what is a 16x clock?

[Dr_Doggy]

i found this encoder/decoder,MCP2122, but what is this 16x clock, does this mean this needs to be clocked 16X faster than my data input? or does it divide my clock input pulse by 16 internally?



https://www.electro-tech-online.com/custompdfs/2010/08/21894c.pdf

 

[MikeMl]

You must feed it a clock. Its baud rate will be at 1/16 of the clock freq.

 

[Dr_Doggy]

so that means i can't send my serial data on each edge of the clock, I can only send it on 1/16 clock pulses?

 

[BrownOut]

The data will typically be asynchronous to the receive clock. By feeding it at 1/16 the clock frequency, the receiver has sufficient resolution to sample the data at the optimum time.

You'll need to use UART protocal, start/stop bits, etc.

 

[Dr_Doggy]

i was hoping to skip that part with the bits n junk, and just feed this one serially, maybe with a parity, I thought though that the data sheet said that it didn't need the protocol, due to the 16x clock,

it says that there is no protocol layer handler, and that the host uart baud rate selection is by the 16x clk, what does that mean?

also i thought i read it somewhere in there that it inserts the start/stop bits for me, i figured that was the whole point to the chip?

 

[BrownOut]

It won't insert the start and stop bits for you. That's not the purpose of the chip. It only encodes the received data to IrDa protocal. It will insert and control bits on the IR side of the device. But, if you're using an embedded UART, they it should insert the bits in the data transmitted from the host processor (also remove the bits from the receive data). As for not having a protocal layer handler, that probably refers to higer level protocal. I ddin't see that statement in the datasheet, but then again I only skimmed over it.