MINIATURE R/C RADIO

I got the idea for this project when I saw these nifty joysticks in a trade magazine and was thinking of

possible applications for them. I had also been wanting to do something with the Micrel single chip

RF receivers that Tom Cantrell wrote about in the Jan. 99 issue of Circuit Cellar, which can enable

someone with absolutely no RF experience to design a wireless device.

This R/C radio and receiver is meant to be an ultra-cheap 'throwaway' product for the increasingly

popular class of micro R/C airplanes, which can be flown in the length of typical front yard or large

living room. Included are photocopies of a current hobby magazine and a typical microlite aircraft.

This whole device with receiver costs under $5 to manufacturer in any kind of quantity. This is much

different from the $150 or so for conventional high-power four channel radios that currently have to

be used even for micro aircraft. Because these micro craft are so small and fly so slow (in many cases

they can't even be flown outside because even light winds are too much) the range required is small

and there is no danger of losing radio signal.

I envision a kid being able to get a real, complete, 4 channel R/C airplane for $20-$30 and fly it with

the ease of a Guillow's rubberband glider that I buy at 7-11 for $1.99. This is exciting to me because

when I was a lad I wasted alot of time and money in the elusive goal of flying a plane, saving and

crashing $225 or so at a time (that's another story).

I think of the Micrel MICRF011 single chip RF receiver as a wideband receiver with a signal

strength indicator. With my 4.9 Mhz crystal the chip operates at around 315 Mhz. You use the

device by simply turning Off and On (OOK) a source of 315 Mhz RF at a meaningfull rate. (I guess

technically this is AM). The Micrel chip receives this and its ouput goes to an on-board comparator

to distinguish what is 'on' and what is 'off'. The comparator's CMOS output goes to pin 8. It is up to

the user, through external capacitors and jumpers, to configure the chip for its intended use. The user

still has to develop a modulation scheme for his OOK data because there would be no way to distin

guish a simple 'carrier on=1/carrier off=0's '0' state from simply no data being sent. And with the 315

Mhz off, there is no way for the chip to distinguish between what it should be and the background

noise. So a modulation scheme is used. A 'burst' technique can be used where 50% duty cycle 315

Mhz RF is transmitted and capacitor CTH is charged to a reference level, and then data-modulated

OOK RF is sent to out as a small segment of serial data, while it has the previously charged CTH

reference to work against. In this case a large capacitor for CTH (which is connected to the comp

reference) would require a very long preamble to charge it and too small a capacitor wouldn't support

any decent length of data. So a compromise has to be made on this capacitor value.

I thought I'd avoid all of that by just using a continuous manchester data stream with its 50% constant

duty cycle, especially since the radio transmitter needs to continuously transmit joystick positions

anyways, unlike, say, a garage door opener, which is a 'burst' application.

The physical design of the transmitter was really determined by cost. It's small because small is

cheap. It feels very natural in the hand with the joysticks operated by both thumbs, much better than

cumbersome conventional RC transmitters. The receiver is meant to be absolutely small and light as

possible, and it is many times lighter than the lightest R/C receiver out there. Unfortunately I could

only obtain DIP '51LPCs, but this will be changed to SOIC with my next board revision.

.