When this ain't THAT

[boji]

Had the dumb idea to buy some 'discount' 1646's on ebay.
DRV134's facelift? Wish I had better gear to test.

Opinions, anecdotes?

Legit 1646 on left. Questionable 1646P on right.

Case on left is clean, as well slightly thicker. Note pin 1 indentations.

lines under factory / batch code?

Higher Rez below:
imgur.com/NidQxun
imgur.com/75mWHX2
imgur.com/gIeC7x5

 

[JohnRoberts]

There are some auto makers getting burned by counterfeit parts since that Japanese factory fire (Renesas) broke the auto industry supply chain.

I wonder what the parts in those packages really are. You might be able to make an educated guess with VOM compared to common dirt cheap parts.

JR

 

[john12ax7]

The big giveaway would be how they handle unbalanced loads, 1646 does a better job than most.

 

[ruffrecords]

My first reaction to that was why would you buy specialist chips like that from eBay when they are readily available from reputable sources? But then I just bought ten LM3915 LED VU chips from eBay so who am I to pass judgement? (my excuse of course is that they are no longer manufactured).

Cheers

Ian

 

[PermO]

When I was looking for one of those I also ran into a Aliexpress offer selling them for 3,- euro

I got the expensive original one from Farnell.

The THAT chip will give you 6dB of gain, maybe check that ?

 

[Gustav]

My last from two different sources.

 

[boji]

The big giveaway would be how they handle unbalanced loads, 1646 does a better job than most.

+1 Blown a couple DRV's patching on the fly. Not so with the 1646's.

 

[JohnRoberts]

That corp is pretty big on protecting I/O pins with clamps, et al.

JR

 

[Bo Deadly]

If they're fake I would think they're just a copy of whatever circuit since DRV134 are actually more expensive than THAT 1646. At least on Mouser they are. In that case I think the proof would be in the resistor matching. For that you could run the CMMR test and compare to ones known to be authentic:


Or you might be able to measure a resistance between various terminals and then compare. It's hard to say if a conventional resistance measurement can be meaningful but from looking at the schematic of DRV134, it seems like the resistance between output and corresponding sense should match the other pair. Meaning the resistance between pins 1 and 2 should match the resistance between 7 and 8. If they match closely in several authentic parts and not as closely in the Ebay sourced, that would be a clue that they're not authentic.

Another possibility is that they are not authentic and perform perfectly fine in which case I would say use them. I have no problem using secondary sourced parts. You just have to test them whereas with a primary source you have much more confidence that they will perform as specified in the datasheet.

 

[boji]

pins 1 and 2 should match the resistance between 7 and 8. If they match closely in several authentic parts and not as closely in the Ebay sourced, that would be a clue that they're not authentic.

That's easy enuf. I'll report back.

Another possibility is that they are not authentic and perform perfectly fine in which case I would say use them.

Some of these cheaper INA/DRV mic pre's don't like being patched directly into desk's 2520 diff. line inputs. I'll test em' with 1240 / 1646's (legit & suspect) and see how they do.

 

[Ricardus]

My first reaction to that was why would you buy specialist chips like that from eBay when they are readily available from reputable sources? But then I just bought ten LM3915 LED VU chips from eBay so who am I to pass judgement? (my excuse of course is that they are no longer manufactured).

There have been times when all major suppliers have been out of 1646s during the pandemic. So looking elsewhere is something I often do. That said I would never buy semis off eBay. Got stung once. That was enough.

I bought a bunch of LM3915s myself two years ago for my SSL9000 preamp builds. But I have a few trusted NOS suppliers I go to when I need stuff like that.