Fakes

[GoranP]

With all the talk lately about fake J201s and 2N5458s and 4558s and such, I was just wondering...

How seriously do you take your building? Especially you guys that have huge stashes and who are selling your stuff. Is this a concern for you? Do you buy bulk, measure and write off the faulty ones or do you stock up on the real deal no matter the cost?

Any other thoughts on the matter welcomed...

[IvIark]

It would concern me more about the JFETs now, but I really don't think anyone is selling fakes at this time.  It would take a hell of an investment to get the manufacturing process and tooling to make small transistors that you will sell for £0.12 each, especially when you would only sell to most customers once if they didn't work.  I reckon more likely they are selling examples in the extreme end of their datasheet characteristics that may have been sorted as borderline.  I've seen figures from Tayda and Polida J201's that are definitely within spec and know people have built effects using them with good results.

Fortunately I bought in bulk from two sellers in the US before they became rare, and it is worth keeping an eye of dpi4parts and ACE-LECTRONICS on eBay for when they have new auctions listed.

ICs are different because characteristics that interest us like gain are functions of external components and you're not relying on what you're given to the same extent.  And as long as you have a socketed IC effect somewhere you can give them a quick test when they arrive and if they do sound utterly shit like they are rejects or something, then you're well protected by eBay to get your money back.

So all in all, as long as I'm not spending big money I think it's a risk I'm still prepared to take, but if you do find a good seller (G&C Supermarket seem to be good and I've no complaints so far), then stock up so you have plenty for the future.  

I've bought a lot of JFETs and ICs now and as someone who will 99% of the time build for himself, I know that I will basically never run out.  The worst thing would be to be left with none and having to move over to surface mount!  Urgghhh.

[GoranP]

Thanks Mark!

I'm sure you are aware of the numerous fake J201 threads on the other forums where people tested whole batches of them and found them to be waaay outside of datasheet specs. These are most likely factory rejects and not outright fakes but still the net effect is the same if one's clone is nowhere near the original. Right now I'm thinking that my Wampler B65 build sounds like crap because of J201 being off spec. I didn't socket back then so I can't check. :)
With J201 prices being what they are, I cannot afford to stock and sort afterwards and it's hard to tell which ones are legit. I'm thinking that Steve at Smallbear might be stocking the real deal but at $0.55 a pop with $25 shipping, there's unfortunately no way I'm buying from him.

Opamps don't bother me as much but fake pt2399 ($3.5ppc at Smallbear, $3 per 10pcs at g&c ?!?) and especially bbd's do.

And I hear ya on the smd, my eyes aren't what they used to be so that's gonna be quite a pain when it happens (and it will, no doubt).

[IvIark]

The thing is, if you buy them from Tayda they'll replace them without question, if you buy them from eBay you're more protected than anywhere else in terms of getting your money back.  So there isn't much of a risk.  If you don't get the answer you want from the seller, immediately open a case with eBay and you will definitely get your money back.

I'm still not convinced what we are seeing are ones way out of spec.  I just think some people don't know what they're buying, or maybe some sellers don't know what they're selling.  On the DIYSB thread some were in the specs shown in Fairchild's datasheet, but unless they're being sold as Fairchild, who's to say the datasheet is correct anyway.  It could be another manufacturers J201 with differing characteristics.   As an example I've got datasheets for the 2N5457 made by both Fairchild and Central Semiconductors and a number of the characteristics are different.  And just to make things even more complicated so is the pinout!

When used as a gain stage it should be less of a concern because you should always be able to get them to bias with trimmers which are invariably included in the layouts (or could be easily incorporated).  So something like the J201 that you don't see used in phasers etc will be less of an issue.

If I wanted to buy some now, I'd be tempted to ask the seller first what manufacturer they are and say that it is important that they are in the datasheet's stated tolerances and that you'll require a refund if they're not.  If he tells you they're Fairchild and they do turn out well out of spec then he can have no complaints.

[GoranP]

All valid points... I'm looking to build a Menatone Howie (6xJ201) soon and I'd hate for it to turn out dud(ish) like WB65.

I gotta build me a jfet tester, either way...

[mirosol]

Felt like i needed to chip in here. I've got now at least two batches of counterfeit TL072s and LF347s from ebay seller gc_supermarket. Avoid the seller. http://www.ebay.com/usr/gc_supermarket

The orders for these were made a long time ago. In cases like these, it's really not much help to have ebay protection for the purchase, if you're going to notice the goods are couterefitted s**t later on. Sadly i noticed this just a few days after placing another order...

[Frank_NH]

This is interesting.  It seems odd that someone would go through the trouble to fake an integrated circuit, especially one as common as the TL072.  Is there that much money in it before you're found out??  Same with JFETs.  99.9999% of the population would think that a JFET was a new furry toy sold by Nintendo.  is there really any money in faking them?

I would think that perhaps the unscrupulous sellers have gotten a hold of bad, defective batches and are trying to sell them as good.  Anyhow, we need to keep lists if the bad eBay sellers.  So far, nearly all of my purchases have been good.  The only bad one was a seller who send me some ICs in an envelope with no protection at all.  Thus their legs were smashed upon arrival...

[rocket88]

I'm really shocked. I've bought a tone of IC's from those guys and every single one was good and spot on. I wonder if it was a bad batch they got or something. But that's still not an excuse. This really sucks that we really have to start looking out for fakes like this.

[dbat69]

I did read on another forum about another seller to be beware of  ** greenhouse408 **  I haven't had anything from this seller so I have no personal knowledge of whether the other comments are true or not.  I just try and avoid him.
I'm sure I have had stuff from gc_supermarket but they are mixed in with parts sourced elsewhere, unfortunately my inventory doesn't go as far as keeping track of where I sourced parts or keeping them apart.

[induction]

It's important to remember that IC's and jfets are two different things, specs-wise. Swap one op-amp for another of the same type, and the result should be damn close to exactly the same in any circuit. Most of the time they don't even have to be they same type, and the circuit will still be fine. The tolerances for most IC characteristics are pretty tight. That's why they were invented in the first place. If you replace one 4558 with another (assuming both are of recent manufacture), there shouldn't be much noticeable difference unless one or both is 'fake' (aka broken).

But the allowable variation in jfet specs is huge. Just because a J201 doesn't sound right in a given circuit doesn't mean it's fake or even outside of spec. This is especially true for Wampler circuits, which are designed around a specific set of J201 characteristics. Any jfet that biases properly in a Black 65 will not bias properly in a Plexi-Drive, Tweed 57, Pinnacle, etc. All of those circuits call for J201's and you can't swap J201's from one circuit to the next and expect it to work. Trimmers can be used to bias them, but it will change the gain and the circuit won't sound like the original. I think many (but not all) of the claims of counterfeit jfets is due to a misunderstanding of the tolerance in jfets and the need for sorting them for the circuit you intend to build.

[Frank_NH]

Your comment about JFETs (especially J201s) is spot on, which is why going through the trouble of measuring Vp and Idss for all your JFETs is worthwhile.  I did come across a bad one when working with my Britannia circuit.  Good thing I use sockets for all transistors and ICs, otherwise removing and swapping parts would be a major pain.

Counterfeiting implies to me that someone thinks they can make some money by making fake but realistic-looking transistors.  If I were going to fake something, I'd think about Rolex watches or rare coins - not transistors!  However, it is reasonable to suspect that someone would get a hold of a bunch of rejected parts and try to pass them off as functional.

[mirosol]

Once i get that new order and a Tayda order with same types, i'll be taking photos.

Noticed that something was wrong with those LF347s when i got a "ghost distortion" in the background. For three separate circuits. This distortion was gone when i replaced the chip with a one i got from Tayda. The apearance of these two chips are not even close to each other, TI logo being at the wrong spot, marking dent looks nothing like the original etc.

I'll get back to you in couple of weeks...

[rocket88]

This is pretty upsetting miro. I think I might have to got through all my ICs and make sure I didn't get any fakes. Was that the only chip you had issues with that may appear counterfit or were there others you got from them?

[dbat69]

Miro
I got mine (LF347s) from the UK company - Anglia-Live

http://www.anglia-live.com/products/kw/lf347/200430001_14P-QUAD-J-FET-OP-AMP

They can be very reasonable for some things, but you need to spend £35 for free delivery (not sure about international delivery though)  However, it is too easy to spend £35  

They don't hold stock for everything, but some things are even cheaper than eBay China sellers !!

[mirosol]

It isn't just a single chip.

These are the earlier ones i got from the G&C, compared with a chip from Tayda (which is genuine):
http://mirosol.kapsi.fi/varasto/tmp/LF347/LF347-p1.jpg
http://mirosol.kapsi.fi/varasto/tmp/LF347/LF347-p2.jpg

And these are the chips i got from G&C today, compared with the same Tayda chip:
http://mirosol.kapsi.fi/varasto/tmp/LF347/LF347-1.jpg
http://mirosol.kapsi.fi/varasto/tmp/LF347/LF347-2.jpg

Here's a clip i recorded:
http://mirosol.kapsi.fi/varasto/tmp/LF347/LF347.mp3

The circuit is Madbean's Boneyard with a DIP-14 socket for the quad opamp. The first three chords are with G&C chip and the latter three with Tayda chip. The same exact "ghost distortion" is present with every single G&C chip from botth batches, so it cannot be simple bad batch. The visual characteristics of the chips differ between the two different G&C batches (and they look nothing alike the genuine article), so yes. There are at least two counterfeit models out there. The ones i have are marked as 1CA29GM and 13WYH8M. Both counterfeits.

Just sent them a message on ebay that i want my money back and asked them to please, stop selling counterfeit opamps.

If i had any energy left i should do a warning image with seller's logo and comparisons. Just to see how far i could spread it over social networks. Way too tired and disappointed about human race right now.

[mirosol]

Not to mention none of the TL072s i got from G&C earlier worked right on a Tremulus Lune. I need to set up a breadboard and test all the other s**t i've bought from this seller. Sooner or later...
+m

[Frank_NH]

Wow.  Still I wonder if the seller got a hold of a bunch of bad ICs from many sources and is now trying to sell them on eBay.  Fake, to me, implies something that looks like the real thing but in fact doesn't even work (i.e. no working inner parts).  In your case, the IC "works" but has some internal defect which causes the ghost distortion.

I wonder if you could contact Texas Instruments about this and see if they know of any bad IC batches for this particular chip.  

[mirosol]

TI customer service seems to be for corporate customers. Rather off putting experience, for the whole site.. I'll see what i can do.

I'm not saying "fakes", but counterfeits. If the chips were used for driving LEDs, the chance of anything out of the ordinary happening would be minimal. I'm guessing these are rebranded as LF347, while they are in fact some very abysmal quad opamp.

I'll try contacting TI. I'm fairly certain these chips aren't from them.
+m

[rocket88]

That's fucked up. You know miro, I think they have a place in nj near me. I might be able to get a number for there, if I'm right, and be able to help out a bit. Let me know if you want me to look into it.

[mirosol]

Just emailed ask-TI email address. Let's hope they'll answer.

Like i said. By the look of the chips, i'm not convinced TI has anything to do with them. We'll see..
+m