8 ms·

There's a term we in random ee blogs call this: being ftdicked If you run a windows platform and you introduce FTDI chips in your design schematic, you run afo

by kefka 11y ago

There's a term we in random ee blogs call this: being ftdicked If you run a windows platform and you introduce FTDI chips in your design schematic, you run afoul of using accidentally counterfeit chips. Digikey and Mouser both have had this problem of counterfeits in their supply chain. And we aren't discussing FTDI-compatible chips, that do not counterfeit or claim to be FTDI. But the FTDI kills them all. The safer alternative is to use a USB-serial chip that's cheaper, and doesn't harass users or nuke their chips: ch340G Datasheet: https://www.olimex.com/Products/Breadboarding/BB-CH340T/resources/CH340DS1.PDF https://www.olimex.com/Products/Breadboarding/BB-CH340T/reso...

Tomte 11y ago
And the real solution is to get a grip on your supply chain. Somehow with FTDI chips it's "normal", "okay" and expected to get counterfeit goods. And heaven forbid the original manufacturer doesn't support you in your innocence. But if someone ships you a fake ATtiny all hell breaks loose and the Chinese ebay seller must be punished, hours and hours are spent communicating with ebay, with PayPal, with other users' on web forums. Hobbyist electronics people are strange. Wake up and give Mouser hell if they send you faked goods!
- quadrature 11y ago
  Well, that's kind of what FTDI is trying to accomplish. You won't buy from the same supplier again after getting a counterfeit.
  - lultimouomo 11y ago
    I probably won't by FTDI either!
  - kefka 11y ago
    Or better yet, you just stay away from FTDI, knowing that their capriciousness can lead to customers' devices being bricked or disabled in various ways. The CH340G is really a nice and cheap chip. And I'm starting to see it in more established American designs too (as opposed to Chinese designed electronics).
  - gvb 11y ago
    For FTDI, their battle is becoming (has become?) a Pyrrhic victory. https://en.wikipedia.org/wiki/Pyrrhic_victory https://en.wikipedia.org/wiki/Pyrrhic_victory
  - deelowe 11y ago
    LOL! Don't buy from farnell, mouser, or digikey? They've all had issues with fakes getting into the supply. Are we supposed to just up and boycott them all? And, once I've exhausted the top 3 distributors, now where do I go? I'm sure I can just call up FTDI and order 10k chips on a weekly basis right? Actually, that's a pretty good idea. Anyone who works in supply chain, switch to ordering directly from FTDI. They'll soon realize the folly in this decision.
    - st3v3r 11y ago
      It's their responsibility to make sure that they're selling me what they promised me. If they can't do that, then maybe I shouldn't be purchasing from them.
      - deelowe 11y ago
        Ok fine. Link me the spec that allows me to determine whether an FTDI part is genuine and I'll gladly inform the supply chain, manufacturing and testing teams I work with to reference that spec moving forward. Until then, we have to avoid anything with an FTDI brand as there's no way to be sure. Going with whomever I think is the most reputable supplier and then praying my products aren't bricked at some point in the future isn't the solution.
        st3v3r 11y ago
        Again, that is all their responsibility.
        deelowe 11y ago
        Who is this they you are referring to? If ftdi doesn't provide specs, resellers can't create them out of thin air. My only option is to not buy ftdi.
    - quadrature 11y ago
      I didn't say it was effective, its a tricky problem to curb.
- rhinoceraptor 11y ago
  And what if the fake FTDI is in a piece of medical equipment keeping someone alive? Surely FTDI has no culpability in bricking it, they should have gotten a grip on their supply chain!
  - Tomte 11y ago
    Yes, they should have. And tested the function. The applicable standards actually demand it. Safety development is serious business. You cannot justify your "I want everything as cheap as possible, counterfeit or not" attitude with some hallucinated safety risk.
    - kefka 11y ago
      > Yes, they should have. And tested the function. The applicable standards actually demand it. I agree. I don't want to put counterfeits in my product. Now, show me FTDI's standards documents that indicate how to identify and test for fakes. (Hint: they don't)
      - Tomte 11y ago
        In this case it's obviously simple: you are not allowed to update drivers unless you have validated them. As far as I understand, this "behaviour" is totally deterministic and certain. You immediately notice the problem in any test system. Don't believe that safety-critical devices are allowed to install some drivers willy-nilly!
        creshal 11y ago
        > As far as I understand, this "behaviour" is totally deterministic and certain. You immediately notice the problem in any test system. FTDI does not provide any guide lines to third parties how to test whether their chips are counterfeit or not. The only way to test it is trying to use it with FTDI's Windows driver, and FTDI randomly changes the both the detection criteria, and the action done on detecting a counterfeit chip – sometimes it bricks them, sometimes it sends gibberish, you have no way of knowing beforehand. FTDI doesn't even guarantee that this behaviour is deterministic!
- kefka 11y ago
  But that's the rub: FTDI doesn't have any good process to identify fake chips, other than "install it in your board and wait and see". And if you've tried ordering from FTDI themselves, they routinely have shortages and backlogs going from 3-6 months out. Meaning you lean on reputable places like Mouser or Digikey. And you also glossed over "FTDI compatible" chips. They too exist, and do not claim to be a legit FTDI (the markings on the case are what make a counterfeit, not the underlying silicon). Yet the FTDI driver disables them all, in one way or another. You assume I'm talking purely about the hobbyist market: anyone who pays for a usb FTDI adapter at $1.50 should know it's likely a counterfeit.
  - daphreak 11y ago
    3-6 months is not that bad. Especially if it means you can guarantee that you get genuine parts. Most real projects can easily handle this lead time for production.
    - DanBC 11y ago
      6 months would be tolerated for a hard to get specialist part - a particular connector that's made in smaller batches twice a year, perhaps. But for a normal part? It's lousy.
- kabdib 11y ago
  Frankly, I don't think about counterfeit chips. If I did, and had a meaningful way to determine a counterfeit, then I'd not buy products with them. But there's no way for an end-user to do this. And since they're the ones being punished by the bricking (any supply chain movement is a second-order effect driven by upset customers), and since I lack the technical capacity to identify a knock-off, then my most rational course of action is to avoid anything that might be counterfeit . . . by insisting on a totally different chip. Not having FTDI in your design is a feature. And that's easily checked by plugging the device into a USB port and looking at the VID/PID. FTDI has a death wish.
- leoedin 11y ago
  I've purchased what I thought was an industrial spec microcontroller board from a legitimate and respected supplier, through Farnell, and it had a counterfeit FTDI chip on it. I built that board into a piece of test equipment that's likely to be used for 5 or 10 years. Fortunately this whole saga happened while I was still developing it - if their chip bricking driver had been released a year later, I'd probably not have known about it. Someone would connect the test PC to Windows Update in 3 years time, and the device would stop working. Counterfeit chips are a reality, and not one that people building devices on a small scale can properly mitigate against. Not buying chips on alibaba is common sense, but not buying pre-built modules from Farnell? If you're not going to buy them from Farnell, where are you going to buy them from?
  - Tomte 11y ago
    I'm not disputing that you can easily get a counterfeit chip without it being your fault. But why do you simply accept that? Have you even complained to Farnell? Read the sundry web forums, very few people complain to their distributor. Everyone bitches about FTDI. And that's what's really bugging me here. In the FTDI case the problem has obviously gotten out of hand. I can totally understand that the manufacturer wants to get tougher.
    - raphman_ 11y ago
      The problem is that FTDI ist taking it out on the customer, and hoping that "the FTDI driver bricked my board" somehow gets those customers to (1) realize that the problem is a counterfeit chip, (2) work their way up the supply chain until they find the party at fault (or get the supply chain to take care of the problem), and (3) manage to get a replacement for the product which uses genuine FTDI chips. Why not just pop up a small window with a message "Hey, this product seems to contain counterfeit chips. Please tell us where you bought it so that _we_ can take care of the problem."
      - jnky 11y ago
        Just FYI, but displaying a Message Box from the context of a Windows device driver is not trivial and would add significant overhead to the driver. I'd prefer it if FTDI weren't utter douchebags around this whole ordeal. I understand they have to make money, but since when is it a problem when devices are API-compatible with existing device drivers? As far as I'm concerned, this is a rather common and generally beneficial thing.
- jdietrich 11y ago
  Nobody has total control over their supply chain. Substitution of counterfeit parts can happen at many points in the supply chain. Until resilient die-level traceability is universal, the best you can hope for is to minimise the risk. FTDI do not provide any useful traceability systems. Rather than providing their customers with a reliable means to identify genuine parts, they have hugely increased the risk posed to their customers by counterfeit parts. http://www.publicintegrity.org/2011/11/07/7323/counterfeit-chips-plague-pentagon-weapons-systems http://www.publicintegrity.org/2011/11/07/7323/counterfeit-c... http://spectrum.ieee.org/computing/hardware/counterfeit-chips-on-the-rise http://spectrum.ieee.org/computing/hardware/counterfeit-chip...
- drzaiusapelord 11y ago
  >Hobbyist electronics people are strange. Popular moralism in a nutshell. "I can steal from big companies because they're evil! But if someone even samples one of my songs I'm lawyering up!" There's something almost psychopathic about how people just rationalize amorality because the entity they're attacking from is "big" or "different" Or "the other." I guess the mp3 wars are over with Google Music, Spotify, Apple Music, etc but for a long time it was creepy to hear these justifications from otherwise normal adults. It made you think, "What if they see me and my interests as 'the other'? What would they do without a conscience to stop them?"
gsteinb88 11y ago
How reliable is the documentation for the ch340g? I don't have experience with Olimex directly, and for production designs I'm wary of datasheets that are translated this poorly -- while they frequently are just fine, i've run into numerous cases where the translation has either introduced ambiguity or is just plain incomprehensible. And that's the rub: if i get a legit ftdi chip, I have a high level of trust that it will do what it says, and there's tons of documentation.
- ArbitraryHobo 11y ago
  > I have a high level of trust that it will do what it says Even after seeing them pull a stunt like this in their drivers?
- tremon 11y ago
  And how will you know that you actually got a legit ftdi chip?
  - gsteinb88 11y ago
    I dont. I'm not defending FTDI's actions w/r/t their driver, just pointing out that ftdi does currently fill the market niche of "reliable, good usb-serial devices", and while I'd love for someone else to provide equivalent chips without the shit ftdi pulls, no one else has.
- baobrien 11y ago
  Cypress and Microchip also have USB-UART chips that cost less than FTDI's parts and are well documented. The CY7C65213 is even a pin-compatible replacement for the FT232R and shows up as a USB CDC device.
  - gsteinb88 11y ago
    Nifty, thanks. I've been lucky enough to be able to default to microcontrollers with usb-serial built in recently, but that's lookkng to be short lived...
analog31 11y ago
Are the CH340G drivers open source? I bought a pile of them, but have been skittish about installing one because I can't figure out where the drivers come from. Granted, knowing that FTDI's drivers come from FTDI doesn't prevent them from messing around with them, but still, the CH340G seems like more of a mystery.
- welterde 11y ago
  The datasheet does not seem to be very specific on what protocol it actually uses. It does however state it is software-compatible to CH341 and for that there is a linux driver[1]. [1] https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/drivers/usb/serial/ch341.c https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux....
- tyingq 11y ago
  For drivers other than Linux: The manufacturer's website is http://wch.cn http://wch.cn Their drivers are here: http://wch.cn/index.php?s=/page-search_content-keyword-USB%E8%BD%AC%E4%B8%B2%E5%8F%A3-type-50-p-2.html http://wch.cn/index.php?s=/page-search_content-keyword-USB%E... They don't appear to be open source, but I believe they are signed drivers for both Windows and OSX.

by kefka 11y ago

Tomte 11y ago

And the real solution is to get a grip on your supply chain. Somehow with FTDI chips it's "normal", "okay" and expected to get counterfeit goods. And heaven forbid the original manufacturer doesn't support you in your innocence. But if someone ships you a fake ATtiny all hell breaks loose and the Chinese ebay seller must be punished, hours and hours are spent communicating with ebay, with PayPal, with other users' on web forums. Hobbyist electronics people are strange. Wake up and give Mouser hell if they send you faked goods!

quadrature 11y ago
Well, that's kind of what FTDI is trying to accomplish. You won't buy from the same supplier again after getting a counterfeit.
- lultimouomo 11y ago
  I probably won't by FTDI either!
- kefka 11y ago
  Or better yet, you just stay away from FTDI, knowing that their capriciousness can lead to customers' devices being bricked or disabled in various ways. The CH340G is really a nice and cheap chip. And I'm starting to see it in more established American designs too (as opposed to Chinese designed electronics).
- gvb 11y ago
  For FTDI, their battle is becoming (has become?) a Pyrrhic victory. https://en.wikipedia.org/wiki/Pyrrhic_victory https://en.wikipedia.org/wiki/Pyrrhic_victory
- deelowe 11y ago
  LOL! Don't buy from farnell, mouser, or digikey? They've all had issues with fakes getting into the supply. Are we supposed to just up and boycott them all? And, once I've exhausted the top 3 distributors, now where do I go? I'm sure I can just call up FTDI and order 10k chips on a weekly basis right? Actually, that's a pretty good idea. Anyone who works in supply chain, switch to ordering directly from FTDI. They'll soon realize the folly in this decision.
  - st3v3r 11y ago
    It's their responsibility to make sure that they're selling me what they promised me. If they can't do that, then maybe I shouldn't be purchasing from them.
    - deelowe 11y ago
      Ok fine. Link me the spec that allows me to determine whether an FTDI part is genuine and I'll gladly inform the supply chain, manufacturing and testing teams I work with to reference that spec moving forward. Until then, we have to avoid anything with an FTDI brand as there's no way to be sure. Going with whomever I think is the most reputable supplier and then praying my products aren't bricked at some point in the future isn't the solution.
      - st3v3r 11y ago
        Again, that is all their responsibility.
        deelowe 11y ago
        Who is this they you are referring to? If ftdi doesn't provide specs, resellers can't create them out of thin air. My only option is to not buy ftdi.
  - quadrature 11y ago
    I didn't say it was effective, its a tricky problem to curb.
rhinoceraptor 11y ago
And what if the fake FTDI is in a piece of medical equipment keeping someone alive? Surely FTDI has no culpability in bricking it, they should have gotten a grip on their supply chain!
- Tomte 11y ago
  Yes, they should have. And tested the function. The applicable standards actually demand it. Safety development is serious business. You cannot justify your "I want everything as cheap as possible, counterfeit or not" attitude with some hallucinated safety risk.
  - kefka 11y ago
    > Yes, they should have. And tested the function. The applicable standards actually demand it. I agree. I don't want to put counterfeits in my product. Now, show me FTDI's standards documents that indicate how to identify and test for fakes. (Hint: they don't)
    - Tomte 11y ago
      In this case it's obviously simple: you are not allowed to update drivers unless you have validated them. As far as I understand, this "behaviour" is totally deterministic and certain. You immediately notice the problem in any test system. Don't believe that safety-critical devices are allowed to install some drivers willy-nilly!
      - creshal 11y ago
        > As far as I understand, this "behaviour" is totally deterministic and certain. You immediately notice the problem in any test system. FTDI does not provide any guide lines to third parties how to test whether their chips are counterfeit or not. The only way to test it is trying to use it with FTDI's Windows driver, and FTDI randomly changes the both the detection criteria, and the action done on detecting a counterfeit chip – sometimes it bricks them, sometimes it sends gibberish, you have no way of knowing beforehand. FTDI doesn't even guarantee that this behaviour is deterministic!
kefka 11y ago
But that's the rub: FTDI doesn't have any good process to identify fake chips, other than "install it in your board and wait and see". And if you've tried ordering from FTDI themselves, they routinely have shortages and backlogs going from 3-6 months out. Meaning you lean on reputable places like Mouser or Digikey. And you also glossed over "FTDI compatible" chips. They too exist, and do not claim to be a legit FTDI (the markings on the case are what make a counterfeit, not the underlying silicon). Yet the FTDI driver disables them all, in one way or another. You assume I'm talking purely about the hobbyist market: anyone who pays for a usb FTDI adapter at $1.50 should know it's likely a counterfeit.
- daphreak 11y ago
  3-6 months is not that bad. Especially if it means you can guarantee that you get genuine parts. Most real projects can easily handle this lead time for production.
  - DanBC 11y ago
    6 months would be tolerated for a hard to get specialist part - a particular connector that's made in smaller batches twice a year, perhaps. But for a normal part? It's lousy.
kabdib 11y ago
Frankly, I don't think about counterfeit chips. If I did, and had a meaningful way to determine a counterfeit, then I'd not buy products with them. But there's no way for an end-user to do this. And since they're the ones being punished by the bricking (any supply chain movement is a second-order effect driven by upset customers), and since I lack the technical capacity to identify a knock-off, then my most rational course of action is to avoid anything that might be counterfeit . . . by insisting on a totally different chip. Not having FTDI in your design is a feature. And that's easily checked by plugging the device into a USB port and looking at the VID/PID. FTDI has a death wish.
leoedin 11y ago
I've purchased what I thought was an industrial spec microcontroller board from a legitimate and respected supplier, through Farnell, and it had a counterfeit FTDI chip on it. I built that board into a piece of test equipment that's likely to be used for 5 or 10 years. Fortunately this whole saga happened while I was still developing it - if their chip bricking driver had been released a year later, I'd probably not have known about it. Someone would connect the test PC to Windows Update in 3 years time, and the device would stop working. Counterfeit chips are a reality, and not one that people building devices on a small scale can properly mitigate against. Not buying chips on alibaba is common sense, but not buying pre-built modules from Farnell? If you're not going to buy them from Farnell, where are you going to buy them from?
- Tomte 11y ago
  I'm not disputing that you can easily get a counterfeit chip without it being your fault. But why do you simply accept that? Have you even complained to Farnell? Read the sundry web forums, very few people complain to their distributor. Everyone bitches about FTDI. And that's what's really bugging me here. In the FTDI case the problem has obviously gotten out of hand. I can totally understand that the manufacturer wants to get tougher.
  - raphman_ 11y ago
    The problem is that FTDI ist taking it out on the customer, and hoping that "the FTDI driver bricked my board" somehow gets those customers to (1) realize that the problem is a counterfeit chip, (2) work their way up the supply chain until they find the party at fault (or get the supply chain to take care of the problem), and (3) manage to get a replacement for the product which uses genuine FTDI chips. Why not just pop up a small window with a message "Hey, this product seems to contain counterfeit chips. Please tell us where you bought it so that _we_ can take care of the problem."
    - jnky 11y ago
      Just FYI, but displaying a Message Box from the context of a Windows device driver is not trivial and would add significant overhead to the driver. I'd prefer it if FTDI weren't utter douchebags around this whole ordeal. I understand they have to make money, but since when is it a problem when devices are API-compatible with existing device drivers? As far as I'm concerned, this is a rather common and generally beneficial thing.
jdietrich 11y ago
Nobody has total control over their supply chain. Substitution of counterfeit parts can happen at many points in the supply chain. Until resilient die-level traceability is universal, the best you can hope for is to minimise the risk. FTDI do not provide any useful traceability systems. Rather than providing their customers with a reliable means to identify genuine parts, they have hugely increased the risk posed to their customers by counterfeit parts. http://www.publicintegrity.org/2011/11/07/7323/counterfeit-chips-plague-pentagon-weapons-systems http://www.publicintegrity.org/2011/11/07/7323/counterfeit-c... http://spectrum.ieee.org/computing/hardware/counterfeit-chips-on-the-rise http://spectrum.ieee.org/computing/hardware/counterfeit-chip...
drzaiusapelord 11y ago
>Hobbyist electronics people are strange. Popular moralism in a nutshell. "I can steal from big companies because they're evil! But if someone even samples one of my songs I'm lawyering up!" There's something almost psychopathic about how people just rationalize amorality because the entity they're attacking from is "big" or "different" Or "the other." I guess the mp3 wars are over with Google Music, Spotify, Apple Music, etc but for a long time it was creepy to hear these justifications from otherwise normal adults. It made you think, "What if they see me and my interests as 'the other'? What would they do without a conscience to stop them?"

gsteinb88 11y ago

How reliable is the documentation for the ch340g? I don't have experience with Olimex directly, and for production designs I'm wary of datasheets that are translated this poorly -- while they frequently are just fine, i've run into numerous cases where the translation has either introduced ambiguity or is just plain incomprehensible. And that's the rub: if i get a legit ftdi chip, I have a high level of trust that it will do what it says, and there's tons of documentation.

ArbitraryHobo 11y ago
> I have a high level of trust that it will do what it says Even after seeing them pull a stunt like this in their drivers?
tremon 11y ago
And how will you know that you actually got a legit ftdi chip?
- gsteinb88 11y ago
  I dont. I'm not defending FTDI's actions w/r/t their driver, just pointing out that ftdi does currently fill the market niche of "reliable, good usb-serial devices", and while I'd love for someone else to provide equivalent chips without the shit ftdi pulls, no one else has.
baobrien 11y ago
Cypress and Microchip also have USB-UART chips that cost less than FTDI's parts and are well documented. The CY7C65213 is even a pin-compatible replacement for the FT232R and shows up as a USB CDC device.
- gsteinb88 11y ago
  Nifty, thanks. I've been lucky enough to be able to default to microcontrollers with usb-serial built in recently, but that's lookkng to be short lived...

analog31 11y ago

Are the CH340G drivers open source? I bought a pile of them, but have been skittish about installing one because I can't figure out where the drivers come from. Granted, knowing that FTDI's drivers come from FTDI doesn't prevent them from messing around with them, but still, the CH340G seems like more of a mystery.

welterde 11y ago
The datasheet does not seem to be very specific on what protocol it actually uses. It does however state it is software-compatible to CH341 and for that there is a linux driver[1]. [1] https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/drivers/usb/serial/ch341.c https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux....
tyingq 11y ago
For drivers other than Linux: The manufacturer's website is http://wch.cn http://wch.cn Their drivers are here: http://wch.cn/index.php?s=/page-search_content-keyword-USB%E8%BD%AC%E4%B8%B2%E5%8F%A3-type-50-p-2.html http://wch.cn/index.php?s=/page-search_content-keyword-USB%E... They don't appear to be open source, but I believe they are signed drivers for both Windows and OSX.