Bargain chips bought online - buyer beware!

[BigEd]

We've had a few discussions recently with a common theme: someone has bought some 6502-family chips online and is concerned that they might not be right.

This is by way of a heads-up: as a buyer, you need to beware! If you buy from a high-reputation seller, you will very probably get the product you expect (and if not, you will probably get a replacement without much bother.)

But if you buy from any other seller, you might get anything. I reckon there's a fairly good chance that you'll get a 6502 if you tried to buy a 6502, but it might not be the speed grade you expected, it might be NMOS or CMOS regardless of what you expected, and it almost certainly will be pre-loved. That is, it will be a chip salvaged from historical equipment, possibly cleaned-up and possibly remarked, and probably not tested.

It may be that the marketplace you used has some kind of reputation system for sellers: if it does, use it! Choose a seller with a higher reputation, and always claim for a replacement if the parts you bought are not what they should be.

You will have your own attitude to risk and reward: you might like the excitement of a lottery, you might have a high preference for spending the minimum amount of money, you might like the adventure of testing a chip to see what it is and how fast it goes (and what temperature it runs at.)

If you prefer to get exactly what the description described, please buy from an official supplier, and be prepared to pay the price, the tax, the shipping, the import duty if applicable, and maybe even the handling charge.

If you prefer to take a risk, be prepared to do some diagnosis and be ready to reverse the transaction.

Please don't treat the suppliers as fraudulent, or describe the whole marketplace or an entire nation as being corrupt: you were playing by their rules, not your rules, and you were aiming to spend infeasibly small amounts for something that mattered to you. The seller may have no idea what a 6502 is, or what the different types are, or that 40 pin packages contain a variety of very different chips. Or they may be very capable and honest but have a bad supply chain. You dealt with someone you don't know, and you accepted their description at face value, and by some reckonings that makes you an ideal customer for sharp practice.

For an interesting take on how the supply chain can work, see this great post on Bunnie's blog. He's literally written the book on doing business in Shenzhen.

Edit: perhaps see also 65xx parts sources, genuine and fake

[BigDumbDinosaur]

One key to avoiding a rip-off is looking at the date codes and considering them in terms of the product that is purportedly being sold. For example, production of Rockwell-branded 65C02s ceased some 20 years ago. If, for example, a supposed Rockwell 65C02 has a datecode of 0421 it is a fake, as no date code after 99xx would be valid.

Similarly, if the seller is claiming the part is a genuine WDC product be aware of how WDC numbers their products. If you aren't sure about what you are getting please post a clear photo of the part in question so we can look at it to see if the combination of part number and date code makes sense.

[BigEd]

(BTW, I was aiming here for a post which could stand as useful advice for a while, perhaps even as a sticky post. Hopefully any comments can support that position. Otherwise, we'll continue to have the same discussions in many places every time the issue comes up.)

[cbmeeks]

What sucks is that several years ago, I bought a bunch of sound chips and video chips from the cheapest sellers on eBay that I could find. I literally have dozens of these chips.

Being naive at the time, I had no idea they could be fake. They looked real enough to me.

So I have this gut feeling that if I were to test them, many of them would be fake or at least different than what I thought I was getting.

Some of them have worked as I have used them. So I guess I need to sit down one day and try to find out how real they are.

[cjs]

So, from my reading of things like Jed Margolin's fabulous and fascinating Atari Email Archive (of messages at Atari coin-op in the 1980s), it seems that this "oops, got parts that were not exactly what I thought they were" thing is is not just an issue for those of us cheaping out and going to random vendors on AliExpress, but is an integral part of working with hardware.

The Atari coin-op division had an Approved Vendors List (AVL) listing all the parts approved for use in manufacturing their video games, including acceptable substitutions from different vendors. To get a part on this list it would have to be qualified via a process called a Component Evaluation Request (CER). Yet even so, this did not always ensure that parts would work, as noted in this message from Ross Cox:

Quote:

Subject: USERS OF 16K X 4 DRAM BEWARE!!
Date: August 01, 1984 14:40; From: KIM::COX; To: @SYS$MAIL:ENGINEER

RAM USERS BEWARE!!!

The Fujitsu version of the 4416 (16K X 4) DRAM as qualified under Atari part number 137323-001 will not work in a board with a RAS/CAS generator designed like the one on Crystal Castles. The CAS HIGH to RAS LOW delay time on both the TI and INMOS parts is specified as 0ns, while the Fujitsu part needs 30ns for the 150ns version. On Crystal Castles, the CAS signal clears an S74 to generate the leading edge of RAS. No way can it work.

I have asked that Fujitsu be disqualified for 137323-001.

The problem of 1980s parts being hard to get is not just one we have now in the 2020s:

Quote:

Subject: Battery RAM
Date: September 17, 1985 13:31; From: CHARM::AVL; To: KIM::MARGOLIN

I've assigned Atari P/N 137442-150 for the 'Zero Power' RAM chip on cer#3842.
The only vendor listed is Mostek #MK48Z02B-15.
If you need a copy of the CER or data, come on down and grab it.

Quote:

Subject: The fabulous battery operated static RAM.
Date: October 21, 1985 11:51; From: ERNIE::ARVIDSON; To: KIM::MARGOLIN

Jed,

Sorry to inform you that I have just removed Mostek as the source
for the 2K x 8, 150ns, static RAM with battery, our part number 137442-150.
As you may have seen in the Mercury News, Mostek is going out of business.

Erwin.

And, of course, there are constant and deep price and performance analyses. After all, a good engineer and a great engineer both produce a product that works just fine; the difference is that the great engineer does it for half the cost.

cbmeeks wrote:

What sucks is that several years ago, I bought a bunch of sound chips and video chips from the cheapest sellers on eBay that I could find...
So I have this gut feeling that if I were to test them, many of them would be fake or at least different than what I thought I was getting.

On the other hand, I ordered CMOS "65C02" parts from AliExpress, even though I wanted NMOS parts (in the hope of trying out the undocumented instructions one day). Lucky me, half the "65C02" chips were actually NMOS chips! Bonus! :-)

[BillO]

I have to agree with Ed. While I have yet to get chips that are not what they purport to be, I have had other issues dealing with eBay (not tried AliExpress yet), like wrong number of parts, lost in the mail, and other things. However, I have only once had to go to eBay for a resolution. Every other time I've had an issue the sell was more than happy to immediately refund my money or re-send the item. In my experience (about 1400 purchases from eBay) I have never found a bum vendor from the far east.

I did have a US vendor that would not help, but it was understandable. In this case it was a $3200 trombone that came to me from Hawaii. It came damaged. It was shipped via eBay Global Shipping so all I had to do was put in a claim with eBay and they, with little effort on my part, refunded the price of the trombone and the shipping. I was quite happy for them to pay to get it fixed, but they did not even ask me to get a quote (which was only $132, BTW), they just gave me back to whole wad. So I got a $3200 trombone for $132. Go figure.

So, so far I've got no complaints WRT buying parts on eBay. I always go with highly rated sellers over the lowest price, and that may have a bearing on my experience. However, they are still very reasonable price wise and have never been difficult to deal with.

Now, the postal services, I've had issues there, but the vendors usually make good regardless.

[1024MAK]

Another problem with buying CMOS chips (which are sensitive to static discharge) from the likes of eBay, is that you have no idea how the chip has been handled.

And then there is the packaging. For example look at this:

Although novel and space saving, there has been no proper regard for protecting the chips from static discharge during handling, packing, shipping and when they get to you, how do you unwrap them without the risk of a static discharge?

Mark

[Chromatix]

I would try taking a syringe and injecting ordinary tap water into the package. That should be conductive enough to suppress discharges while unwrapping. Don't use distilled water for this - it's an insulator itself.

[BigDumbDinosaur]

1024MAK wrote:

Another problem with buying CMOS chips (which are sensitive to static discharge) from the likes of eBay, is that you have no idea how the chip has been handled.

And then there is the packaging...

Geesh! Anyone who packages chips—CMOS or otherwise—like that should be taken out back and shot.

Quote:

...how do you unwrap them without the risk of a static discharge?

You don't.

[GARTHWILSON]

I would cut a slit in each end and blow my humid breath through, then unwrap it rather normally.  Virtually all CMOS parts have static-protection diodes which can discharge small amounts of static, without damage.  They have to be tiny so they don't add a substantial capacitive load.  I doubt the scavengers in China exercised any kind of static-handling precautions though.

[BillO]

1024MAK wrote:

... how do you unwrap them without the risk of a static discharge?

Unwrap them under water.

[GARTHWILSON]

Link to Atlantis's topic "Identifying what 6502s from Aliexpress really are":
viewtopic.php?f=4&t=6027