BUF634P Silk Screen Differences???
Mar 8, 2006 at 7:08 AM Thread Starter Post #1 of 14

gtp

1000+ Head-Fier
Joined
Sep 2, 2005
Posts
1,068
Likes
11
Hi,

I bought three Buf634P for my Pimeta on EBay. When I tried them out, I found that one was dead. Hmmm, strange...so I bought a replacement from a HeadFi member. Tonight I noticed that the Ebay ones have a different silk screening than the one from the HeadFi member (who is well-known and trustworthy). The latter has the same silkscreen as every other picture I've seen:
-----------------
|........BB(R)......|
|.....BUF634P.....|
|.....33ZJ3Y7.....|
-----------------
The Ebay ones have a silk screen that looks like this:
-----------------
|..BB(R)............|
|.....BUF634P.....|
|..0050.....5627..|
-----------------
So as you can see, the location of the BB logo is different, and the ID numbers are totally different.

The question is: Why are the silkscreens different, are these from different factory sites or are the Ebay ones possibly fake???

Thanks,
George
 
Mar 8, 2006 at 5:37 PM Post #2 of 14

tangent

Top Mall-Fi poster. The T in META42.
Formerly with Tangentsoft Parts Store
Joined
Sep 27, 2001
Posts
5,969
Likes
56
Call TI. They'll be able to tell you whether what you're seeing is normal.
 
Mar 8, 2006 at 7:32 PM Post #4 of 14

gtp

1000+ Head-Fier
Joined
Sep 2, 2005
Posts
1,068
Likes
11
Thanks for this. So, according to that, the Ebay BUF634P's DO NOT have appropriate markings!!!!!! Has anybody else experienced anything like this???? Do fake BUF634P's exist????

I started investigating this problem when I found that my Pimeta sounded thin and tinny in comparison to my Cmoys! All have G=3, run on identical 9.6V batteries. I swapped around opamps and still found the same result. Then I remembered the DOA BUF634P and started to wonder if something was fishy. I am going to email the seller.

Quote:

Originally Posted by tess
If you go to TI's web site you can view there Top Side Markings, look under packaging.

http://focus.ti.com/docs/prod/folder...4.html#symbols



 
Mar 8, 2006 at 7:50 PM Post #5 of 14

firefox360

500+ Head-Fier
Joined
May 23, 2005
Posts
765
Likes
10
I'm guessing that it's plausible that someone would make fake BUF634's. It's a highly sought after part in the audio DIY community for one, and it's not an exactly cheap part either.

P.S. Those numbers on the bottom of a genuine TI/BB chip will always consist of letters & numbers. The numbers are suppose to represent the date the chip was manufactured, as well as some other details of the chip. Chances are, you got scammed.
 
Mar 8, 2006 at 9:45 PM Post #6 of 14

Xakepa

500+ Head-Fier
Joined
Sep 2, 2005
Posts
751
Likes
0
I wouldn't freak out. I bough from Digikey 3 634P that have no white silk screen, but rather dark brown inscription with "BB" (not a logo) on the second line. At the same time I got from Ebay 2 634P with nice white silkscreen and BB logo on the first line. All chips work fine, have the prescribed quiescent current and infinite resistane b/w their legs.

I'll be pretty surprized if the seller is not willing to replace yours.
 
Mar 9, 2006 at 1:20 AM Post #7 of 14

mono

1000+ Head-Fier
Joined
Jan 10, 2005
Posts
1,412
Likes
11
Considering the potential for other Head-Fi members to buy BUF634 from ebay, "IF" it turns out yours are fake, it would be useful to know who the seller was so they can be avoided.
 
Mar 9, 2006 at 1:43 AM Post #8 of 14

gtp

1000+ Head-Fier
Joined
Sep 2, 2005
Posts
1,068
Likes
11
Quote:

Originally Posted by mono
Considering the potential for other Head-Fi members to buy BUF634 from ebay, "IF" it turns out yours are fake, it would be useful to know who the seller was so they can be avoided.


Fair enough. If I can verify that these are fake, I will post the name of the seller.
 
Mar 9, 2006 at 6:40 PM Post #10 of 14

tangent

Top Mall-Fi poster. The T in META42.
Formerly with Tangentsoft Parts Store
Joined
Sep 27, 2001
Posts
5,969
Likes
56
Quote:

Originally Posted by Xakepa
I bough from Digikey 3 634P that have no white silk screen, but rather dark brown inscription with "BB" (not a logo) on the second line.


That's a laser etching process. It's usually a better indicator of a genuine chip than a silk screened label, since laser etching costs more to do. Manufacturers feel that pinch, too, though, so they generally only do laser etching on their higher-end chips.

That said, I've recieved even OPA627s and 637s from Digi-Key with both laser etching and silkscreening, so I'd doubt that TI is consistent on this with any cheaper chips, either.

Quote:

Originally Posted by z2trillion
The differences could just be from TI standardizing the silkscreen on burr brown chips after TI bought them.


Um, no. That was completed over five years ago. Surely the two entities are about as well merged by now as they're going to get. Unifying manufacturing would have been one of the first things to be finished.

Far more likely, we're looking at one consequence of the fact that TI probably has multiple fabs in multiple countries, and makes chips at whichever one happens to be least busy at the moment. We've already seen from the BUF634 and TLE2426 shortages in the past few years that they don't continuously manufacture the chips we use. They're done in batches.

Speaking of shortages, that's one point in favor of the relabeled chip argument. Shortages increase the incentives to try this.

Someone above brought up a good point: try measuring the chip's current while you change the bandwidth resistor value. I don't know of any other chip with this pinout with that behavior. If you get reasonable measurements, I'd say that would mitigate heavily against it being a fake chip. Relabeled chips aren't manufactured clones, usually, they're some other kind of chip that's close enough to function in the same circuit. If one were to decide to fake BUF634s, I'd imagine you'd start with a cheap 1-channel op-amp with no function on pin 1, so it'd just drop into the circuit. I guess they'd have to close the feedback loop somehow, too, but that's still easier than fabbing true clones.
 
Mar 9, 2006 at 6:48 PM Post #11 of 14

MASantos

Headphoneus Supremus
Joined
Dec 21, 2004
Posts
2,436
Likes
38
WHen I got the OPAs for my PIMETA and for future amps I bought OPA627 and 637 in both AP and BP versions. There were also differences between the silkscreen in these and they came from digikey. So I would agree with Tangent.
 
Mar 9, 2006 at 7:10 PM Post #12 of 14

gtp

1000+ Head-Fier
Joined
Sep 2, 2005
Posts
1,068
Likes
11
tangent said:
try measuring the chip's current while you change the bandwidth resistor value. I don't know of any other chip with this pinout with that behavior. If you get reasonable measurements, I'd say that would mitigate heavily against it being a fake chip.
Yep, good idea. I'll try this as soon as I have free "hobby" time (but who knows when that will be, hahaha).
 
Mar 9, 2006 at 7:20 PM Post #13 of 14

gtp

1000+ Head-Fier
Joined
Sep 2, 2005
Posts
1,068
Likes
11
Quote:

Originally Posted by z2trillion
It would really help if you could post some pics of two buf634s. The differences could just be from TI standardizing the silkscreen on burr brown chips after TI bought them.


I can't quickly/easily take pictures and post, but here are examples I found from other auctions. This one is what every other BUF634P looks like (obviously with OPA2111KP replaced by BUF634P):
http://cgi.ebay.com/Op-Amp-TI-OPA211...QQcmdZViewItem
and here is one with the silkscreening that I got, with the strange BB location and strange ID numbers:
http://cgi.ebay.com/Op-Amp-TI-OPA235...QQcmdZViewItem

Hopefully this gives a pictoral representation of what I was trying to draw before.

When I have time, I'll try to do the current testing to verify/disqualify any assumptions/questions put forth in this thread.

Thanks,
George
 
Mar 9, 2006 at 9:21 PM Post #14 of 14

pmillett

Pete Millet
Sponsor: TTVJ/Apex Hi-Fi
Joined
Feb 10, 2002
Posts
211
Likes
55
Don't sweat it. The silicon inside the part could have come from one of two or three fabs running that process, and then it could have been packaged in one of 3 or 4 assembly sites, each of which probably has 2 or 3 different flows for a DIP8 package. Throw in the fact that the two parts probably came out of fab 3 years apart, and it is highly unlikely that the top marking would be similiar.

It's highly unlikely that somebody would "fake" a BUF634. This is a very low volume IC, and it would take an investment of 100's of thousands of $$ to clone one to sell a few thousand parts. No return on investment there.

Pete
 

Users who are viewing this thread

Top