NefMoto

hey so I haven't seen this myself but apparently the newest FT232 windows drivers reset the PID of the controller to '0' rendering it basically unusable

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

figured I would post this here, I know a lot of people buy the knock-off ebay cables which are probably affected

Thanks for the heads up. Stickied.

hackaday article: http://hackaday.com/2014/10/22/watch-that-windows-update-ftdi-drivers-are-killing-fake-chips/

apparently you can re-set the PID with their utility on winxp/linux: http://www.ftdichip.com/Support/Utilities.htm

A comparison of the die

http://zeptobars.ru/en/read/FTDI-FT232RL-real-vs-fake-supereal

Anyone found a programatic way of identifying fake FTDI chips? I've emailed FTDI asking how I tell if my chip is real or fake :) I suggest everyone else does the same.

Newer drivers (just before the bricking ones) apparently don't work with the clones... I wish I knew the version numbers.. maybe somebody can do a bit of research.

Clone owners HAD to use old drivers ...

Update from FTDI following coverage on zdnet, reddit, el reg, etc:

http://www.ftdichipblog.com/?p=1053

Theyre pulling the destructive driver and will push something out next week, but this sets a very awkward precedent.

So could this be why some generic KKL cables don't work at all out of the box while seemingly identical other ones work just fine? I'm glad they pulled the bricking driver though, that was not a well-thought out move. They could've easily just given the end user a warning but they decided on the scorched earth approach.

A few weeks old, but still important information:

http://hackaday.com/2014/10/22/watch-that-windows-update-ftdi-drivers-are-killing-fake-chips/ (http://hackaday.com/2014/10/22/watch-that-windows-update-ftdi-drivers-are-killing-fake-chips/)

http://m.slashdot.org/story/208837 (http://m.slashdot.org/story/208837)

So it may just be coincidence, but my laptop just did a windows update, rebooted and now the FTDI cable doesnt work. Anyone else having issues?

EDIT: So I decided to do some debugging...

The bad FTDI drivers set fake(?) chips to a VID:0403, PID:0000, when the chip should report 4003:6001.

I configured windows to never download drivers from windows update, then I set the PID back to 6001, plugged it in and let windows detect it with the bad (2.12.00 drivers), then hit 'uninstall driver, and delete driver', which causes the driver to be deleted from the 'Driver Store' - the internal database windows uses to locate drivers.

Unsurprisingly, the bad drivers had set my PID to 0x0000 again, so I reset that, plugged it back in and windows loaded up the drivers from January 2014, iirc the 2.10.00 set. Everything was looking peachy, so I fired up NEF and could talk to a spare me7.5 ecu. So I unplugged it again, plugged it back in, and surprisingly it wouldnt work again - because my PID had been set to 0000 again.

So I tested this a few times, set the PID to 6001, plugged it into windows, unplugged it and then checked the pid, everytime the windows laptop would reset the PID to 0000. So I went through and uninstalled all the FTDI drivers from the windows store, even with no drivers installed at all (not even the 2006 ones that ship with windows), when the port would appear as a 'USB Serial port, manufacturer: unknown, device type: other devices' and not work, it would reset the PID when I plug the cable in - so what the heck have FTDI left on my system?

I did a search on system32 and system32/drivers and it turned out quite a lot, so I first tried deleting all the files installed by the 2.12 driver pack - but that didnt work, so I used "pnputil.exe -f -d oem<xy>.inf" to delete the inf for my FTDI drivers. Along the way, I told windows to delete the old ftdi drivers too, given the wierd behaviour I've been seeing.

At this point, I can plug and unplug my FTDI lead without it zeroing the PID. I reinstalled the 2.10 drivers and I can still plug and replug it. Read a eeprom using argdubs tool, and flash file with nef and it all seems to be happy. Of course, it might break tomorrow!

My cable died after Windows update  ;) All blue cables use fake FTDI chip.

Set the PID back to 6001 using linux, disable windows update for drivers, delete the bad ftdi drivers using pnputil, install the good ftdi drivers, your lead will work again :)

really, do ALL blue leads have fake chips? Shame, I like my blue lead!

Is there any way to identify a fake chip without plugging it into a computer with "destructive" drivers?
I normally use an XP-machine so I'm safe, but it would be good to know if the adapters I have has real or fake FTDI chips so I know if I can use them in other computers or not.

I haven't yet found one, I emailed FTDI about this months ago, but never got a response. Maybe it's time to reach out again...

Price for blue cable and original FTDI is the same  ;)

This method doesnt work in 100% cases, but you can compare "FTDI" label on the chip. Original printing is deeper and made by laser.

The concern is that I dont know which leads have genuine chips in and which ones dont. Obviously I'd rather buy genuine given the choice, but how do you tell from an ebay ad?

I know that you can tell the difference by looking at the chips, but the one in my lead is so small you can barely see the writing - do you use a magnifying glass or something?

No, its easy to see difference without any glasses etc...
I dont know how to explain... ori label is engraved. Fake label is printed. You can see it on the photo.

By the way I'm using fake FTDI for some months and everything is fine) I think even some fake chips work with new driver.

I wonder if there is a list of common tuning devices that will be killed by these drivers? I don't own a lot of clone stuff but it would be worthwhile to know if your galetto or mpps cable will be killed by this update.

It's fairly trial and error, but the problem seems to have gone away with later drivers. I would assume that most fake ('clone') cables and most blue kkl leads use fake chips

again...


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

There's a simpler way. Download 2.10 the old unsupported version worked for me.

Doesnt work if your cable has been bricked, because the old unsupported drivers wont recognise it as a FTDI cable.

Ah I see, luckily mine didn't brick. It just froze up and I repeatedly tried installing old drivers and it randomly worked. Ill consider my self lucky this time.

here , possible solution... are you guys usb or serial?

Download 2.10.0.0 drivers here: http://www.ftdichip.com/Drivers/CDM/CDM v2.10.00 WHQL Certified.zip