FTDI's driver issues with fake chips and Parallax
Ken Gracey
Posts: 7,400
Hey all,
By now you've probably already come across the news about FTDI's drivers, the Windows updates, and what they do to fake FTDI chips. The internet is abuzz with discussion about how these USB bridge chips get "bricked" if they're imposters when new drivers are installed, whether or not FTDI's driver license should be able to disable imposters, etc.
I want to assure our customers that to the best of our knowledge all of our chips are original FTDI chips. This morning I looked at all of our products that use FTDI chips and did a bit of custody research with our Purchasing Manager and Asian supplier. Parallax was an FTDI distributor when they first came to the USA. Since then we've bought all of our chips directly from FTDI since we first started using them (maybe 12 years ago?). One concern I had are that we still have a few Asian-manufactured products, like USB to Serial adapters and the USB Board of Education. The chips we have used in China are also from FTDI, consigned and carried to the suppliers by Parallax. I don't think we've ever bought any Asian-manufactured FTDI imposters. It's difficult to be 100% certain unless somebody internally is personally interrogating every single purchase. For example, at times FTDI may have been in a backorder and we may have bought from Digi-Key. Digi-Key likely has better processes in place for screening imposter parts than Parallax since they're in the component business.
What you need to know is that if a Parallax product failed, we'll replace it with a return process that favors you, the customer. Warranty and support is one aspect of our business we hold as high as ingenuity and quality, so we'll never let you down.
We'll keep a close watch on the FTDI part supply chain and any trends we notice from you about returns or replacements.
Sincerely,
Ken Gracey
By now you've probably already come across the news about FTDI's drivers, the Windows updates, and what they do to fake FTDI chips. The internet is abuzz with discussion about how these USB bridge chips get "bricked" if they're imposters when new drivers are installed, whether or not FTDI's driver license should be able to disable imposters, etc.
I want to assure our customers that to the best of our knowledge all of our chips are original FTDI chips. This morning I looked at all of our products that use FTDI chips and did a bit of custody research with our Purchasing Manager and Asian supplier. Parallax was an FTDI distributor when they first came to the USA. Since then we've bought all of our chips directly from FTDI since we first started using them (maybe 12 years ago?). One concern I had are that we still have a few Asian-manufactured products, like USB to Serial adapters and the USB Board of Education. The chips we have used in China are also from FTDI, consigned and carried to the suppliers by Parallax. I don't think we've ever bought any Asian-manufactured FTDI imposters. It's difficult to be 100% certain unless somebody internally is personally interrogating every single purchase. For example, at times FTDI may have been in a backorder and we may have bought from Digi-Key. Digi-Key likely has better processes in place for screening imposter parts than Parallax since they're in the component business.
What you need to know is that if a Parallax product failed, we'll replace it with a return process that favors you, the customer. Warranty and support is one aspect of our business we hold as high as ingenuity and quality, so we'll never let you down.
We'll keep a close watch on the FTDI part supply chain and any trends we notice from you about returns or replacements.
Sincerely,
Ken Gracey
Comments
I also got embroiled in a ginormous twitter chain about it - coining it #FTDIGate and wrote a little thing on the current arguments out there for and against: http://tymkrs.tumblr.com/post/100749183024/ftdigate
On the one hand we have fake chips. Not good.
On the other hand we have the creator of the real chips disabling/damaging your systems that may accidentally use fake chips. Not good.
On the third hand. What on Earth is the problem with someone making a device that does what yours does?
That is not the issue. They are marked with the ftdi logo and name, and programed with a vid and pid that is used by the ftdi driver. It also appears to be made with completely different hardware. If the clone mfg had written their own driver and used their own name on the chip, this problem would not exist.
What a mess!
This brutal "it's not ours, brick it" play is going to the courts. I'll be watching with great interest. Some new and potentially painful law will likely come from it. I can't believe they would do that and somehow believe it would all be OK.
@Ken, awesome! Thanks for the support and ongoing great business. We all sure appreciate it.
But there are perfectly good laws for handling this type of counterfeiting. This is certainly true of the US, one of FTDI's prime markets, and the UK, where the company is located. Customs and revenue officials in both countries interdict illegal shipments all the time.
Sabatoging another person's property because it contains one allegedly non-genuine part, especially if that person has no reasonable way of determining compliance, is an egregious method of dealing with the situation.
["Update: Microsoft has given us a statement:
Yesterday FTDI removed two driver versions from Windows Update. Our engineering team is engaging with FTDI to prevent these problems with their future driver updates via Windows Update."]
I am not sure if you are disagreeing with, or misunderstanding what I said. Please enlighten me. I do not disagree with your previous statements.
I might have misunderstand your response to Heater. My point was this: Producing true counterfeit chips, including FTDI logo, etc., is against trademark laws, Lantham act (US), passing off laws (UK), etc. Pretty easy to get customs to seize those kinds of products without first having to go to court. Cloned parts that function like FTDI chips internally but are not marked as such are harder to get an interdiction action, I would believe. That would almost assuredly require actually going to court and proving infringement.
"...if the clone mfg had written their own driver and used their own name on the chip..."
What I find appalling about all this is that we even need a special driver for a USB/UART chip.
The USB standards define a communications device class (USB CDC). CDC is the specification for USB to UART adapters. Once I have a USB driver stack in my operating system with support for CDC I should be able to use any USB/UART adapter chip that implements the standard.
I do agree of course that creating fakes with fake names and logos is not acceptable, indeed there plenty of laws against it. However it is not acceptable that FTDI take the law into their own hands and cause the machines of innocent parties to fail.
FTDI is a great product and Parallax has done a great deal to support them. But if I were Parallax I would certainly begin to look for another USB device that doesn't have an 'avenging software driver'. The end users are going to resent the fact that whatever they purchased has been bricked.
FTDI could simply have the driver refuse to use the fakes... suggest that the owner seek out a driver elsewhere. Bricking is just too aggressive and very possibly illegal in some places.
I suspect the dubious Arduino clone market is one place where fake FTDI are likely to show up in abundance. These boards seem to come out of nowhere. Most companies that have a real identity want to provide the real thing.
One way for Parallax to avoid the issue is to compile their own driver, buy their own ID and use the FTDI-tool to change the ID on all products they ship.
Getting a Id is costly, but...
It would be nice to see 'Parallax Board' listed in the device manager on Windows. ;-)
It would probably make it easier to diagnose some issues, too.
You wouldn't, it's only for ftdi chips (or pirate clones). Would you use a Seagate utility on a Fujitsu drive?
Gordon: We agree on this.
Actually this is one area where FTDI stood out from the crowd. Although you can use your own ID that becomes a huge pain since you then have to manage all the drivers too. However, I have used the FTDI-Tool to change the description on the FTDI USB chip (leaving the ID codes alone) so that when Windows finds the device it shows the name that you want. I did this on my original HERO 1 USB interfaces and it was pretty cool to watch Windows "Find" a HERO 1 robot when you plugged in the interface. It should work for any of the Parallax devices but if they wanted to do it would add a lot of overhead since they would need to program that in each chip.
Robert
The similar (16KF micro USB) SiLabs C8051F38C-GQ* in an easy to use LQFP32 is 1,000: $1.35, so the days of custom USB do look numbered.
* The C8051F380/1/2/3/4/5/6/7/C have Integrated clock recovery; no external crystal required for full-speed or low-speed. These parts also have fractional baud UARTS, so can give the same .Baud granularity as FT23x series.
( The FT232RL shows at 2,000: $2.65, (ouch) and the newer, smaller FT230XS-U is 1,000: $1.46 )
We use a Silabs CP2105, which is USB-DualUART, (I think based on the older, slower OTP USBuC, 24MHz vs 48MHz for newer F38C above) and price is 1,000: $1.30
How do I +10000000 your +10000000?
Ken,
I think/hope I speak for most on this forum when I say that Parallax is a one of a kind company and sets the standards for quality, support and integrity!!! While I have to admit that it's been better the 15 years since I stopped working as an EE I've dealt with a lot of suppliers and vendors and when I discovered Parallax over a decade after moving from EE to IT I was shocked when I started working on a project that made use of a MicroC and started with a basic stamp and called asking about the diff between it and the propeller. Imagine my surprise when after asking a few questions of the first person I got on the phone I was asked to hold and after a few minutes ended up on the phone with you and at no time did I feel any rush by you to get me off the phone!! At that moment I knew I was dealing with a very special company.
This email about the dealings of Parallax with the FTDI chips confirms the feeling I got from you and Parallax when I first talked to you many years ago! I don't think that anyone that has dealt with Parallax would think that you would ever knowingly use counterfeit chips!!
Agreed! Ken's post at the top of this thread (and what if really means) is exactly why I love working with Parallax! What I mean is that this thread is not just a public relation's statement. It represents the way Parallax does business and how the company goes above and beyond to support customers.
I am by no means a big customer (meaning high volume), but Parallax has always treated me like I was and I see them do this all the time. There support of hobbyist as well as corporation is exemplary!
Thanks Parallax for being truly one of a kind!
And perhaps Parallax should step away from a company that has no respect for it's own or Parallax's potential customer base.
There must by now be usable and economical alternatives to FTDI devices that don't require messing with proprietary driver installations.
SiLabs have CP2110 USB-UART HID based at 1000 $1.2750
claims "Standard USB class device requires no custom driver"
- and I see they have a new CP2104-MINIEK complete USB-0.1" PCB for just $5 (shows RTS.CTS) (but the CP2104 does need drivers) - no mention of a CP2110-MINIEK ?
Come on over and get to work. I've got about 25 PCBs that need redesign, stencils and Pick-n-Place programs that need to be changed, BOMs that need to be updated, and a flurry of associated documentation and tutorials. You should be busy for about - oh, let me count - five engineer years. Then there's the software and drivers - by the time you finish that there will be a new set of challenges because Chromebooks and Macs could be supplanted by yet another hardware with its own OS.
I'm not poking fun at you, but I certainly thought about the consequence of such a change all weekend and it's far more complex than one might envision. Casual conversation about it is one thing, doing it is yet another.
Ken Gracey
Thank you, 4x5n! I wish I could talk to more customers every day. It seems my primary role puts me mostly in front of the educational customers.
Thank you for the post.
Ken Gracey
Ken Gracey
Now, I did say "step away from" not "run". I know very well that making such changes to existing designs is a huge effort. Besides it's a thankless task that nobody really wants to do even if you are paying them to do it.
New designs however are up for grabs. As you say "a different vendor on a new board."
And make it known why you have changed vendor.
This could be a wake-up call for us to diversify our engineering approach. I've always been told by some engineers (ahem, very senior ones) to have a second source. Until now I always poo-poo'd the idea, thinking that anything commercial would continue to exist as long as people would buy it. And that's basically true, but it doesn't account for other changes in supply you may wish to consider, like this case exemplified. I'm still an ardent supporter of FTDI and will use millions, but it's true that it's time to take a look at alternatives just in case.
For the Propeller, one reason we open-sourced the whole product was to allow other suppliers to exist should they actually decide to. Some larger customers have complained about a lack of second source. That's still true - only we make it - but at least it could be built by another company if desired.
Ken Gracey
Ray
Back in the day we were not allowed to specify any parts in our systems that did not have at least two sources. That of course was in the world of military / government systems where things are expected to be maintainable for a decade or two.
Around about that time Intel had to licence it's processor designs to AMD so as to provide a second source and thus allow Intel to supply devices to the military demands. A big and important customer when they started out.
In the commercial world people have not paid attention to this idea. In many cases this has cost them dearly as processors or operating systems went the way of the Dodo. Not to mention the myriad of other devices.
This is one of my major gripes about the world basing it's compute systems on MicroSoft.
There is a reason why we have standards for all kinds of things, mechanical, electrical, protocol, etc. So as to ensure that anyone anywhere can make stuff that adheres to the standard and thus provide a mass of second sources.
This whole FTDI thing is the antithesis of that concept.
Time to walk away.
I've seen an opening for SiLabs here, since they dropped the price of their Flash devices.
I've been pushing SiLabs to release a Library Source, that allows their cheapest flash USB, to 'drop into' Serial apps, as there are many 'other things' you can add to a smart-download part. I believe they are working on this.
eg Like the Reset-line control on the Prop, and you could maybe even power the Prop from the 100mA regulator.
Other things like an i2c slave ADC are in the part, just waiting some small code to enable.
SiLabs candidates :
16KFlash 48MHz uC, fractional Baud support - C8051F38C-GQ LQFP32 1K @: $1.35
and their standard bridge parts
CP2110 USB-UART HID QFN (1K @ $1.2750) claims "Standard USB class device requires no custom driver"
CP2105, USB-DualUART, QFN 1K @: $1.30 (also CP2104 etc) - needs Drivers.
Compare : FT232RL (2K @: $2.65) or FT230XS-U ( 1K @: $1.46 )
( Seems the older FT232RL has a NRFND price point, but the FT23xX series seem to be competitive)
SiLabs AN433/AN434 (CP2110) mention a DLL for HID wrapper, but suggests source for Win/Max/Linux.
I've not used the CP2110, so there may be practical caveats lurking in the USB-UART HID.
Looking at the lib calls, it should be simple to expand the Param values, to widen the usage.
The Prop2 is going to need some better means to download, and the CP2130 data suggests the underlying CPU and USB hardware, ( in Table 10. Typical SPI Throughput) can manage sustained > 2.6 Mbps (tho that is with their driver).
Truthfully, it's sorta a non-issue from our point of view. Our products use original chips and they work reliably.
So, we wouldn't be involved in any kind of boycott movement. In the meantime, keep sending us your money. We'll spend it developing things you want to buy.
Ken Gracey