A CYBOT TECHNICAL SITE

Now I'm down to writing the software, I've never played with these data modules before, but I'm aware that there are a number of problems not found when doing a serial link down a piece of wire!.

Firstly you can't send a continuous high level, after a certain time the output of the receiver switches randomly between high and low, and then settles in the low position, so my initial change is to invert the data from the PIC, so it stays low normally (the opposite was true in my original RS232 routines). Having done this, I've been doing simple tests sending two bytes of data (actually 0xAA and 0xFF), separated by a low level for 20 bit-times. I monitored both the input to the transmitter, and the output from the receiver, using a 20MHz double beam oscilloscope, this showed that the receiver was outputting an accurate replica of the transmitted signal. I tried the scope with the data the other way, with a high between the data bytes, and about half way between the bytes the signal corrupted and fell to a low level.

Secondly you can't rely on 100% reliability, any interference, or obstruction between TX and RX could give data loss, either one bit or more, so it becomes difficult to be sure you are starting reading at the beginning of the data byte. I'm currently considering how I might best overcome this problem, searching the net I've found some interesting schemes for this, but I've not yet decided how I'm going to do it. I'm open to suggestions if anyone has any simple workable schemes?. At the moment I'm considering simply maintaining a long gap between data bytes, I could then wait for a low period of this length before looking for a start bit, as I'm not looking for terribly high data transfer rates this scheme would help to ensure correctly locating the start of the data.

Update - 29 November 2001.

OK, I've now played a while longer, and I've got simple working routines, I'm not doing any error checking yet (apart from checking the stop bit is correct), but it seems to work pretty well!. I'll try implementing the scheme I mentioned above (waiting for a long low between reads, my transmit routines already insert a suitably long low signal). But if you want to try it now, the routines are here - please note these are actually written using a couple of 16C84's (I have a number of them lying about!), to use a 16F84 simply make the following changes at the start of the files.

The transmit routine simply checks the condition of the 6 switches on the transmitter, connected to PortB pins 6 to 1, and transmits a code corresponding to which key is pressed. The receiver routine reads the incoming code, and sets Cybot accordingly. In this test routine I have buttons for Forward, Backward, Left, Right, and two buttons for Stop, you can easily alter them as you wish. Once a key has been pressed, the transmitter continually outputs that code, until you press a different key.

For aerials, I'm simply using two pieces of stiff wire, about 15cm long, I actually used the live and neutral conductors from 1.5mm twin and earth mains cable. To use the system, turn the transmitter on first, otherwise the random noise from the FM receiver makes Cybot do random movements.

I've now updated the receiver software to wait for a long low level, this seems to cure any problems, and it now doesn't matter which you turn on first. The original routine is still there to download, and the new 'MK 2' version is there as well. Following on from this, the latest version now includes the PWM speed control routines, I've set the turning speed to slow, this makes it much easier to try and point Cybot where you want to it go.

Cybot graphics used by kind permission of Eaglemoss