Got some bum 6502's

bit_slicer

New member
Joined
Oct 17, 2007
Messages
2,984
Reaction score
17
Location
Florida
I picked up a batch of 6502A's from a known source (which will go unnamed) and out of the 11 I received 4 were bad. :( Not a good ratio. These were the R6502P parts made in Mexico. Anyone else have a similar experience recently?
 
Not sure if yours are the same that I picked up from a known source, but I think I only found one bad one in my batch of 20. Although mine are R6502AP made in Mexico.
 
Man, nothing sucks more than getting a bum part that you fully expect to be working. It can really eff up the troubleshooting process and send you on a wild goose hunt.
 
Bad as in dead or bad as in flaky? I just put one in a board I'm working on and it is very flaky right now. I thought it may be other problems, but I'll pull it and see where the chip was made.

Can you PM me the supplier to verify the source.

Thanks,

-JM
 
Man, nothing sucks more than getting a bum part that you fully expect to be working. It can really eff up the troubleshooting process and send you on a wild goose hunt.

Tell me about it. I didn't get as far as replacing any chips, but I did scope out the watchdog circuit (it was fine - toggling the reset line as it should), the sync circuit, then noticed that SYNC wasn't firing. I traced it back to the 6502 and then just happen to notice that A(0) wasn't toggling. WTF??? Swapped in a used 6502 from another board - came right up! D-ohhh.

Yeah these chips were completely dead - no outputs toggling. If I have time this week I'll take them into work where they have the x-ray machine. It will confirm or deny my suspicion.
 
Worst part about it is I'm probably out $20 of overhead for the month already. :(
 
Yeah these chips were completely dead - no outputs toggling. If I have time this week I'll take them into work where they have the x-ray machine. It will confirm or deny my suspicion.

...just curious...What are your suspicions?

Edward
 
Counterfeit parts.

You know, if I hadn't seen evidence of this before, I'd never believe that anyone would go to all the trouble of counterfeiting one of the most common processors ever.

The subject I read about, however, was the SID chip, as used in the Commodore 64. There was a site detailing counterfiet SID chips. In this case, they were simply known defective chips repainted with a different manufacturer code.

Again, IMHO, it seems like an awful lot of work for not much gain... like making counterfeit dimes.

-Ian
 
Well not counterfeits. :p Just for grins I took the chips over to assembly and had them x-rayed. It was funny, there was a younger gentleman there noodling about with something mundane. I approached the subject cautiously, can we x-ray some parts? His eyes got wide, "yeah! It's over here let's go!" LOL Like a 16-year-old kid with a new drivers license. Any opportunity to use the x-ray machine! :D It was cool - this was the first time I've seen one in action. Basically it's a big lead box with a blast door and fancy camera inside mounted above an XY table. He put the chips inside the door, closed it (woosh), and turned on the x-ray:

Bad part:

chip4.jpg


Good part:

chip5.jpg


So yes the die is there on all of them. :p Some are just plain broke.
 
You know, if I hadn't seen evidence of this before, I'd never believe that anyone would go to all the trouble of counterfeiting one of the most common processors ever.

The subject I read about, however, was the SID chip, as used in the Commodore 64. There was a site detailing counterfiet SID chips. In this case, they were simply known defective chips repainted with a different manufacturer code.

Again, IMHO, it seems like an awful lot of work for not much gain... like making counterfeit dimes.

-Ian

http://www.stellacoinnews.com/index.php/mercury-dimes-ch-4-counterfeit-altered-coins/

You'd be surprised. :)

As to the images of the dies, it's hard to tell w/ them (i'm assuming based on the long line of contacts and 2 points going down the next axis on part of it) flipped relative to each other so they're not showing from the same orientation/angle. The odd thing is that it almost seems like there's more contacts on the known working die but there's not as many connected to the traces. The pads (or whatever they are) on the good one look a hell of a lot larger in places, may just be sketchy manufacture and they're not making a good connection. Certainly differences, though I'm not versed enough in chip architecture to know where the issue is looking at it, hopefully someone will pipe up with an explanation.

*e* If you could flip one 180 degrees and edit it back in, that'd work w/o needing a bunch more pictures. If you don't have a graphic editor handy let me know and I'll take care of it real quick.
 
Last edited:
I've got more images of bad ones if it would help. I just didn't want to inundate the thread with large images. They are all almost certainly flipped 180 degrees. I also just noticed one of the bonding wires crossing an adjacent wire on the frame. Probably not making an errant connection, or maybe so? :)
 
I also just noticed one of the bonding wires crossing an adjacent wire on the frame. Probably not making an errant connection, or maybe so? :)

The Xray's are TOO COOL ! I think with that picture it might be giving a distorted view of the bond wire, it may actually be clear of the pad as from the images I've seen the bond wires were in an arc from the chip to the lead. So when looked at vertically you can't see the clearance. I'm sure we'll now discover there's a pad-bonding expert on the board who can comment :)

- James
 
I noticed the bad one doesn't have a bonding wire connecting to the "plane" (ground? maybe?) that surrounds the die.

Edward
 
I don't think the one on the top is actually a 6502 chip. A real 6502 has pins 35 and 36 as N/C. This is clearly visible on the bottom where there are two contacts on the leadframe that aren't connected to the die. The one on the top has all leadframe pins connected.

-Ian
 
All of the chips were in 6502 packages. One thing is for certain, the manufacturer clearly failed to test these for quality control. Having a 30-40% failure ratio is inexcusable for the manufacturer.

BTW, I did contact the vendor and the issue is all squared away.
 
Brand new. R6502P, 11365-12, Mexico, 0023, A22336-3. I think the 0023 might be the date code?
 
Counterfeit chips are a big issue these days. You have to be real careful buying overseas especially on the more expensive stuff like this. Most of my ICs come from Bulgaria and Holland from my dealer who has trusted contacts there. Luckily no issues yet.
Hope you didnt pay too much for the lot.
 
No it wasn't me. I got these from a vendor. If anything he's out the $$d'oh.

If they really were counterfeit, I would think the chip carriers would be empty. Why put a wafer in a carrier you knew wasn't going to work anyway? I think they're just bad. BTW, I did check the other 'bad' pictures and the rest had bonding to the largest wire in the frame - GND I guess. I can post the rest of the pictures if anyone is really interested.
 
Back
Top Bottom