RS232 – Sending/Receiving multiple characters

[viki2000]

I send 3 integers via RS232 using PIC10F322.
I receive them in PIC18F4550 using EUSART.
The baud rate is 2400. I display them on a gLCD.
The code works.

My question is how do you do it more simple than that?

What I do is I send in fact 6 integers: 3 with data and 3 integers (1, 2 and 3) for each real data.
They are sent continuously in a loop. I need to know which is the first data, second and third. To identify the data position I sent also the integer 1 for first data, 2 for second and 3 for 3 third.
The problem is that I cannot send data with integers 1 or 2 or 3 because is confused with identifiers.
How do you make the code simpler without identifiers associated with data stream in order to know exactly which data is?

And later I would like to extend to 4 or more data. Then instead of integers I would like to try float numbers and later strings with a known length in bytes. Then the situation changes for my code again.

One idea was for example at the sender to convert the data in string, concatenate in one string with known length in bytes and at receiver side to read the data as string and break it again in 1 byte strings which will be converted in integers or floats depending what was sent.
Any other more simpler ideas?

Here is the code for sender:

[temtronic]

You need to research both 'packets' and 'parsing'.

Packets refers to a group of data being sent or received. You might look at the Xbee modules to see how they do it. Generally you compose and send the following of only one transmitter to one receiver.

start of packet byte
qty of data byte
actual data bytes
end of packet byte.


The start of packet and end of packet bytes need to be unique and NOT to be confused with actual data. qty of data byte,1 - 255, should be more than enough for you. Actual data bytes, format depends on you.

There are hundreds of 'packet' formats,most will include a 'checksum' to validate the actual data. I'd keep it simple to start with though.


Parsing is the subroutine that 'decodes' the incoming data. It breaks down the various bytes into a known 'format'. You can search this forum for threads that deal with GPS to see how some programmers do it. Again, there are countless ways to do it BUT you should (must) use an ISR to acquire the data and a 'buffer' to store the incoming data . CCS does provide an example ( ex_sisr.c) in the examples folder.

So do some research, start small, test and build upon what you do.

Jay