sockets and things like tl072s from them with no issues, so i'm not sure what the case is here. Since they still work and the cost was pretty cheap, I didn't bother to contact them on it.
"Noise is what the Earth is made of" - David Bowie

User avatar
valis
aliens killed bigfoot
Posts: 2133
Joined: Tue Sep 30, 2008 3:53 pm
Location: on the palouse, usa

Post by valis » Fri Jan 04, 2013 1:33 pm

negativspace wrote:I have a large (50+) batch of CA3080 that are plainly blacktopped but are 100% good. I tested every one vs. a known sample pulled from a Doepfer module as well as one from a batch of pulls I bought years ago. Makes me curious... why fake working chips? No ROHS shenanigans with those, so... :despair:
This is what I'm curious about too. Maybe it's the copyright infringement that makes it lucrative.

User avatar
consumed
International Filter Conspiracy
Posts: 3700
Joined: Fri May 25, 2007 1:38 pm
Location: nor cal

Post by consumed » Fri Jan 04, 2013 1:45 pm

interesting. i recently purchased a discontinued AD chip from HK for a module. ill have to go have a close look at it.

User avatar
Psychlist1972
Veteran Wiggler
Posts: 508
Joined: Sat Oct 13, 2012 11:55 am
Location: Near Annapolis, Maryland, USA

Post by Psychlist1972 » Sun Jan 06, 2013 3:39 am

djs wrote: note- the seller was thaishine- Usually i buy only sockets and things like tl072s from them with no issues, so i'm not sure what the case is here. Since they still work and the cost was pretty cheap, I didn't bother to contact them on it.
thaishine is Tayda Electronics. I've bought a fair bit from them on their regular website. No issues so far. They claim to only purchase direct from the manufacturer.
http://www.taydaelectronics.com/

Pete
Pete Brown
@pete_brown | SoundCloud | YouTube | Flickr

[I work in Windows at Microsoft, but participate here in a personal capacity. My opinions are my own.]

User avatar
neil.johnson
Super Deluxe Wiggler
Posts: 1402
Joined: Thu Apr 08, 2010 3:51 pm
Location: Cambridge, UK
Contact:

Post by neil.johnson » Sun Jan 06, 2013 4:09 am

I'm sure Tayda has superior quality control and vendor qualification than military electronics firms:
Fake chips from China sold to U.S. defense contractors

:despair:

Neil
Random ramblings :: http://www.njohnson.co.uk
SSI parts and more :: https://www.cesyg.co.uk

User avatar
Tim Stinchcombe
Uncommon Wiggler
Posts: 1420
Joined: Fri Mar 13, 2009 1:55 pm
Location: Cheltenham, Glos, UK

Post by Tim Stinchcombe » Tue May 28, 2013 5:42 pm

Necroing this thread rather than starting another - not specifically 'synth diy', but of possible general interest to the subject of counterfeiting.

In the couple of years since I last posted, we have continued to see more counterfeits at work, especially of several of the chips we use lots of that are now going obsolete. Thankfully we are getting better at spotting them, and most of the following are sample chips, which funnily enough we declined to buy!

First up they got sloppy with this one, and didn't abrade off the original engraving enough, so characters are still visible underneath - you can make out '3' or '5' 'G', under the '4C', and note also that the pin 1 dimple is quite matt, and not shiny as a real chip would be:
Image

Sometimes curiosity gets the better of me, and you know you just have to try and find out what the chips really are. This is an example where on scraping off the blacktopping the original markings were plain enough to see, and also easy enough to get a snap through the microscope - the '0349' date code is the original; the '1135' is the remarked one, so these appear to be non-RoHS chips made to look like younger, RoHS ones:
Image

Here's the recent photo I was pleased with, prompting me to make this post - whilst the chips were clearly blacktopped, they we also otherwise in pristine condition (they're a large 208-pin QFP package), prompting me to wonder why. On removing the topping, the answer appeared to be that they simply weren't the right chips, so presumably cheap ones made to look like ones that can command a higher price! The original engraving was very hard to make out, so an ordinary photo would have taken a lot of messing around to get the light right. So I rubbed a highlighter pen over it, and used a UV light - the result has made the original markings visible. Instead of the chips being 'XCS20XL', as now marked, underneath that you can make out that they are 'XC2S200E'- not even the same family of FPGA!:
Image

And here is another easy spot - Glink DRAM samples. The font immediately looked different, and a simple 'ink permanency' test with some acetone just wiped the paint right off, along with a good deal of blacktopping, as the swab testifies! Top and bottom are bad, in the middle is a genuine chip (the acetone simply doesn't touch it!):
Image

I might sift through a few more snaps depending on interest/as time allows.

Tim

Post Reply

Return to “Music Tech DIY”