oldest usb to serial chip/when will the patent expire

[peterthenovice]

does anybody know when the first usb to serial chip was released? when will the patents expire. maybe when the first chip is patent free we will not have to deaal with ftdi/prolific anticounterfeit drivers.

[ataradov]

There is no patent on USB to serial converters, what are you talking about?

And FTDI is free to mess up their drivers until the end of times, there is no time limit on that.

There are plenty of genuine parts, but it makes sense to fake a well known brand.

[peterthenovice]

ftdi's die implementation is patented. any one can make a usb to serial chips, but they can not legally make clones of ftdi chips. if clones can be made legally then ftdi's chip implementation would become a de facto industry standard. ti , microchip , ONsemi, nxp could then make ft230X clones. like 555 timers ,74 series logic, etc.

[donmr]

A chip implementation can have a copyright and/or a maskworks but not a patent.  Patents are on a process and chip might use a patented process (dabatable) but the chip itself won't be patented.

[ataradov]

but they can not legally make clones of ftdi chips

No one is cloning the die. It is a pointless exercise. FTDI chips are MCUs with USB firmware. All clones are re-implementing the firmware on different MCUs.

And all those clones of 555 are not actual clones, they use different masks. They are just second source supplier. FTDI does not need second source suppliers.

[ataradov]

other semiconductor manufacturers could reimplement  the ft230X with their own die mask.

But why would they? They are free to implement their own thing. And why would FTDI not try to resist this?

There is always going to be differences, so FTDI will just make a new firmware update that will kill all the clones again.

You can freely create USB to serial adapters, just use your own VID/PID and write your own drivers.

You don't even need drivers. All OSes support standard USB ACM. I use it all the time with my own designs and it works perfectly. On Linux it just works out of the box. On Windows you need an INF file, but no actual drivers required, they are a part of the OS already.

[peterthenovice]

usb flash drives use a mass storage driver. usb to serial devices could use a cdc driver that works across devices.

[ataradov]

usb to serial devices could use a cdc driver that works across devices.

They already can. Why none of the manufacturers implement CDC ACM and prefer their own proprietary solution - ask those manufacturers.

[peterthenovice]

why do they not implement a cdc acm complatible device? i think the Microchip MCP2200 is cdc acm.

[ataradov]

why do they not implement a cdc acm complatible device?

I don't know.

are there any cdc acm compatible chips out there?

I may be wrong, but I think some Cypress IC implemented it. But I can't verify it right away, their site does not seem to mention it.

I've only seen ACM used in software implementations. For example, Atmel EDBG debugger IC uses CDC ACM for USB to UART translation.

[ataradov]

On further search, here is the quote from Cypress employee here http://www.cypress.com/forum/usb-full-speed-peripherals/cy7c64225-linux-devttyusb-support

USB driver stack in Linux has a built-in driver for CDC-ACM class devices. As CY7C64225 supports CDC-ACM class, CDC-ACM driver gets automatically bound to the device and creates a device node in /dev/ttyACM*(* -The name of device node will vary based on the number of devices connected). If one device is connected, /dev/ttyACM0 node will be created which will be used by the application to communicate with the device.

So here you go, one sane manufacturer out there.

[Monkeh]

Both Cypress and Silicon Labs make CDC ACM driven chips.

E: At least I think Silabs do. That may be my brain misfiring.

[peterthenovice]

on cyress's site for there usb to serial in the driver section if ound this.

http://www.cypress.com/products/usb-serial-bridge-controller
Cypress provides free USB CDC and USB Vendor Class drivers for USB-Serial Bridge Controllers.

[ataradov]

Silicon Labs

Part numbers? CP210x use proprietary protocol.

[Monkeh]

Silicon Labs

Part numbers? CP210x use proprietary protocol.

I may be mistaken. I'm sure I recall one of their chips operating under CDC ACM or a proprietary extension, but it may have been from another manufacturer.

It may have been one of the Cypress chips and I've got my wires crossed.

[peterthenovice]

https://www.silabs.com/Support%20Documents/TechnicalDocs/cp2104.pdf

it says in their datasheet that there drivers are royalty-
free not cdc acm.

[ataradov]

it says in their datasheet that there drivers are royalty-free

If it is real CDC ACM, then you don't need any drivers at all. It is possible that all their "drivers" are doing is just a simple install of INF files.

[peterthenovice]

http://www.microchip.com/wwwproducts/en/MCP2200

microchip supports cdc

[ataradov]

microchip supports cdc

Yep, one more to the list.

Now the challenge would be to find cheap Chinese IC with CDC ACM

[Monkeh]

I distinctly recall a chip which supported both CDC ACM and an extension thereof, but I can't seem to find any reference to it now.. Just ignore me until my marbles roll back.

microchip supports cdc

Yep, one more to the list.

Now the challenge would be to find cheap Chinese IC with CDC ACM

And then the challenge will be to get it working right!

[peterthenovice]

ch340 is the cheapest i think. but no cdc acm and bad mac os x support

[nctnico]

usb to serial devices could use a cdc driver that works across devices.

They already can. Why none of the manufacturers implement CDC ACM and prefer their own proprietary solution - ask those manufacturers.

The problem is that whoever made the CDC specification totally forgot about plain serial ports! The CDC ACM devices identify themselves as a modem to the computer. This is why Linux doesn't enumerate CDC ACM devices as ttyUSBx but ttyACMx.

[ataradov]

The CDC ACM devices identify themselves as a modem to the computer.

What difference does it make how they identify themselves? They work as COM ports just fine.

[oPossum]

I distinctly recall a chip which supported both CDC ACM and an extension thereof, but I can't seem to find any reference to it now.

Prolific PL2303 is CDC ACM with extensions to support all the handshake lines.

This can be verified by looking at the Linux drives for both.

[oPossum]

It is also mentioned in the PL2303 data sheet.

1.0 Features
 * Fully compliant with USB Specification v1.1 and USB CDC v1.1