AlieExpress FTDI to be trusted?

[Jackster]

So I need some FT232 ICs, like over 100.

Pretty low budget project so looked at what AliExpress has to offer.
Found one seller doing them for $1.58 each, which when imported is just over a third the cost of getting them from local stockists.


Am I playing with the devil here?
I saw the video on the fake FTDI chips being killed by FTDI so I am sort of worried that these are  probably fake for the price.


Thoughts?

Thanks
Jack,

[ady.price]

the chip bricking problem doesn't exist anymore except if you have that driver version. but that doesn't mean they will all work.

[Monkeh]

Don't do it. Just don't.

FTDI do cheaper chips, and other manufacturers do cheaper still. Question if you really need FTDI.

[sleemanj]

I would go with the CP210x, or the CH340x rather than an FTDI unless you have some particular need to do so.

The CP210x does not need an external crystal, but it only comes in QFN package which might be a pain.

The CH340x needs an external crystal, but it comes in a more easily soldered package.

Both work fine, the CP210x drivers are probably better supported.

Cost is not too dissimilar, at least not for westerners.

[Monkeh]

CP2102 is also cloned

[citation needed]

so does CH340

Why would anyone want to make a cheap and nasty clone of an already cheap and nasty chip?

[Monkeh]

CP2102 is also cloned

[citation needed]

From HAD http://hackaday.com/2017/03/07/fixing-fake-ftdis/

There is fake CP2102 also, and they are worse: they ouput 4.4v even when configured for 3.3v operation. Nice when you use 3.3v only devices

Random comment on a Hackaday post (which makes no sense considering there's nothing on a CP2102 to 'configure for 3.3v operation' as such) holds no weight. Single faulty chip/board != cloned chips.

[bson]

FTDI isn't to be trusted.  Why willingly add a potential time bomb to your designs... I won't touch them with a 10ft pole.

[Jackster]

I have already have printed the boards to have the FTDI chip. Well only 10. Will be ordering a batch of 50 later on.
It also provides 3.3v so I don't have to also use a regulator for another component.
So that is why I want to use it.

Originally was going to use the CH340g but reasons lead to the FT232.


After thinking about it and some of the replies here, I am going to attempt to cancel my order on Ali and buy in bulk in the UK.
Thanks for the help everyone.

[Monkeh]

It also provides 3.3v so I don't have to also use a regulator for another component.

So do alternatives, FTDI and otherwise.

Ponder your component choice more carefully before ordering boards.

[FreddyVictor]

I've yet to understand why the FT232RL gets used so much, FTDI have an FT230X which is smaller and cheaper and, AFAIK, hasn't been cloned

'fraid my experience with CH340g is not good ...

[LukeW]

"I have already have printed the boards to have the FTDI chip"

Well, think about the availability of the components during the design phase.

Or if you don't want to take the risks with AliExpress, then don't - simply buy them from a reputable distributor.

FT230XQ is a much better choice IMO - smaller, cheaper, less external BOM.

[amyk]

https://www.eevblog.com/forum/projects/cypress-cy7c65213-a-pin-compatible-ft232r-replacement/ CY7C65213 may work in your design.

[Rick Law]

I've yet to understand why the FT232RL gets used so much, FTDI have an FT230X which is smaller and cheaper and, AFAIK, hasn't been cloned

'fraid my experience with CH340g is not good ...

I have at least 1/2 dozen Arduino NANO with CH340g.  Using them as a user has been uneventful.

Tinkering with them is another story.  So this may reflect the work at the design end may be more difficult.  I think the 340g's TTL-load capability various a lot between chips (see below for why and see if I am wrong).  Second issue: while the FT232RL datasheet said the 3.3v can support 50mA, I can't find out for sure how much the CH340g can support.  So 3.3v load is also an issue.

The reason I think the 340g's TTL output load capability varies between chips is this:

I attempted to connect Arduino's TX and RX to use both the Arduino's internal USB and a blue tooth (and only one in used at a time).  With the TX (send by Arduino's MCU), it is simple - one can have multiple listeners listening to the signal so the Bluetooth can just listen in on the signal.  Two "talkers" to the RX (send to Arduino's MCU) is a different story - even only one active at a time.  It means either connection must be able to pull the RX up or down while the internal chip may be at rest but connected.

I can make one configuration that works for all my UNO's and the FTDI NANO however I swap them.  With them, I can just plug in my modified Bluetooth card and both the internal USB and BT can listen to the MCU and take turn talking to the MCU.  But, with the CH340g NANOs, I can't.  I think I found one NANO that barely could.  In essence, each NANO (each 340g) will need adjustments to resistors to allow both the internal and the BT to pull the RX to the right state.

My conclusion is that  the TTL output load capability of the 340g varies too much.   Besides that Bluetooth experiment, I have done no further work to confirm/repute that assumption so my conclusion may not be accurate.  Experienced input welcomed.

[james_s]

FTDI chips from Chinese sellers are virtually certain to be counterfeit. Many of those do work just fine, but FTDI's effort to fight this by bricking them with a driver change put me off from ever wanting to use FTDI chips in anything. Genuine or not, it's just not worth the gamble. I understand their wish to disable counterfeit chips but they did it in a way that only hurt innocent end users and that is not ok IMO.