Interesting PIC to PIC USART question

[NJ Roadmap]

Ok so I'm testing the system step by step and heres what I've come across:

1.Master is sending a 01010101 in a loop out of a a generic digital I/O pin
2.AM TX module modulates amplitude between about 10V RMS and 6V RMS
3.AM RX module on Slave gives pulsed output, but since I only have a multimeter I can't really say whether the high pulse out of the RX module's digital out pin is over 2V for TTL high.

This is because I need a fairly long high pulse for the multimeter to settle and show me the right voltage. I understand that if I do try to send too long a high pulse from the TX module, the output from the RX will fall low after a period of a few 100us.

I've tried varying the distance between the two modules for the same duration high pulse, and either:

1.the voltage for digital output high from the RX module decreases with increase in distance, or

2.the pulse duration gets shorter with distance - I can't say which again because of not having a scope and only a multimeter.

Ok so I'll get access to a scope in the morning but I'd rather have an answer now than wait till tomorrow. Basically what I want to know is whether these AM RX module's voltage for digital output high varies with distance from the TX module, or is it the pulse duration which shortens? If it is the latter I guess I'm going to have to go with Manchester coding after all...if it's the first, this is a much bigger problem!

Another couple of questions (if you don't mind ):
-the TX module can take up to 14V in and is currently powered via a 5V regulator, I'm thinking of giving it about 12.6V out of the 12V regulated mains PSU which powers the whole board. Would I need another 12V regulator just for the TX module or can I do without it?
-I got a couple of 1/4 wave 433MHz flexible whip antennas (monopole), one connected to the antenna pin on the TX module and another to the antenna pin on the RX module. I'm guessing that's good enough for about 5 metres through one wall?

Thanks in advance!

 

[NJ Roadmap]

bump..

..sorry.

 

[Nigel Goodwin]

Get the scope on and see what's happening, I've never used AM modules so I can't comment on them.

 

[NJ Roadmap]

I haven't looked at the scope but I have managed to get it working without Manchester coding and using the USART over a distance of about 10m through a 1-foot concrete wall with minimal errors. I am however going to use a modified version of Manchester Coding by sending 8 normal bits followed 8 inverted bits so that the number of 1's and 0's balance out.

Now all I have to do is remember how to convert an ASCII char into binary and invert it!

 

[Nigel Goodwin]

Now all I have to do is remember how to convert an ASCII char into binary and invert it!

What's to convert?, it's presumably already stored as a binary value in a GPR (or W) just send it a bit at a time (check my RS232 tutorial).

 

[NJ Roadmap]

I send 8 bits at a time by loading TXREG. I plan to send this:

11001100

followed by another 8 bits, except inverted:

00110011

Net number of 0's and 1's over each set of 8 bits are the equal.

 

[Nigel Goodwin]

I forgot you were using the USART, to invert the byte just XOR with 0xFF.

 

[NJ Roadmap]

thanks. i'm using the '~' command in C.

 

[NJ Roadmap]

P.s. do you know anything about antenna's and ground planes? i understand that having a ground plane close to the antenna base reduces losses..

 

[Nigel Goodwin]

P.s. do you know anything about antenna's and ground planes? i understand that having a ground plane close to the antenna base reduces losses..

A 1/4 wave whip is essentially a dipole, with the other half 'reflected' in the ground plane - so a 1/4 wave whip mounted on a ground plane works a LOT better than without one.

 

[NJ Roadmap]

I guessed so, since it was working almost as well without the antenna!

I'm guessing the ground plane has to be perpendicular to the whip? The whip is mounted parallel to the veroboard (at one end basically). Now I'm not sure how I can create a ground plane if it has to be perpendicular to the veroboard?

 

[Nigel Goodwin]

I guessed so, since it was working almost as well without the antenna!

Just the whip should make a HUGE difference to no aerial at all, far more than adding the groundplane will.

I'm guessing the ground plane has to be perpendicular to the whip? The whip is mounted parallel to the veroboard (at one end basically). Now I'm not sure how I can create a ground plane if it has to be perpendicular to the veroboard?

What distance are you looking for?.

 

[NJ Roadmap]

about 15m through 1 or 2 walls.

 

[Nigel Goodwin]

Shouldn't be any problem without a groundplane.

 

[NJ Roadmap]

It does 1 wall quite comfortably but struggles with 2 walls unless its aligned in a particular direction or two. I'd rather have 2 walls stablility without having to worry about alignment, even though I imagine that this will most probably have to work with only 1 wall in practice.