We are going to be using a TTL RF transmitter to send commands via PC to a micro controller. What we are thinking about doing is putting an Arudino board between the RF receiver and the micro controller which would validate the serial TTL commands and if validated then the Arduino board would send the command to the micro controller.
The reason for this in the event the the RF receiver is out of range or there is a bad command sent the micro controller wouldn't receiver it since commands are being validated with Arduino. It also allows for failsafe, if no data is received within 3 seconds then it can execute a failsafe command.
Can someone recommend a board that I could use for this?
All we really need are two pins on the board, one for TTL input and another for TTL output. I've briefly checked the Arduino language and it looks like there is support for serial writing and reading.
Also, can arduino language write bytes from hex to a TTL output? Ideally we'd like to send plain text commands, the Arduino can read those commands then convert it to hex and write a byte out.
If the communication is purely in one direction, you can connect the serial RX pin (pin 0) to the module and the serial transmit pin TX to the controller.
If any of those two components expect to use both pins, you'll need two sets of serial pins. If the data rate is rather low and one of the communications is mostly transmitting, you can use the software serial library to use any digital pin as a serial device. In your case, I would connect in this scenario Pins 0 and 1 to the wireless module and send the data to the controller via softserial.
Any Arduino will do, take whatever fits your project best. If you fancy one with included space for the XBee module, good. If you prefer the cheaper Arduino Pro Mini where you don't pay for the FTDY-USB chip, also good. If you have any version laying around in drawer and you want to use that version, fine again.
If this data transfer is that critical, then what about adding a checksum to it?
You need to consider the probability of getting a valid command by accident from a corrupt stream of data.
Also why is another board inherently more reliable than just writing the validation software on the end microcontroller? The more components, the greater the mean time between failures for the system.
Hardware fails, software is either right or broken and testing tells you which
