A 433 MHz 7 channel radio control system for antweight robots.

Radio control kit can be expensive. Scarily so. So, rather than immediately shell out a hundred quid or more for a commercial radio control system for antweight robots, I figured it was worth investigating building my own.

A radio control system has essentially four main components: (1)The encoder which takes information from the controls and assembles it into a form the transmitter can send. (2)The transmitter which takes the encoded data and sends it via radio to the receiver. (3)The receiver which extracts the encoded data from the radio signal. (4)The decoder which extracts individual channel information from the encoded data. All these components are straightforward enough to build using ordinary tools and test equipment. However the effort involved in building a 27MHz or 40Mhz transmitter and receiver is such that it makes far more sense to buy ready made transmitter and receiver modules.

RF Solutions make a range of hybrid low power transmitter and receiver modules intended for use in radio controlled key fobs, door openers and other short range data links. These modules are inexpensive and easy to use, which makes them ideal for a home made radio control system. The pair I bought cost me about 10 pounds from Maplin Electronic Supplies and work at a frequency of 433.92 MHz.

For my encoder and decoder circuit I dug out a 20 year old hobby electronic project book, "Cost Effective Projects around the Home", by John Watson. This book, which I bought at a remaindered book store in the 1980s, contains a variety of electronic projects including a 7 channel 27 MHz digital proportional radio control system. While the 27MHz radio circuits were of little use to me in this application, the encoder and decoder circuits based around readily available 4000 series CMOS integrated circuits proved to be simple and cheap to construct. I have no wish to infringe Mr. Watson's copyright so I can not publish the circuits he uses on this web page, but I can publish a block diagram showing their operation.

The encoder circuit is shown above. The main component is a 4022 8 step shift register whose outputs step through in turn 8 R/C networks. R/C networks 1 to 7 provide variable length pulses between 1 and 2 mS to represent the 7 channels and R/C network 8 provides a 6 mS sync pulse. The resulting pulses are combined into a single pulse train by a 4078 8 input NOR gate. A further R/C network provides a gap between pulses of about 250 uS and a 4001 quad 2 input NOR gate is used to clean up the output. A feedback line is taken from the 4001 to the clock input of the 4022 to ensure that at the end of each pulse the 4022 advances to the next R/C network.

The output from the encoder circuit was fed to the digital input of the RF Solutions transmitter module and when power was applied to the circuit and to the RF Solutions receiver module, the pulse train was found at the receiver output. The decoder fed from the receiver was very simple, a 4017 counter with an R/C network selected to detect the 6mS sync pulse feeding its reset pin. This provides the 7 different pulses generated in the encoder on its output pins in a suitable form to drive standard radio control servos.

Both circuits were assembled on standard strip board. The reciever is small enough to fit in the same space as a miniature radio control receiver, and the transmitter board has a socket to allow connection of variable resistors and switches for controls. Both boards are shown below, along with a control board with 2 variable resistors.

So far so good.
The circuit did not unfortunately meet expectations when it was tested in a real environment. Range with quarter wavelength antennas was a creditable 10 yards or so, but the instability of the super regenerative receiver design used by RF Solutions proved to be the system's downfall. At extreme close range the receiver overloads and produces garbage, while at normal operating ranges it proved impossible to drive a servo from the same battery as the receiver because the voltage drop induced by the servo starting up proved to be more than the instability of the receiver could take.