FTDI driver kills fake FTDI FT232??

[Richard Crowley]

You can connect it to a different or the same computer with an legacy driver with PID 0000 instead of PID xxxx this I believe is proven to work. You can either edit the .inf file, or force the driver to install in device manager, or use a GUI tool to do it, or use regedit, or use CLI tools, or use a linux tool, or... (So many things still work notice that the list is a massive list of options you have all of which are in my opinion a joke, my printer doesn't automatically install drivers and I have to use the force driver method in normal install anyways)

You can roll back the driver, you can force it to work with it even if the IDs don't match. You can as shown before connected to a different os and use completely FOSS tools to not only correct but even prevent it from being modified again afterwards.

FTDI did a bad/mean/evil DRM ala HDCP style detection and modification but the workaround/bypass is a joke.

If you need help fixing this particular issue I can help for free via remote desktop. (After I finish some work stuff of course)

In all the above cases, what  a210210200 means by "you" is the experimenter, hacker, hobbyist, programmer, developer, engineer, etc.

Alas, virtually no END USER of a commercial product who has the misfortune of using an appliance assembled with counterfeit chip will be in a position to mitigate the damage caused by FTDI.

It appears that FTDI has rather "queered the pitch" for USB to serial in general and "fouled their own nest" in particular.

[Kjelt]

If its real eeprom it can be written a few million times.
If it is flash it can be written 100000 times.
If it is something else, who knows

[f4eru]

Hence two writes, and the EEPROM retains its original values.  Objections?

Yes,objections : interrupting the sequence leaves the device with an incorrect checksum (probably bricked)

[markb82]

Hence two writes, and the EEPROM retains its original values.  Objections?

Yes,objections : interrupting the sequence leaves the device with an incorrect checksum (probably bricked)

Ditto for FTDI's solution, not any worse.  Is it?

Edit:

The best would be to find a test for a clone which doesn't require an EEPROM write.  I don't have a clone, and don't really want to dedicate any time to this as I don't see any benefit in it for me personally.

[a210210200]

You can connect it to a different or the same computer with an legacy driver with PID 0000 instead of PID xxxx this I believe is proven to work. You can either edit the .inf file, or force the driver to install in device manager, or use a GUI tool to do it, or use regedit, or use CLI tools, or use a linux tool, or... (So many things still work notice that the list is a massive list of options you have all of which are in my opinion a joke, my printer doesn't automatically install drivers and I have to use the force driver method in normal install anyways)

You can roll back the driver, you can force it to work with it even if the IDs don't match. You can as shown before connected to a different os and use completely FOSS tools to not only correct but even prevent it from being modified again afterwards.

FTDI did a bad/mean/evil DRM ala HDCP style detection and modification but the workaround/bypass is a joke.

If you need help fixing this particular issue I can help for free via remote desktop. (After I finish some work stuff of course)

In all the above cases, what  a210210200 means by "you" is the experimenter, hacker, hobbyist, programmer, developer, engineer, etc.

Alas, virtually no END USER of a commercial product who has the misfortune of using an appliance assembled with counterfeit chip will be in a position to mitigate the damage caused by FTDI.

It appears that FTDI has rather "queered the pitch" for USB to serial in general and "fouled their own nest" in particular.

1) GUI tools are easy to use and are meant for non-expert users
2) Device manager and its prompts are something a user can see with many devices that they have to install.
3) Command line tools that work like this (Your device has been PID 0000'd by FTDI do you want to correct? [yes], press enter, done) easy. Some command line stuff works so easily you literally just press enter constantly till it says done. (Some level of reading comprehension is required, high school English should be sufficient)

Yes FTDI is mean/bad/evil for breaking plug-in play compatibility but it literally is a joke to fix (You know before 2000 automatic plug and play detection was crap and regular computer users had trouble with cards, drivers, devices not working out of the box or even conflicting or getting the wrong settings coded). Compared to HDCP which does require some level expertise to "fix" it is a trivial problem that even 1st year students have figured out without reading my guide which is only given to the other project based course.

I don't understand if you use windows or linux you have to have at least some level of skill in you know pressing the buttons to do things and using google when your at a loss. It is like I'm walking into APSC160 all over again and people are like "how do I open Excel..." when they should be configuring a USB DAQ to dump data into an excel file... (Then there are students who go nothing shows up and when you look at their station the setup isn't turned on and isn't even connected to the computer... when the guide has picture by picture steps telling them to do just that)

I had probably 1000's of students all of them are certainly end users of an FTDI products (a box of USB>RS-232 adapter cables counts as a commercial product no?) know how to do something as simple as that, students that get development boards know even more about them and I don't see how this is so hard. (The com port numbers get screwy sometimes or some student made software had a dumb limitation of only allowing users to select 10 com port numbers when they could have auto-detected. but anyways people figured out how to fix that easily its just point click interface)

No user who doesn't know how to use google would be unaware of how to fix it. Also need help ask someone it works a lot of the time and for this problem most people should know how to figure it out.

Again if your having problem's with an FTDI device being bricked I can help you via remote desktop. (For free, work permitting)

Edit: For #2 here is a pictorial guide easy stuff if only it was that easy to "fix" HDCP problems, https://www.eevblog.com/forum/reviews/what-should-ftdi%27s-next-driver-update-look-like/msg541025/#msg541025

[MicroBoy]

If the detection is to write the PID, then check it is 0000 and then rewrite it to what it was, is that not also unauthorized modification of a users product without permission, this time twice, or many more times each time it is plugged in?

Don't forget that EEPROM wears out...

You will exceed the insertion spec of your USB connector before you wear out the EEPROM (provided the clone uses a real eeprom, some have suggested it could be some sort of OTP ROM by looking at the die shots? I'm no expert)

Not necessarily. Rebooting your PC with the device always attached will make the driver write the EEPROM every time.

If its real eeprom it can be written a few million times.
If it is flash it can be written 100000 times.
If it is something else, who knows

Normal EEPROM would be around one to  ten millions.
If it's Flash, i know some uC that only guaranteed one hundred times (a couple of PIC18FXJXX? can't remember the exact model). Others go up to ten to hundred thousands.

[a210210200]

If the detection is to write the PID, then check it is 0000 and then rewrite it to what it was, is that not also unauthorized modification of a users product without permission, this time twice, or many more times each time it is plugged in?

Don't forget that EEPROM wears out...

You will exceed the insertion spec of your USB connector before you wear out the EEPROM (provided the clone uses a real eeprom, some have suggested it could be some sort of OTP ROM by looking at the die shots? I'm no expert)

Not necessarily. Rebooting your PC with the device always attached will make the driver write the EEPROM every time.

If its real eeprom it can be written a few million times.
If it is flash it can be written 100000 times.
If it is something else, who knows

Normal EEPROM would be around one to  ten millions.
If it's Flash, i know some uC that only guaranteed one hundred times (a couple of PIC18FXJXX? can't remember the exact model). Others go up to ten to hundred thousands.

Which is why FTDI should just make the driver report itself as failed/error to windows so it doesn't attempt to do anything for that device. Only if the user forces it to re-install or tries again would it happen again (plugging it into another USB port would also cause it as well due to the way windows keeps track of things).

[(*steve*)]

Hence two writes, and the EEPROM retains its original values.  Objections?

Yes,objections : interrupting the sequence leaves the device with an incorrect checksum (probably bricked)

I wonder if the chips respond differently to read requests on odd and even addresses.  I know that a similar means of detecting a clone is used in another product.  In this case you don't need to do writes at all, resulting in identification without bricking.

[markb82]

I wonder if the chips respond differently to read requests on odd and even addresses.  I know that a similar means of detecting a clone is used in another product.  In this case you don't need to do writes at all, resulting in identification without bricking.

I think you're thinking something like an EEPROM write status register, which I don't believe the driver API would have access to unless there is a specific command for that?

[markb82]

If the detection is to write the PID, then check it is 0000 and then rewrite it to what it was, is that not also unauthorized modification of a users product without permission, this time twice, or many more times each time it is plugged in?

Don't forget that EEPROM wears out...

You will exceed the insertion spec of your USB connector before you wear out the EEPROM (provided the clone uses a real eeprom, some have suggested it could be some sort of OTP ROM by looking at the die shots? I'm no expert)

Not necessarily. Rebooting your PC with the device always attached will make the driver write the EEPROM every time.

Not if you keep a list of serial numbers you've determined to be clones inside the driver (windows registry?), then you need to run the check once and store the result.  (as on page 88 of this thread)

[Richard Crowley]

I had probably 1000's of students all of them are certainly end users of an FTDI products (a box of USB>RS-232 adapter cables counts as a commercial product no?) know how to do something as simple as that, students that get development boards know even more about them and I don't see how this is so hard. (The com port numbers get screwy sometimes or some student made software had a dumb limitation of only allowing users to select 10 com port numbers when they could have auto-detected. but anyways people figured out how to fix that easily its just point click interface)

No user who doesn't know how to use google would be unaware of how to fix it. Also need help ask someone it works a lot of the time and for this problem most people should know how to figure it out.

Wow, you don't have much (any?) experience with Dumb End Users of appliances.  They make your "1000's of students" look like geniuses. It might take you 30 minutes just to get the USB cable plugged back into the right place after the janitor accidentally pulled it out last night.  (And that is after 10 minutes to identify which is the USB cable.)

[ozwolf]

No user who doesn't know how to use google would be unaware of how to fix it. Also need help ask someone it works a lot of the time and for this problem most people should know how to figure it out.

Really? Are you serious?  Look around you and open your eyes.  There are lots of everyday users who are very competent using their devices, but don't need to know (or wish to know) the insides and how it operates. 

Let's not forget our customers/users/family members who didn't ask for their devices to be bricked.  How much time has been wasted by the end user because the device doesn't work today, but used to work a few weeks ago.

How many devices have ended up on the trash heap because they just stopped working? 

Just because members of this forum have an interest (and ability) in electronics/computing, we should never forget our customers/users/family members may not have that same ability/talent. 

FTDI has the right to protect their IP, no argument there.  However, a good corporate citizen that respects the final customer (the end user), would have merely notified of the suspect device - "Not an FTDI device - please contact your supplier/manufacturer for the correct driver."

FTDI does NOT have the right to vandalise my device.

Ozwolf

[a210210200]

I had probably 1000's of students all of them are certainly end users of an FTDI products (a box of USB>RS-232 adapter cables counts as a commercial product no?) know how to do something as simple as that, students that get development boards know even more about them and I don't see how this is so hard. (The com port numbers get screwy sometimes or some student made software had a dumb limitation of only allowing users to select 10 com port numbers when they could have auto-detected. but anyways people figured out how to fix that easily its just point click interface)

No user who doesn't know how to use google would be unaware of how to fix it. Also need help ask someone it works a lot of the time and for this problem most people should know how to figure it out.

Wow, you don't have much (any?) experience with Dumb End Users of appliances.  They make your "1000's of students" look like geniuses. It might take you 30 minutes just to get the USB cable plugged back into the right place after the janitor accidentally pulled it out last night.  (And that is after 10 minutes to identify which is the USB cable.)

Students labs are designed to be as dumb end user ready as physically possible. Everything that can go wrong will and then more so since some really know what they are doing and do bad things that they knew would cause the worst possible result. (Students unlike dumb users can be both dumb and hostile actors) If you don't think universities are a difficult operating environment consider that a student lab has to work for years (decades even) and labs get handed down from others and thousands of students using the same setups. The chances of something going wrong is far worse than a typical consumer application. (It is really obvious in the lab reports when something went "wrong", results are sometimes hilarious, and sometimes very strange)

If it takes 30 minutes to plug a USB cable back in and 10 minutes to find it how long is that USB cable, I find that absurd. You shouldn't have removable cables if that is your situation and recovery time. Just epoxy the cable in and fill in all the other ports and if it gets ripped out just replace the entire system. Why is the cable in the way of a janitor if your used to dumb proofing that is very poor placement. (Finger safe rated enclosure I think is what you need to keep your janitor out of the area)

[a210210200]

No user who doesn't know how to use google would be unaware of how to fix it. Also need help ask someone it works a lot of the time and for this problem most people should know how to figure it out.

Really? Are you serious?  Look around you and open your eyes.  There are lots of everyday users who are very competent using their devices, but don't need to know (or wish to know) the insides and how it operates. 

Let's not forget our customers/users/family members who didn't ask for their devices to be bricked.  How much time has been wasted by the end user because the device doesn't work today, but used to work a few weeks ago.

How many devices have ended up on the trash heap because they just stopped working? 

Just because members of this forum have an interest (and ability) in electronics/computing, we should never forget our customers/users/family members may not have that same ability/talent. 

FTDI has the right to protect their IP, no argument there.  However, a good corporate citizen that respects the final customer (the end user), would have merely notified of the suspect device - "Not an FTDI device - please contact your supplier/manufacturer for the correct driver."

FTDI does NOT have the right to vandalise my device.

Ozwolf

Open you eyes type "FTDI" into google and read the answers are there...

A user cannot be "very competent" if they are not capable of learning or using something called Google. At best they can be graded as having an "minimally acceptable level of basic knowledge with no further interest in the subject, C 61%". Any competent person should be able to use Google (or any other not as good search engine), otherwise people around them should assist them in carrying out said tasks and teach them who is able to assist them.

HDCP does the (it was working just last week and now it isn't) that much more frequently and there is no effective uproar (You so much as turn on the system in the wrong order and it will quit working for a while). FTDI did the same and the uproar is extreme and correct, although the consequences are in my view trivial, a joke, so far from having a killed device its miles apart, ....

Let me be clear I think FTDI is really mean/evil DRM type actions with that driver but the bypass is a joke sorry it is, ever install a printer that windows refused to install, same process. (There is a pictorial guide on that in the other thread)

I do family tech support and people do ask me to help, it isn't that hard and it isn't a technical skill to ask for help. (interpersonal skills its not a tech thing) Plus how many consumer devices that use the default VID/PID VCP driver are there.

Isn't the most common use a USB to RS-232 adapter cable what kind of non-technically inclined user is going to know today what an RS-232 serial port is or what a null modem cable does or what a huge difference is for TTL Serial signal levels.

FTDI put your device into a non-plug and play reduced functionality state by altering the PID to 0000, it is in no way physically damaged / vandalized (Now students and vandalism that is a real problem). I can help via remote desktop if someone is for some reason unable to follow the picture instructions for this issue. Also plug and play is not something that always existed and not all devices even today are plug-in-play.

A person who is unwilling to learn or ask for help is not going to get much done in at all. (This is broadly applicable and is not something that only certain people have to have)

But yes FTDI should not have left people in a HDCP your device half works just not in the way you really want it to state. They should have just caused a Microsoft driver error (generating custom error messages may add bloat but I'm not sure) and let the device driver fail for that combination.

End result is the same someone who is capable is going to have to look into it (There is no difference really, other than that one warning message you can still bypass it with the same instructions).

[Mr.B]

99% of the end users are not technical people.
They don't know what FTDI is or means.
They would not know to use that as a search term.
Most of them don't know jargon such as 'bricked' and so would not know where to start.

I 100% agree with ozwolf and Richard.

[a210210200]

99% of the end users are not technical people.
They don't know what FTDI is or means.
They would not know to use that as a search term.
Most of them don't know jargon such as 'bricked' and so would not know where to start.

I 100% agree with ozwolf and Richard.

Seriously no one is going to try typing that this device doesn't work?



Have you even attempted to see what a non-default PID says it isn't like windows doesn't give you nothing to go on. (Or is everyone just guessing that it vanishes from windows and says nothing, it gives you a bubble tells you the name of the hardware which you then google)

Type FT232R into google, FTDI. Microsoft asks you select mfg. google answers FTDI. (Doesn't require you to even know what it means windows tells you directly)

Googling are not reserved for the tech elite last time I checked its been many years since the search engine has gone mainstream.

I also highly doubt 99% of FTDI's chips are being used by completely tech illiterate people (prove it, reference(s) please).

[a210210200]

99% of the end users are not technical people.
They don't know what FTDI is or means.
They would not know to use that as a search term.
Most of them don't know jargon such as 'bricked' and so would not know where to start.

I 100% agree with ozwolf and Richard.

Also this entire point is moot, I think we all agree FTDI can have the driver refuse to work with a fake as long as it doesn't mess around with the PID. In the end a technically inclined user is still going to be required.

What we are talking about here is bypassing FTDI DRM which obviously should require a modicum of skill. (Like enough to install a printer with garbage drivers, very common)

[Mr.B]

99% is a totally inaccurate statement on my part and of course I can not prove it.
The point I am trying to make is that there will be a fairly large percentage of people who really only know how to use facebook and ebay, have purchased a widget from ebay that worked a couple of weeks ago and then suddenly does not.
They may choose to junk it.
They may choose to 'phone a friend'.
A lot of them don't have the skills to open and view the driver details, nor understand what it means.

[Mr.B]

99% of the end users are not technical people.
They don't know what FTDI is or means.
They would not know to use that as a search term.
Most of them don't know jargon such as 'bricked' and so would not know where to start.

I 100% agree with ozwolf and Richard.

Also this entire point is moot, I think we all agree FTDI can have the driver refuse to work with a fake as long as it doesn't mess around with the PID. In the end a technically inclined user is still going to be required.

What we are talking about here is bypassing FTDI DRM which obviously should require a modicum of skill. (Like enough to install a printer with garbage drivers, very common)

Agreed.

Oh, and printers are evil... Don't get me started on that subject... 

[a210210200]

99% is a totally inaccurate statement on my part and of course I can not prove it.
The point I am trying to make is that there will be a fairly large percentage of people who really only know how to use facebook and ebay, have purchased a widget from ebay that worked a couple of weeks ago and then suddenly does not.
They may choose to junk it.
They may choose to 'phone a friend'.
A lot of them don't have the skills to open and view the driver details, nor understand what it means.

If they can buy stuff off ebay they probably know how to use enough google fu to figure out that FTDI did something mean/stupid (social media, on twitter wasn't it, was where it exploded) also if they can use ebay and facebook they probably can use forums in general and google to figure it out. The listings typically like advertising their "brand real genuine good FTDI FT232" probably too so it would be plastered all over one of those highly likely to be fake listings.

I'd junk something that I find out is fake (other than the fakes I'm intentionally ordering for purely academic purposes of course). (AC power adapters that are fake/substandard are burn it with fire level of throw out, after tear down of course)(Very badly designed integrated lithum battery devices are also relegated to the this device may spontaneously explode sections disposal or re-work, wouldn't advise tearing down batteries without a proper setup)

Windows has a info bubble popup that says device not installed successfully and basically invites you to click on it and if you click on it says "Device driver software not successfully installed" "FT232R USB UART No driver found". Then even without knowing about the device manager right now they have enough knowledge use google to find help. Windows doesn't say the device is non-functional it says it has no driver for it and asks you to seek assistance from the mfg (which is going to get you an angry you should buy real parts as the response)(but the user probably will go elsewhere slightly or more agitated than before after getting told off in such a manner and find a real solution bypass the change).

Also this entire point is moot, I think we all agree FTDI can have the driver refuse to work with a fake as long as it doesn't mess around with the PID. In the end a technically inclined user is still going to be required.

What we are talking about here is bypassing FTDI DRM which obviously should require a modicum of skill. (Like enough to install a printer with garbage drivers, very common)

Agreed.

Oh, and printers are evil... Don't get me started on that subject... 

Yes lets not rag on those printers so much lost time.

But if it is possible the driver install error should just say "It's a faaaaaake" and then the google results would be more entertaining to take the edge off finding out that they were scammed.

i just saw this ... is it that easy to unbrick?

You don't need to do that to use the fake chips you just need to force the driver to install it requires no third party programs. Arguably for unskilled users it is easier to just tell them that is the only way and they will learn quickly to just keep doing it on other computers. Just look a bit earlier and it is just like installing a bad printer driver. Also why use MProg when FT_PROG works fine... (Mprog is not needed and FT_PROG is simpler). And if a user has a mac/linux/windows machine then they clearly should know at least a little bit about computers...

You literally just go into FT_PROG and go select default VID/PID, program, done.

[pickle9000]

99% is a totally inaccurate statement on my part and of course I can not prove it.
The point I am trying to make is that there will be a fairly large percentage of people who really only know how to use facebook and ebay, have purchased a widget from ebay that worked a couple of weeks ago and then suddenly does not.
They may choose to junk it.
They may choose to 'phone a friend'.
A lot of them don't have the skills to open and view the driver details, nor understand what it means.

I agree 99% is probably a little low, should be 99.99943%. 

[a210210200]

99% is a totally inaccurate statement on my part and of course I can not prove it.
The point I am trying to make is that there will be a fairly large percentage of people who really only know how to use facebook and ebay, have purchased a widget from ebay that worked a couple of weeks ago and then suddenly does not.
They may choose to junk it.
They may choose to 'phone a friend'.
A lot of them don't have the skills to open and view the driver details, nor understand what it means.

I agree 99% is probably a little low, should be 99.99943%. 

Still moot, 0-100% tech/non-tech use of FTDI parts doesn't change the pretty clear conclusion that FTDI can make the new driver block fake devices from working but not mess up the PID of fakes because that is just mean/evil stuff.

Also an unskilled user who cannot even use google is not going to notice the difference between the acceptable driver level block and a non-acceptable PID alteration which does the same thing just is mean/evil because it breaks plug-in play functionality on other OSes. If a person has windows, linux and a mac in the same room then they probably know enough to figure it out.

[Richard Crowley]

Wow, that philosophy wouldn't last two minutes in a proper customer-support organization. How astonishingly clueless about the end-user support world.

[ozwolf]

Wow, that philosophy wouldn't last two minutes in a proper customer-support organization. How astonishingly clueless about the end-user support world.

+1

[a210210200]

Wow, that philosophy wouldn't last two minutes in a proper customer-support organization. How astonishingly clueless about the end-user support world.

A proper product customer support person should be "We will RMA that part right away not sure why that didn't work" (internally panic that fake parts are in their products). But I'm not into product support I'm more into computer IT tech support. (Big differences in operation and objectives, its quite astonishing you simplify every support type into one blob)

And a proper computer tech support organization will start with have you:

1) restarted your computer?,
2) did you search the internal site for guides?
3) did you search for it on google?
and so on before continuing. Most problems would be fixed if you follow that order. Time is limited and IT computer technical support is supposed to solve actual problems not hand hold people through every little task they have to do with a computer. If a user is able to do something themselves then that is much more efficient and if they can be directed to ready made guides even better. Treating them like VIPs is not going to work out in a large scale support situation with limited staff/time/money.

IT support and product support are different (there is no such thing as a one size fits all "customer" or "support organization").

What would you say as a proper customer support oriented organization if someone called you up and said your product contains fake parts google told me, explain right now my IT people said that I should report you and I want a refund right now.