FTDI driver kills fake FTDI FT232??

[firewalker]

So, the drivers just changes the PID:VID to something else? On Linux, I guess, someone could just end a patch to the kernel tree for the driver adding the new VID:PID.

 drivers/usb/serial/ftdi_sio_ids.h

Alexander.

[krater]

Okay nice. I don't think there are only hobbyists affected of this thing....

So, why should I ever use FTDI chips again when I must control the complete delivery chain for this chips ? That sounds to expensive and risky for me. There are good alternatives...

As if you didn't need before driver update. Or are you OK with using fake parts in your product?

No, but it can happen. It happened many big manufacturers without there liability. Most time the chip-manifacturer then loses a little bit of money because the not selled chips. Now that shall be my risk, but I don't just lose some non existent profit. I lose costs of manufactoring, shipping, time to rebuild, reputation of customers, etc,etc....
So why just eat this toad whitout the need ?

But now you have a great tool how to check to be genuine right away  . However fakers likely will fix them soon.

Hehe, yes now I can But just after the manufacturing.
Yes, thats the joke on this, FTDI loses reputation and the faker will fix their chips...

[rsjsouza]

Does it only kill devices with FTDI's VID? (Can the fakes have the VID reprogrammed?)

If you have your own VID/PID, then you should supply your own device drivers - even if they redistribute the FTDI binaries under the hood (most of them do).

In this case, perhaps you will be shielded if you don't update the binaries in your own device driver support package.

[sacherjj]

Put yourself in FTDI's shoes... It's a bold move, but I understand exactly why they did it.
What makes you think FTDI haven't already been trying to fight the cloners from the supplying end already?
FTDI enjoys a decent premium over the knockoffs and similar products due to brand recognition. Do you think they are going to piss away all that potential money just because some chinese cloners started selling counterfeit product?
Buying fake chips has ALWAYS been a bad deal for everybody involved.

I can understand why FTDI would implement a block on counterfeit chips.  But to silently break them, it is doing them no good.  Inform the user that their device is using an counterfeit chip.  This does nothing to educate people about the problem, it just makes engineers never use an FTDI part ever again.

[Rufus]

Containing a stolen VID the device isn't USB standard compliant so you should not have expectations of it working in the first place. Possessors of stolen property generally have few rights regardless of their knowledge of it being stolen.

FTDI should have 'reclaimed' their VID, presumably not and option, I can't blame them for trashing the PID instead.

Does it only kill devices with FTDI's VID? (Can the fakes have the VID reprogrammed?)

An FTDI driver wouldn't get installed or loaded if the device didn't have a matching VID/PID. I don't know if the VID can be reprogrammed. Removal of a VID owned by FTDI would be more justifiable but what would they replace it with? They couldn't really use someone else's VID or one that had not yet been sold.  0000 seems not to have been allocated but I don't know that 0000 is reserved for uninitialised or unknown like it appears to be for PIDs.

We don't even know how many types of FTDI clone there are. Maybe some are better clones and can't be detected by FTDI drivers.

[ve7xen]

So, the drivers just changes the PID:VID to something else? On Linux, I guess, someone could just end a patch to the kernel tree for the driver adding the new VID:PID.

 drivers/usb/serial/ftdi_sio_ids.h

Alexander.

You don't even need to patch the kernel, you can add the new ID at runtime via sysfs.

[MikeH]

For us, the circuit designers are the ones who decide which chip to use (or replace), and supply chain is a whole different department.  If I had a lot of my products come back with broken FTDI chips, I'd probably conclude that the chips are crap and I'd better use something else. 

[krater]

For us, the circuit designers are the ones who decide which chip to use (or replace), and supply chain is a whole different department.  If I had a lot of my products come back with broken FTDI chips, I'd probably conclude that the chips are crap and I'd better use something else.

Thats how it works in reallity Maybee FTDI doesn't know that...

[Rufus]

For us, the circuit designers are the ones who decide which chip to use (or replace), and supply chain is a whole different department.  If I had a lot of my products come back with broken FTDI chips, I'd probably conclude that the chips are crap and I'd better use something else.

You mean something like real FTDI chips instead of Supereal Microelectronics chips with FTDI written on them?

I had a bunch of boards manufactured in China with fake Fairchild MOSFETS that didn't work - from that I should have concluded that Fairchild makes crap MOSFETs?

[firewalker]

You don't even need to patch the kernel, you can add the new ID at runtime via sysfs.

Yes. Such a patch to the official kernel tree would be nice though.

Alexander.

[mikeselectricstuff]

An FTDI driver wouldn't get installed or loaded if the device didn't have a matching VID/PID.

Not the default one, but one with a suitably tweaked .INF file will.
Anyone who ships a FTDI device other than a simple serial converter with the factory VID&PID should be shot as it can cause major compatibility issues.

[jadew]

This thread is currently going at ~100 views / minute.

I wonder if anyone from FTDI is following it, because it's the story of how people stopped using FTDI chips.

[zapta]

Please note that FTDI`s driver licence agreement will be broken if used with counterfeit devices.

It's your driver that decides to attach itself to the device, not the other way around. The user doesn't 'select' your driver to be used with that competing device.

[Rufus]

This thread is currently going at ~100 views / minute.

I wonder if anyone from FTDI is following it, because it's the story of how people stopped using FTDI chips.

A new user called FTDI chip has already posted in this thread. FTDI don't care if people stop using clones of their chips.

It's your driver that decides to attach itself to the device, not the other way around. The user doesn't 'select' your driver to be used with that competing device.

The device decides what driver it wants to attach to it. The user decided to buy a device which asks for the wrong driver.

An FTDI driver wouldn't get installed or loaded if the device didn't have a matching VID/PID.

Anyone who ships a FTDI device other than a simple serial converter with the factory VID&PID should be shot as it can cause major compatibility issues.

The same as anyone who ships a non-FTDI device with and FTDI VID.

[mrflibble]

Made it to slashdot and hackaday. Well, they're boned. And rightfully so.

Make your drivers simply refuse to work with non-genuine ICs ... fair enough I suppose.

Willfully brick devices from another vendor, just because you are having legal issues with that other vendor .... not very nice, and possibly illegal.

[sacherjj]

There is a very simple way to do this CORRECTLY.

1 - Windows uses device driver based on PID/VID codes.
2 - Driver detects counterfeit device.
3 - Driver informs users that device is counterfeit and does not allow it to function.

Same end result of not working, but the user actually can know why and they complain to the company the made it.  Complains work up the line until we find out where the fake chip came into the mix.  Of all the wrong ways this couldn't be done, I can't think of a better way to get no design engineer to want to risk using your product again.

The product doesn't work.  How do we know if we had an ESD failure or a dishonest Chinese CM?  This is just stupid.

A new user called FTDI chip has already posted in this thread. FTDI don't care if people stop using clones of their chips.

And they are too stupid to understand that people will stop using their chips as well.

[diyaudio]

I bought some RS485 boards from Ebay with an FTDI FT232 chip on them. Appearantly the chips are fake. I used them with Linux and they work fine. After plugging them into a Windows PC with the latest drivers they quit working (even with Linux). I think the FTDI driver somehow kills the fake FTDI chip. Not nice if you got a lot of these boards in the field 

Another one bites the dust.

I encountered the same problem last week, https://www.eevblog.com/forum/microcontrollers/stm32f103c8tc6-usart-help/15/ I am reworking the 3 boards I bought off ebay with the originals.

[langwadt]

For us, the circuit designers are the ones who decide which chip to use (or replace), and supply chain is a whole different department.  If I had a lot of my products come back with broken FTDI chips, I'd probably conclude that the chips are crap and I'd better use something else.

Thats how it works in reallity Maybee FTDI doesn't know that...

yeh, it used to be cp21xx that had a bad reputation because there were so many fakes and they constantly updated the drivers to break them so it seemed the they never worked 

I can also  imagine it could get FTDI in legal trouble, You can't just deliberately break other peoples stuff because you think they violate your IP

[bingo600]

Damm .... FTDI 

I really liked your chips , and considered them the best (most ComPort compatible)
But killing the product , this is unacceptable.

I'll switch to Silabs , on my next designs.

/Bingo

[Rufus]

Make your drivers simply refuse to work with non-genuine ICs ... fair enough I suppose.

Willfully brick devices from another vendor, just because you are having legal issues with that other vendor .... not very nice, and possibly illegal.

They haven't bricked anything. They changed the PID so their older drivers will also simply refuse to work with non-genuine ICs. It would be an issue if there were any other legitimate drivers that worked with these non-genuine chips but there are not and FTDI know it because the non-genuine chips use FTDI's VID and FTDI are the only ones able to create a legitimate driver for something with their VID.

[KPR8]

This is hilarious, so much heat and noise from so many unimportants:

-Try pursuing cloners in China, totally impossible

-If the FTDI drivers displayed a message box 99.9% of users will click cancel and carry on regardless. This way gets their attention

-FTDI don't give a monkey's about all this commotion, this move is aimed at mfrs that ship 10-100K devices. If devices start failing in the field then end users will complain to their supplier not FTDI - it's unlikely the end user will even know or care the USB UART is a fake FTDI. The supplier will notice/care when the RMAs start piling up (and then take action to control their supply chain - FTDI win)

-If you can't trust your mfr then buy the FTDI parts from Mouser/Dodgeykey/RS/Farnell and give them to the Chinese factory FOC. This forces you to control your supply chain, you should be doing that anyway

-Go ahead & use an alternate part, you will not benefit from the 'built-in' FTDI drivers in Windoze and have your end user suffer the inconvenience of downloading/installing a driver - 0/10 for ease of use

-Stop using Tarduino + FTDI USB UARTs, start designing with grown up ARM chips and implement your own USB CDC stacks - problem solved

I say good for FTDI, they invented the USB UART chip business and deserve to benefit financially, instead they are being ripped of by Chinese fakers facilitated by outfits that are too cheap/lazy to control their own supply chains

Flame away

[teraflop]

It would be an issue if there were any other legitimate drivers that worked with these non-genuine chips but there are not and FTDI know it because the non-genuine chips use FTDI's VID and FTDI are the only ones able to create a legitimate driver for something with their VID.

I beg to differ: http://lxr.free-electrons.com/source/drivers/usb/serial/ftdi_sio.c

[jadew]

Make your drivers simply refuse to work with non-genuine ICs ... fair enough I suppose.

Willfully brick devices from another vendor, just because you are having legal issues with that other vendor .... not very nice, and possibly illegal.

They haven't bricked anything. They changed the PID so their older drivers will also simply refuse to work with non-genuine ICs. It would be an issue if there were any other legitimate drivers that worked with these non-genuine chips but there are not and FTDI know it because the non-genuine chips use FTDI's VID and FTDI are the only ones able to create a legitimate driver for something with their VID.

And now the question is if they have any legal rights over the VID, which is just a number allocated by the USB Consortium at $5k a pop. As far as I know, this allocation is not backed by any law.

[mrflibble]

They haven't bricked anything. They changed the PID so their older drivers will also simply refuse to work with non-genuine ICs.

Sure they bricked it. They made changes to a device you paid for & own, without notifying you of these changes and the slight 100% chance of inoperability that might possibly occur. As for bricking semantics, these changes make this device inoperable for the average consumer who has no truck with popping in the debian rescue cd/stick/whatever.

So they don't want me to use their windoze drivers? Fair enough. But after I reboot to linux it'd be nice if that  non-FTDI device still worked.

[linux-works]

So, the drivers just changes the PID:VID to something else? On Linux, I guess, someone could just end a patch to the kernel tree for the driver adding the new VID:PID.

 drivers/usb/serial/ftdi_sio_ids.h

Alexander.

You don't even need to patch the kernel, you can add the new ID at runtime via sysfs.

I tried.  didn't work.  what is the exact command?  I have a 'bad' ftdi board (ruined just today) and so I'm willing to try the hot fix on linux.  the one that was posted before (earlier today) did not work.