Counterfeit MC34063A - Opinions

[rfdes]

Awhile back (year or two ago) I purchased a handful of MC34063A ICs from a reputable dealer (or so I thought).  This week while trying one of them in a design, it was acting super noisy.  Looking closely at the IC laser etching I see a date code year of 1991 and the logo appears to be ON (for ON semiconductor).  ON semiconductor was not around in 1991 per my research.

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.
 

[coppice]

Who would counterfeit a part that's only 3 cents for a genuine one?

[wraper]

Who would counterfeit a part that's only 3 cents for a genuine one?

Why this question gets asked every time a thread like this appears? Like were you born last night?

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.

100% counterfeit. Sanded, blacktopped and new laser marking applied. Try rubbing it with acetone or other similar solvent. Most likely marking will go away.

[wraper]

Also logo looks to be DN rather than ON.

[Stray Electron]

   The very first question that needs to be answered is; where did you buy the part?

  FWIW EVERY IC that I've tried to buy from anyone on Ebay, etc has been counterfeit.  The problem is so bad that IMO anything bought from anyone other than a very well established dealer like Digikey is likely to be counterfeit.

[rfdes]

Also logo looks to be DN rather than ON.

Is there a company with the logo DN?

[rfdes]

   The very first question that needs to be answered is; where did you buy the part?

  FWIW EVERY IC that I've tried to buy from anyone on Ebay, etc has been counterfeit.  The problem is so bad that IMO anything bought from anyone other than a very well established dealer like Digikey is likely to be counterfeit.

Honestly, I don't recall where I bought them.  Certainly not Digikey, Mouser, etc.
Bitten again.  Oh well, live and learn.

[wraper]

Also logo looks to be DN rather than ON.

Is there a company with the logo DN?

Obviously it's supposed to imitate ON logo. There are fake ST vregs with logo looking like S7 if you look closely. Maybe they do it for legal reasons, who knows.

[Mechatrommer]

Awhile back (year or two ago) I purchased a handful of MC34063A ICs from a reputable dealer (or so I thought).  This week while trying one of them in a design, it was acting super noisy.  Looking closely at the IC laser etching I see a date code year of 1991 and the logo appears to be ON (for ON semiconductor).  ON semiconductor was not around in 1991 per my research.

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.

i was just curious how your circuit looks like? and what noise are we talking about? 100mVpp? 500mVpp? 1Vpp? afaik, no smps ic can perform very well when you give it wrong components. ymmv.

[magic]

I found NXP NE5534 even thought NXP was spun off from Philips only after Philips discontinued this IC. They were recycled equivalent TI chips with original markings sanded off and replaced.

Your chip doesn't seem to have its marking altered, so most likely it's a Chinese 34063 substitute which was never marked with its true manufacturer's logo, or maybe the manufacturer is indeed "DN Semiconductor", everything is possible in China

Same thing happens if you buy LM358 from China, or even several other part numbers which may also contain LM358 inside.
https://www.eevblog.com/forum/projects/whats-inside-the-cheapest-and-fakest-jellybean-opamps/

[rfdes]

Awhile back (year or two ago) I purchased a handful of MC34063A ICs from a reputable dealer (or so I thought).  This week while trying one of them in a design, it was acting super noisy.  Looking closely at the IC laser etching I see a date code year of 1991 and the logo appears to be ON (for ON semiconductor).  ON semiconductor was not around in 1991 per my research.

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.

i was just curious how your circuit looks like? and what noise are we talking about? 100mVpp? 500mVpp? 1Vpp? afaik, no smps ic can perform very well when you give it wrong components. ymmv.

It was greater than 1.5 Vpp on a 5V dc output (configured in buck mode).  This was a proven design and the current demand was only 200mA.  The excess noise made absolutely no sense.

[magic]

Was it noise (random drift of regulated voltage over time) or ripple at the switching frequency? Was the frequency as specified? (IIRC, there is some means of adjustment, maybe you connected something wrong to a wrong pin?)

Did you try a legit chip on the exact same board and it went away?

edit
And by the way, 34063 really isn't a great chip for buck because of fairly high voltage drop of its Darlington switch. And there is no thermal shutdown in the basic version, though there are supposed to exist improved variants which include it.

[Mechatrommer]

Awhile back (year or two ago) I purchased a handful of MC34063A ICs from a reputable dealer (or so I thought).  This week while trying one of them in a design, it was acting super noisy.  Looking closely at the IC laser etching I see a date code year of 1991 and the logo appears to be ON (for ON semiconductor).  ON semiconductor was not around in 1991 per my research.

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.

i was just curious how your circuit looks like? and what noise are we talking about? 100mVpp? 500mVpp? 1Vpp? afaik, no smps ic can perform very well when you give it wrong components. ymmv.

It was greater than 1.5 Vpp on a 5V dc output (configured in buck mode).  This was a proven design and the current demand was only 200mA.  The excess noise made absolutely no sense.

that means you should be very well versed in designing such circuit already, buying parts from legit source and then replace the counterfeit, if it works ok, then its confirmed counterfeit. probing emitter or transistor gate may also reveals poor switch frequency. counterfeit ic is real business not limited to china region, sometime thailand, mexico, italy batch/designation whatever. sometime we cannot brain why they even bother to counterfeit such already cheap ic... maybe one cent profit when multiplied by millions sale is still large profit. the ic design maybe they got free in dustbin from 1977's design  i once bought stuffs from honest seller admitting counterfeit parts (opamp), but i buy anyway because they can work to my spec. cheap is key and verify your counterfeit parts when you buy it for testing in small quantity. good advice? ymmv.

[coppice]

Who would counterfeit a part that's only 3 cents for a genuine one?

Why this question gets asked every time a thread like this appears? Like were you born last night?

Well, what is the answer? Nobody ever has one. Nobody fakes 1 cent coins. So, what is the rationale for faking a 3 cent part, when they could put the label of something more valuable on there. There is some real weird behaviour amongst the counterfeiters.

[wraper]

Your chip doesn't seem to have its marking altered, so most likely it's a Chinese 34063 substitute which was never marked with its true manufacturer's logo, or maybe the manufacturer is indeed "DN Semiconductor", everything is possible in China

I'm 95% sure it's sanded/blacktopped. Does not look right and has sharp edges.

[wraper]

Who would counterfeit a part that's only 3 cents for a genuine one?

Why this question gets asked every time a thread like this appears? Like were you born last night?

Well, what is the answer? Nobody ever has one. Nobody fakes 1 cent coins.

The answer is that they manage to make money on this, probably by cheap labor and sheer volume of this junk produced.

So, what is the rationale for faking a 3 cent part, when they could put the label of something more valuable on there. There is some real weird behaviour amongst the counterfeiters.

Don't worry, they counterfeit those too and provide full assortment from cheap to expensive components .

[coppice]

Was it noise (random drift of regulated voltage over time) or ripple at the switching frequency? Was the frequency as specified? (IIRC, there is some means of adjustment, maybe you connected something wrong to a wrong pin?)

Did you try a legit chip on the exact same board and it went away?

edit
And by the way, 34063 really isn't a great chip for buck because of fairly high voltage drop of its Darlington switch. And there is no thermal shutdown in the basic version, though there are supposed to exist improved variants which include it.

In 2023 the MC34063 makes no sense for purely economic reasons. Even if its free, the cost of the ferrite needed to run at such a low frequency exceeds the cost of a smaller transformer, and a modern chip that can drive it at a higher frequency.

[ejeffrey]

Who would counterfeit a part that's only 3 cents for a genuine one?

Why this question gets asked every time a thread like this appears? Like were you born last night?

Well, what is the answer? Nobody ever has one. Nobody fakes 1 cent coins. So, what is the rationale for faking a 3 cent part, when they could put the label of something more valuable on there. There is some real weird behaviour amongst the counterfeiters.

How much did the OP pay?  Probably more than $.03 total.  They might also make money on "shipping" charges, or selling customer information.  And there are no shortage of scrap ICs, so it's not like they are giving up the opportunity to sell a fake $.30 chip.  They key is probably that it's a well known part number that people search for.  It's entirely possible that the remarked chip is technically more expensive than the MC34063A but if it's a part from a no name manufacturer that nobody has heard of and nobody has a datasheet you can't sell it on eBay at any price.

[Gyro]

...
  FWIW EVERY IC that I've tried to buy from anyone on Ebay, etc has been counterfeit.  The problem is so bad that IMO anything bought from anyone other than a very well established dealer like Digikey is likely to be counterfeit.

Not a 100% rule (but probably damn close), I've had some very useful NOS chip scores. You really need to do your research though, including checking photos carefully, checking date codes against photos in other listings for the same part. Most important is to know your seller, that means checking that the majority of their other items can't possibly have been conterfeited (thinking in terms of electromechanical and other parts), and the number and range of items tells a plausible story. They also need to be very clearly your own countrymen.

[magic]

Nobody fakes 1 cent coins. So, what is the rationale for faking a 3 cent part, when they could put the label of something more valuable on there. There is some real weird behaviour amongst the counterfeiters.

You are already making those 3 cent parts (MC34063, NE555, LM358, LM317, TL431, ...).
You can sell more of them by printing TI logo and listing them on eBay.

What in hell would be the reason not to do it?


That's how I assume it works. I only don't know if it's a semi-official policy of the manufacturers to make such things, or rogue employees doing it, or manufacturers turning blind eye to eBay dealers ordering such custom "artwork" on otherwise legitimate chips, or manufacturers selling blanks to be marked by dealers themselves.

And of course they also fake more expensive stuff. But there is market for fake low-end chips too, so supply meets demand and it isn't that hard to compete with DigiKey on price...

[Mechatrommer]

$0.01 x 1,000,000 pcs = $10,000 is math so hard?

[Gyro]

It seems as if laser marking is a bit of an own goal for the semiconductor industry, every man and his dog seems to be able to do it. It really isn't that bothered about what is / was underneath and it isn't very high res. Screen printing, on the other hand is hard - it's hard to get consistency and clean edges, it's very dependent on the underlying texture, it's hard to get the colour exactly right etc. The OP's photo is a prime example, all those blobs at the ends of the strokes are a dead giveaway. Getting screen printing right (enough) on a worthwhile number of parts is a problem without the high volume printing technology used by the real manufacturers.

[schmitt trigger]

edit
And by the way, 34063 really isn't a great chip for buck because of fairly high voltage drop of its Darlington switch. And there is no thermal shutdown in the basic version, though there are supposed to exist improved variants which include it.
In 2023 the MC34063 makes no sense for purely economic reasons. Even if its free, the cost of the ferrite needed to run at such a low frequency exceeds the cost of a smaller transformer, and a modern chip that can drive it at a higher frequency.

Just like the 555 or 741, the 34063 is so ingrained in the collective EE psyche, that I would not be surprised if 10 years from now it is still being used on one of the World’s most remote corners.

[magic]

Actually, 741 seems to be mostly dead nowadays. I have only seen it once in some cheap soldering station. LM358/324/321, on the other hand, is everywhere.

[WattsThat]

Semiconductors have never been screen printed as it is slow and does not scale down well.

Before laser marking, pad transfer, aka Tampo printing was by far the most common method.

[CJay]

   The very first question that needs to be answered is; where did you buy the part?

  FWIW EVERY IC that I've tried to buy from anyone on Ebay, etc has been counterfeit.  The problem is so bad that IMO anything bought from anyone other than a very well established dealer like Digikey is likely to be counterfeit.

You have been insanely unlucky then.

[Andy Chee]

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.

As others have commented, if true, I'm surprised that there's any value in counterfeiting this particular part.  I myself scavenge this part from e-waste, as shown in the attached picture.

[Mechatrommer]

As others have commented, if true, I'm surprised that there's any value in counterfeiting this particular part.  I myself scavenge this part from e-waste, as shown in the attached picture.

counterfeit usually works so so... and in most low tech products can work within spec, just put bulky caps on. scavenging is different business $0 and maximum time involved, and scavenging counterfeit parts is a part of it.

[rteodor]

I was just curious if anyone had an opinion regarding the legitimacy of these parts.  They look counterfeit to me.

As others have commented, if true, I'm surprised that there's any value in counterfeiting this particular part.  I myself scavenge this part from e-waste, as shown in the attached picture.

It might be just a bad batch that failed tests. Instead of being discarded into a furnace it was "recovered" and sold by some "entrepreneurs".
This sort of things happen all the time.

[Jay_Diddy_B]

Hi,

The only thing that makes this counterfeit is the use of the On Semi logo.

There are many parts that have produced to be pin compatible with the original manufacturers part.

The 555 was developed by Signetics. Any other 555 could be considered a clone. I would not think twice about using a clone from a major manufacturer.

The MC34063A that is shown in the photograph is almost certainly a counterfeit. The On Semi datasheet shows the marking should be three lines:



There are many reasons why the circuit is noisy. Is the diode in the circuit correct? Have you substituted a known good part?

Regards,

Jay_Diddy_B

[GnomeZA]

Who would counterfeit a part that's only 3 cents for a genuine one?

This seems to be a common problem for computer chips too, ie. Intel ethernet cards are practically guaranteed to be fake from eBay unless it has YottoMark or BradyId.  They work, they just don't work very well.

The reality of why it makes sense is as follows (IMO):
You have a "friend" that works for the factory that makes this chip. (in reality probably lots of different chips)
The chips that don't make spec aren't marked and are to be thrown out.
Friend then boxes up the 1000s of chips that didn't make spec, you mark them and sell them on a local market for pure profit.
You kick some money back to friend.

Taking this further, the reason RS and Digikey have all accidentally sold fakes in the past, is likely (again opinion) because when they've bought from sellers who deal in large volume and aren't afraid to compromise their business by throwing a couple of fakes in there to pad the bottom line.

Just like viruses have evolved in many ways, I'm sure there are variations on this theme.

ie. you used to make a chip for OnSemi and they terminated their contract with you (maybe not even because you did anything wrong), so now you sell all the inventory that didn't meet spec to "recoup" your "loses" and so on.

Or Digikey bought from you because they were introduced to you by your uncle (a large reputable seller).  The twist is you only sell fakes and promised your uncle you are legit (or he is in on it or variations).  Later uncle says he didn't know when questioned by Digikey.  But Digikey wouldn't have bought from you without that recommendation but here we are now.

etc. etc.

[magic]

It really is a matter of supply meeting demand. If there is enough people looking for cheap MC34063, somebody will figure out an industrial process to make them cheaply, possibly even cheaper than possible.

You might as well ask why anyone even makes 3 cent parts at all. Would you go through all the effort of manufacturing, testing and packaging an integrated circuit to sell it for pennies? Yet somebody does.