Counterfeit parts..

[peter-h]

I don't think FTDI were ever a player in the Far East.

I've used the FT232 for 15 years in OEM volumes and it was always at least 3x the price of Prolific's PL2303.

What most likely happened is that people cloned the FT232 solely to take advantage of FTDI's (obviously, generously provided for free) drivers which were of very high quality and much better than the buggy junk which Prolific produced (which also had weird limitations e.g. no drivers after win7 for the original PL2303).

FTDI were completely entitled to prevent the use of their drivers with these fake chips.

[SiliconWizard]

What most likely happened is that people cloned the FT232 solely to take advantage of FTDI's (obviously, generously provided for free) drivers which were of very high quality and much better than the buggy junk which Prolific produced (which also had weird limitations e.g. no drivers after win7 for the original PL2303).

FTDI were completely entitled to prevent the use of their drivers with these fake chips.

I've expressed the same before and I fully agree.

One of the major benefits of using FTDI products was, and still is, their drivers (especially on Windows, drivers for the rest of supported OSs being mainly based on libusb and not particularly interesting.)
But writing good and robust drivers on Windows was really hard (and still kind of is, even though this has slightly improved), and the WHQL certification is painful and expensive, you get all this for free with FTDI.

Their major mistake was probably to have, at some point, made their drivers brick fake devices. Customers were furious because that basically "killed" their products. FTDI could have just decided to detect and stop supporting fake devices with their drivers - that would have been enough to seriously harm the "cloners" (since, as you said, they are mainly interested in leveraging FTDI's drivers), and customers could not have said that FTDI killed anything. It would also have made it much clearer who was at fault... FTDI were entitled to do something about fake chips, but not this. This decision was probably one of the worst they've ever taken and it has harmed them tremendously. This story still bites them, and has likely harmed them much more than it has harmed counterfeiters, which was certainly not the point.

As to counterfeiting, in itself it's an offense in most countries that I know of, so there shouldn't be much point in defending it in any form.
China's tolerance towards counterfeiting should, IMO, stop. The chinese government should definitely do something about it if they want to change the image China still bears.

For those trying to say that FTDI products are not even good, why even counterfeit/clone them then? Come on. This is just completely dishonest.

[Gyro]

Yes of course the USSR could clone an 8086 and 8087. They cloned the VAX11/780 too.

But that is the technical might of a whole country.

Some chinese fake chip operation is not going to be cloning an 8087.

Irrc, it was either the MicroVAX 2 chip or the Alpha (Alpha I think) that had a little joker playing card on the edge of the die together with the words "Only copy the best" written in Cyrillic.

I remember all the extra security coming in, lectures about end-user certification,tight full height turnstiles etc. in the UK development office.

[peter-h]

As I wrote before, it is not hard for a "nation state" to clone a CPU, from that era.

Anybody with a "real" university degree in what in the old days (1970s) was called Computer Science will know about microcode and how a CPU is built.

I've not been in "serious" logic design for many years but some 25 years ago was doing ASIC designs where we used Xilinx X3000 and X4000 series FPGAs for the prototyping, and synthesising a CPU using some language like VHDL was really quick. To make it timing-compatible with the original you would need to be more clever but there are many clever people around, especially in the USSR which like the rest of the communist bloc always had rigorous science/engineering education, and since they had few resources they copied and reverse engineered all they could. It was a national game.

Re FTDI, yes it was excessively aggressive of them to clobber the fake chip's EEPROM contents, but since there were no non-FTDI drivers for the FTDI chips, what difference did this make? It is not as if there were other drivers which could have been used if the FTDI ones stopped working.

[Karel]

One of the major benefits of using FTDI products was, and still is, their drivers (especially on Windows, drivers for the rest of supported OSs being mainly based on libusb and not particularly interesting.)
But writing good and robust drivers on Windows was really hard (and still kind of is, even though this has slightly improved), and the WHQL certification is painful and expensive, you get all this for free with FTDI.

Not free. You pay for these high quality drivers when you buy the chips. That's why it's so unfair to use fake chips,
you use these high quality drivers without paying for it (by not buying the genuine chips).

FTDI could have just decided to detect and stop supporting fake devices with their drivers

And what difference does that make? The device still doesn't work. It will be "bricked" as well.
Windows will still think it's an FTDI device and will load a driver that will refuse to work.
How is that a solution compared to setting the PID and VID of the fake chip to zero?
In order to let the fake chip work with the generic virtual comport drivers of windows,
you have to change the VID and PID of the fake chip anyway.
So, from a customer point of view, the FTDI driver refusing to work or setting the VID & PID to zero causes the same trouble.

[OwO]

The Linux driver always supported the chips, clones or not. By overwriting registers you disable using the device on Linux and other operating systems. Also the driver in the Linux kernel is GPL and FTDI has no basis for disallowing using clone chips with it. They even attempted to upstream a patch with clone detection which was promptly rejected.

[iMo]

Disabling the chip that way is illegal, of course. An average lawyer would win the case against FTDI.

[all_repair]

Silliness does not seem to stop even people are not buying FDTI marked cable anymore.  Regardless who did the cable, buying them means they could be bricked anytime.  Shallow views still persist even though they are screwing the end-customers and system integrators that bought these Cables. Half through the projects and even half through the operation, the FDTI marked cables can suddenly don't work.  ȘI and customers safest way to stay out of the cross-fire and not got killed by these bullets are to stay away from not only FDTI cables, but anything FDTI related devices and chips.
FDTI chips and drivers eventually are part of a bigger system, and people paying for these child play are not the cloner.  FDTI screwed up the end customer and their operations, and screwed themselves up at the end because end customer will not want to touch FDTI related devices with a 3 foot pole.

[peter-h]

Unless of course you are actually making the stuff and buying the real FTDI chips, like I do in my business

Then you know they won't get bricked.

[EEEnthusiast]

There are very few alternatives to FTDI for USB to FIFO ICS. For a UART to USB bridge, there are many solutions which work well as FTDI. So FTDI devices will stay here for long.
Provided you are lucky to get the real FTDI ICs, they run very well on Windows with very less compatibility issues. Programming them is easy and they are nice to work with.

[peter-h]

I would probably not purchase FTDI parts in China

Also I would not use a chinese assembly outfit to locally source FTDI chips. Or any other electronic component for that matter unless I did what all the smartphone etc makers do: have my own people out there watching everything.

[rsjsouza]

I have been reading an interesting PDF on counterfeit parts.

This is very interesting; thank you for sharing! I haven't heard of several techniques as well.

Far more likely these fake 8087s are real ones, recovered from scrap boards, or subspec ones marked with a higher spec number, or some old stock somebody found and maybe "repainted".

I think so too. A few months ago I was looking for its newer brother 80387SX and I could find only the 25MHz versions in buckets, but the 16MHz version I wanted (the main CPU was a 80386SX-16) was quite rare. Either Intel mostly sold the faster version or they were repainted.

Regarding the FTDI, that horse was already beaten to death many times over, with nobody able to convice the other side. Just search around and I am pretty sure nobody here will bring anything new.

https://www.eevblog.com/forum/reviews/ftdi-driver-kills-fake-ftdi-ft232/

https://www.eevblog.com/forum/microcontrollers/ftdi-gate-2-0/

[SeanB]

387sx in 16MHz a 25Mhz part would work fine, though I suspect that if you actually bought a couple dozen of the /25 parts you would find they are almost all likely fake or repainted sevices.

I used a freeware 387 emulator that, while a lot slower than the actual 387 chip itself, it would at least allow you to use a system without a 387 to run stuff that absolutely wanted a hardware chip. Simply by running an emulation that captured the interrupts that the 387 used, and then run them on the 386 processor instead at a slower rate. But would run, and only timing would give it away.

Still have a PC with a 386processor on it, 386SX 33 IIRC, i used it as a controller for years, must look and see if i can connect it to a slightly more modern display, as I no longer have any CGA/EGA/HGC monitors around.

[rsjsouza]

387sx in 16MHz a 25Mhz part would work fine.

Thanks for that. I could never find clear cut information about that compatibility and it wasn't pressing enough for me to dive into  the timing diagrams of both speed classes to evaluate compatibility.

I used a freeware 387 emulator that, while a lot slower than the actual 387 chip itself, it would at least allow you to use a system without a 387 to run stuff that absolutely wanted a hardware chip. Simply by running an emulation that captured the interrupts that the 387 used, and then run them on the 386 processor instead at a slower rate. But would run, and only timing would give it away.

I used an emulator as well, but it was an 8087 on our then "brand new" 286. It made possible to run Autocad 9 in DOS, but too slow to be practical.

Still have a PC with a 386processor on it, 386SX 33 IIRC, i used it as a controller for years, must look and see if i can connect it to a slightly more modern display, as I no longer have any CGA/EGA/HGC monitors around.

Perhaps some day you could get a used ATI AllinWonder VGA card to fit on your PC. These were the epitome of consumer 2D perdormance on the ISA bus.

I got a working T2000LE Toshiba laptop that fortunately has a VGA external connector. The built-in video was very flaky and full of leaky Al capacitors all around. After a very large number of hours unfortunately my replacements couldn't yet bring it back to life. It is in the "postponed projects" section of my bin.