FTDI driver kills fake FTDI FT232??

[dannyf]

You may have meant 'salvaged', but savaged sounds quite good too

I actually think "savaged" here is the right word,

[edavid]

You generally get what you pay for and buying from some countries directly increases the risk, my supplier won't supply me atmega chips simply because they are not made in china and I can get them cheaper than he can, so yes assume that most arduino chips that are cheap and from china are fake.

Has anyone ever seen a fake AVR?  (Not a "slug", a functional part.)

[Rick Law]

The trouble is how is the end user going to know.  You can assume that a $100 new Rolex is probably going to be a fake, but you cannot assume that a $10,000 Rolex is going to be real unless you get it appraised.

Whether it was sourced from eBay is hardly an indicator.  Even if 100% of eBay is fake, avoiding eBay still would not assure you got something real.  Paying a high cost merely assures it is expensive, but it does not assures it is 100% genuine.

This is timely (from yesterday, Nov 17): Consumers Report, a magazine trusted by millions for objective reviews of products.  They reviewed a Chinese brand tire(tyre) and the results were poor.  The USA rep of the manufacturer (API) called about the surprising result that doesn't even come close to their internal tests.   Upon investigation - it is a counterfeit.

Consumers Report, being the kind of magazine they are, dug deep to investigate as well.  In their article following the supply path of the product,  readers can discern a hole which may suggest how it got into the supply chain.  But it is not a proof nor did Consumer Report suggested that was how the fake got in to the supply chain.  Granted, this is a low cost brand - but the same way this got into the supply chain is no different than had it been a high-cost brand.

http://www.consumerreports.org/cro/news/2014/11/counterfeit-car-tires-pose-consumer-risk/index.htm

Whenever there is a profit, there will always be someone willing to do bad things to gain from it.  Punishing the consumer is not the way to stop it.

Rick

[Simon]

You generally get what you pay for and buying from some countries directly increases the risk, my supplier won't supply me atmega chips simply because they are not made in china and I can get them cheaper than he can, so yes assume that most arduino chips that are cheap and from china are fake.

Has anyone ever seen a fake AVR?  (Not a "slug", a functional part.)

I've never been able to source parts for less than I can get from Farnell in the same volume, when I have been offered parts that seem quite cheap and ask if they are genuine i get no reply, so I don't buy. When i asked my chinese supplier for AVRs he refused saying that he could not compete so no point in him trying to supply.

[sciencelookers]

I'm confused again.  On Friday, two Arduino Nanos I was working with were bricked.  Near as I can tell, the FTDI drivers don't talk to them AND their stop bits got set to zero.  Is there a new update thats doing this again?  I got update 2.11 and it bricked the Arduinos.  Going back to 2.08 lets me program Arduinos without a problem, but the Arduinos that got bricked on Friday are still bricks.  How did my PID get set to zero when supposedly that stopped a month ago?  If anyone has any new information on the situation, I would sure appreciate hearing whats up now.

[marshallh]

I got the update from Windows Update a week ago... I don't think it was ever pulled?

[TheEnd]

Driver packages 2.11 and 2.12 (mentioned, but not seen by me) are the "spiked" drivers.
I guess everyone will have to wait and see what happens with 2.13, or maybe a 2.11.X release.

There are workarounds to get them back up and running again, I got a USB lead a week or so ago which got hit, but I decided to send it back instead of the workaround as who knows what the next driver could do.

I'm fairly sure FTDI won't just walk away from this, there will probably be something up their sleeve, 5 or 10 minute resets or something similarly annoying in a sort of "demo" mode.

[nitro2k01]

I'm fairly sure FTDI won't just walk away from this, there will probably be something up their sleeve, 5 or 10 minute resets or something similarly annoying in a sort of "demo" mode.

Or, it may simply refuse to work with clones, without bricking them.

[dannyf]

I don't think it was ever pulled?

Based on my experience, the driver was definitely pulled?

[nitro2k01]

Based on my experience, the driver was definitely pulled?

Are you sure you ever got the driver in the first place? If you didn't install the driver through the online service, or if you disabled non-essential updates, I don't think the driver would have been pushed to you in the first place. Both of those are true for me, and I still have the driver version I originally installed.

[f4eru]

Well I cant think of anybody who would buy their coponents on ebay if it is for production and sales. Cause then he or she could face some serious legal problems or economical issues when the circuits fails

I don't think people in a procurement department will buy from Ebay but they would buy from any parts broker if there are no other suppliers able to deliver a part.

So what you are saying is...production comes before security ? Ok this sounds like gambling

Yep. Production comes before security, because production means profit, and security does not mean direct profit. Most of the actual world is centered on profit.

Heck, even most Nuclear reactors are running with critical security parts coming from ebay. why ? because the spare parts are not available otherwise. Between closing the nuke plant and purchasing the spare part on ebay, they choose the most profitable option over security :
http://enenews.com/report-every-nuclear-plant-staff-workers-looking-ebay-buy-buy-parts-audio

[Rigby]

Heck, even most Nuclear reactors are running with critical security parts coming from ebay.

Bullshit detector just went off.  I'm gonna need to see a legitimate source for that information.

[f4eru]

I don't think it's bullshit.

When you have a safety certified system, any change, as small as it is is a big risk, and therefore is associated with a lot of reviews, paperwork, partial or complete recertification.
It is much more cost effective to get old hardware somewhere and just check it's integrity and autenticity. If properly checked, it is probably even less risk than to redesign the system with modern but unproven hardware.

Mr Gundersen is an industry insider, yes he is oriented anti-nuke, but he knows the practices of the industry.

[timb]

The trouble is how is the end user going to know.  You can assume that a $100 new Rolex is probably going to be a fake, but you cannot assume that a $10,000 Rolex is going to be real unless you get it appraised.

Whether it was sourced from eBay is hardly an indicator.  Even if 100% of eBay is fake, avoiding eBay still would not assure you got something real.  Paying a high cost merely assures it is expensive, but it does not assures it is 100% genuine.

This is timely (from yesterday, Nov 17): Consumers Report, a magazine trusted by millions for objective reviews of products.  They reviewed a Chinese brand tire(tyre) and the results were poor.  The USA rep of the manufacturer (API) called about the surprising result that doesn't even come close to their internal tests.   Upon investigation - it is a counterfeit.

Consumers Report, being the kind of magazine they are, dug deep to investigate as well.  In their article following the supply path of the product,  readers can discern a hole which may suggest how it got into the supply chain.  But it is not a proof nor did Consumer Report suggested that was how the fake got in to the supply chain.  Granted, this is a low cost brand - but the same way this got into the supply chain is no different than had it been a high-cost brand.

http://www.consumerreports.org/cro/news/2014/11/counterfeit-car-tires-pose-consumer-risk/index.htm

Whenever there is a profit, there will always be someone willing to do bad things to gain from it.  Punishing the consumer is not the way to stop it.

Rick

Everyone complaining that designers should "just buy authentic parts" needs to read that CR article, as it directly reflects what's going on here.

You have the manufacturer (FTDI), importers, resellers, buyers/designers and end user of the product the part is designed into.

How can I verify the chip is legitimate as an end user or, hell, as the designer even?

I can't. So does FTDI have the right to brick my device? No. This is common sense stuff guys.


Sent from my Smartphone

[kourosh]

hi,
there is simple solution, as FTDI let you change PID , they also provide a program which is able to config their driver to work custom PIDs,
thus you can go http://www.ftdichip.com/Support/Utilities/FT_INF_v2.0.61.242_Installer.exe ,  and download that, then you need an old ftdi driver
like this:
32bit:  http://www.ftdichip.com/Drivers/CDM/CDM%202.08.30%20WHQL%20Certified.zip
64bit: http://www.ftdichip.com/Drivers/CDM/CDM%202.08.30%20WHQL%20Certified.zip

then run the FT_INF,set chip type to FT232/245R and PID to 0000, then extract above zip,
in Time Stamp/Filename, select ftdibus and ftdiport, then using this >> to add your prefer config to right menu.
now click Generate files.
2 new inf will be generated. copy them to the folder where you extracted zip files, and overwrite the.
then manually point windows to your new driver.
 I did this on win7 64bit, win7 two time asked for driver, and then win recognized my device.

you can download modified driver from here, win7 64bit: http://www.4shared.com/rar/GK7OOeS7ce/FTDI_unlocked.html

[alm2k]

After updating the drivers through Windows Update I now have version 2.12.0.0 (dated 26-8-2014) on the "USB Serial Port" on my Windows 7 64bit box. However the "USB Serial Converter" below "Universal Serial Bus controllers" still shows version 2.10.0.0 (dated 27-1-2014). I used to have version 2.10.0.0 since earlier this year early 2014.

I'm wondering if these drivers are indeed the same drivers as before, as they were supposed to have been pulled or has something changed?

Can people who had their devices bricked tell what versions of the drivers they had? Did it show 2.12.0.0 in both places in Device Manager or only at the "USB Serial Port" device as with me?

My devices still work with this version (at least they are recognized and unplugging and plugging them back in doesn't change anything), can I be sure now that they are indeed genuine? One of these was bought from Mini-in-the-box (China), but it was about 7 or 8 bucks, so not a 2$ piece.

[alm2k]

Nobody who knows the answer to my question above?

[0xdeadbeef]

There was a perl script posted in October and I used it to check my devices under Win7 back then:
https://www.eevblog.com/forum/reviews/ftdi-driver-kills-fake-ftdi-ft232/msg539237/#msg539237

[Rubi]

I run in this trap recently while building a Elektor LCR meter.

I reflowed the chip and it worked.
The I reflowed the micro controller and the ft232 did not enumerate correctly anymore.
It cost me hours to find out what was wrong, this is so emberassing.
Howe could thery do that without warning ?
On the other hand I learned a lot about how to figure out usb troubles and how to use an usb event viewer.
In my onw designs I will only use the cp210x. I liked to use them since they have an integrated 3V3 regulator and are easier to reflow since they have no pins.

Cheers
Michael

[nitro2k01]

Rubi: Before we suspect that the chip is fake, are you sure that the test pin is grounded on the board? A trap for young players, as a certain crazy Aussie would put it.

[Rubi]

Hi

Yes the test pin is grounded.
I checked the enumeration process and it is perfectly ok, except the pid is 0.
I then found a workaround for the problem and could assign the driver directly accepting a pid of 0.
Since then the chip works great, programmed the uc and works with the assigned programm.

So it was definitley a fake chip and thanks to the internet I could find a description how to get around it.

Cheers
Michael

[toysareforboys]

Super freaky brick problem. This is NOT a fake FTDI chip (well, at least I assume it's not). The manufacturer of the equipment remotely bricks the chip any time you open their software while connected to the internet (just to be bastards and make more money).

This is what it looked like before the brick:



When the software bricks it, it resets BOTH the PID and VID to 0000



This renders the device completely unusable as you would imagine. I tried using the FTDI Inf generator program to see if I could set the inf's up to use VID 0000 and PID 0000 but no luck getting them working. Using the custom INF files it recognized it as USB Serial Converter instead of "Unknown Device" but it still says "The device cannot start. (Code 10)"



In reading online it appears that windows will refuse to load a driver if the PID and VID are set to the same value, so not sure we can hack the FTDI driver to override that.

I read a few posts about using Linux to reset the PID (but unfortunately not the VID) but when I boot up Ubuntu and do "lsusb" the FTDI converter isn't even listed at all! Does anyone know of a way to get Linux to recognize the VID of 0000 so I can overwrite it?

Any help you can provide would be greatly appreciated. There are a LOT of these devices out there that have this problem.

Thanks!

-Jamie M.

[nitro2k01]

Jamie, what does dmesg output after you plug in the device in Linux?

[toysareforboys]

Jamie, what does dmesg output after you plug in the device in Linux?

[   69.144050] usb 8-1: new full-speed USB device number 2 using uhci_hcd
[   69.268064] usb 8-1: device descriptor read/64, error -71
[   69.492047] usb 8-1: device descriptor read/64, error -71
[   69.708069] usb 8-1: new full-speed USB device number 3 using uhci_hcd
[   70.036064] usb 8-1: device descriptor read/64, error -71
[   70.260069] usb 8-1: device descriptor read/64, error -71
[   70.476060] usb 8-1: new full-speed USB device number 4 using uhci_hcd
[   70.888039] usb 8-1: device not accepting address 4, error -71
[   71.204082] usb 8-1: new full-speed USB device number 5 using uhci_hcd
[   71.616047] usb 8-1: device not accepting address 5, error -71
[   71.616096] usb usb8-port1: unable to enumerate USB device

-Jamie M.

[linux-works]

you should be able to use the ftdi fixer (not by ftdi, but by some linux guys) to set the pid and vid.

the key is to have only one dongle plugged in and refer to it by /dev/ttyUSB0.  it will then set the vid and pid for you if you give it both arguments to the CLI.