• Please review our updated Terms and Rules here

Fake chips in IC catalogs

paul

Veteran Member
Joined
Mar 18, 2004
Messages
817
Location
New Zealand
Does anyone remember the fake chips that could be found in some IC catalogs?

Other than the Signetics write-only memory that seems to be well-documented, I thought there were others, including some in the famous orange Texas Instruments TTL books.
 
I don't know about mention in any catalogue, perhaps the mainboard manufacturer just had them specially made, but I've got a pentium board with two real cache chips and two fake ones - all surface mount. The fake ones are shiny, and brightly marked "WRITE BACK". This was obviously something that a salesman could point to. What he wouldn't point out is that there weren't any traces leading to these two chips. I decided to use this board in one of my DOS builds because, to my mind, it is a classic.

Edit: I just turned that box on and checked the details. It is a FuguTech M507. I suspect FuguTech was "run outa town" a long time ago. :)
 
Last edited:
The M507 sounds a lot like a PC Chips model, in fact it is. I would be willing to bet that the FuguTech is just a relabel of a PC Chips to sound like a better board.
 
The M507 sounds a lot like a PC Chips model, in fact it is. I would be willing to bet that the FuguTech is just a relabel of a PC Chips to sound like a better board.

Yep, from my research it's a PC Chips board alright. By the way, I think I made a mistake. The other two chips are not cache, and this board does, in fact, have none - unless one can find a very rare "COAST-Socket" module for it. It's quite abnormal in other ways too. Among other things, I found this quote in a review:

You can find solder pads just beside the keyboard connector. So why did they not put the headers on it? AFAIK the used UMC 8663 Multi I/O has a bug, which prevents the use of the keyboard AND the mouse at the same time.

I wouldn't mind finding one of the PC Chips boards with the imitation "VX Pro" chipset. I think these famous historical quirks are quite collectible. There's a million quality mother boards out there, but these memorable ones seem to disappear.
 
I think that several manafucturers produced data sheets for "write ony memory" , most with a 6.3V AC line for the heaters.

In a similar vein there are also one or two 1st April RFCs.....
 
I think that people are getting confused here over the meaning of "fake". I can see three interpretations:

1. Prank

Paul has already mentioned the Signetics 25120 WOM, but it was just a single datasheet, not in any catalog.

Similarly, the NS LM0901A1411090451C is another joke datasheet. TI had a datasheet for a "Dark Emitting Diode", but I can't find it on the web.

2. Non-functional

This meaning pertains to ICs added to a board, but serving no practical purpose or chips speciously lableled. PCChips was famous for this.

3. Chips in databook, but never produced.

This pertains to chips that were announced but canceled. The MC6855 serial DMA controller is an example--announcement in the dtabook, but never produced.

Since Paul mentioned the Signetics WOM (Bob Pease wrote about it in extenso in his ED column some years ago), I'm assuming that Paul intends meaning (1).
 
Thanks for the replies and links, guys ... the EEs at work did not believe these existed.
 
I wouldn't mind finding one of the PC Chips boards with the imitation "VX Pro" chipset. I think these famous historical quirks are quite collectible. There's a million quality mother boards out there, but these memorable ones seem to disappear.

I have a PC Chips VX Pro board in my Cyrix MII machine:

 
Well, I still have and use two Amptron 8600 boards, in more or less continuous use since new. Both are running Win98SE and I have yet to have a problem. One is the 8600C model and the other is the 8600D. I think one is using an IBM 686MX chip and the other is a Cyrix something or the other.

These boards weren't as bad as Red Hill makes them out to be.
 
Thanks for posting that. If you're interested, I found a bit of information about the VX Pro on the redhill site. There's a bunch more on about PC Chips and fake cache chips on that page too.

I have the infamous PC Chips M919 "VIP" motherboard with fake cache in my Cyrix 5x86 machine. Mine actually has the fake "WRITE BACK" chips installed. Later revisions did not, but you could still see the circuit board traces going around in circles between where the the two fake chips would be, and not connected to the rest of the board's real circuitry.

Here is an earlier ISA/VLB PC Chips board with fake cache...
Part 1:
http://www.youtube.com/watch?v=wNpq0sybH6Q

Part 2:
http://www.youtube.com/watch?v=vUzlkwwH1Aw
 
Here's the squib on the VXPro motherboards that Redhill refers to. Quite frankly, I think either I've been uncannily fortunate, having owned 3 of these (now only 2 after having sold the third) and not having any problems whatsoever (even USB worked fine, which is unusual for a 1996 design). Or it could be that there simply is an over-reaction. The boards are still working fine after 15 years.

It used to be that flash chips were the most commonly counterfeited. You bought either a USB drive or just the chips which were labeled as being 512MB and soon discovered that what you'd received was 64MB chips. Oddly, they'd pass diagnostics, but then develop corruption from overwriting previously-written areas.

I suspect that there are still a number of these in the wild, the owners never having filled memory to the level that would cause corruption.
 
Here's the squib on the VXPro motherboards that Redhill refers to. Quite frankly, I think either I've been uncannily fortunate, having owned 3 of these (now only 2 after having sold the third) and not having any problems whatsoever (even USB worked fine, which is unusual for a 1996 design). Or it could be that there simply is an over-reaction. The boards are still working fine after 15 years.

I think you're right, that it is over-reaction. When they say "Winbench 97 shows at least 10% slower CPUMarks" it sounds like bull to me. In a world where orders of magnitude are the norm, ten percent is just "I'm telling mommy!" talk.
 
It used to be that flash chips were the most commonly counterfeited. You bought either a USB drive or just the chips which were labeled as being 512MB and soon discovered that what you'd received was 64MB chips. Oddly, they'd pass diagnostics, but then develop corruption from overwriting previously-written areas.

A few years ago I bought on eBay from China a 2.5" IDE-to-Compact Flash adapter which came with a no-name "8 GB" Compact Flash card. It came formatted as an 8 GB FAT32 drive. But when trying to install Windows XP on it, after a few minutes the laptop locked up hard, and upon reboot the CF card was unreadable. Then during numerous attempts to re-partition and re-format it, including both on the computer and in a digital camera that takes CF cards, I could never get it to format to more than 32 MB.

So I found out it was a counterfeit -- a fake "8 GB" CF card that was actually only 32 MB, and which would appear as an 8 GB card until you tried to write more than 32 MB of data to it, at which point massive data corruption would ensue.

But even that is better than some fake SD cards which are nothing more than a piece of plastic with some metal contacts on them, and thus are totally non-functional! I once bought a camera which came with one of these fake cards, although instead of trying to pass it off as functional, they put a sticker on it saying, "Additional memory card purchased must same size as this sample." (sic)
 
Back
Top