[Prev][Next][Index][Thread]
Re: 16C84 driver for qcam
> How fast can a radio link be? I am controlling a 'robot' attached by
> wires. The various wires have to deliver 2 quickcam pictures one way and
> servo instructions the other.
> A suitably quick radio link could make the whole thing more
> useful and flexible. It would need to approach existing PC parallel port
> speeds? How fast is that? What radio links are available?
The system i have designed is very simple but has the advantage to
be extremely cheap. However, the max baud rate is around 300 byte/s,
so a picture takes just less than 4 minutes to go to the ground. This is
obvioulsy unusable for real time, interactive applications, but is good
enough for my project. Commercial radio modems (too expensive for me) can go
up to 9600 bps. My design is based on 8 NE555 oscillators triggered by
the output of 8 digital signals from an Io card. I sum these 8 sine waves
and send this signal to an FM emitter (US$20 Free Radio Berkeley FM xmitter,
illegal in .fr, but anyway ...). I then receive this signal on my scanner,
send it to a sound blaster compatible sound card, and use a stupid
time-frequency representation to find what the byte sent are. This is
a very crude and inefficient design, I agree. A PLL would be much
nicer and much more efficient, but I dont know how to build it :((
> Just a thought, but it might be cheaper and more flexible to use a large
> enough balloon to carry all the equipment, including harddisk, or tape
> for storage.
The main reason, although this solution would certainly be easiest, is
for safety (I dont care much about safety, but in this case i'll try
to be careful) : official regulations require that a balloon can be completely
destroyed (balloon and apparatus) if caught in a jet's reactor. The constraints
on the maximum size of the components and their weight are enormous (and
I'll obviously not respect them, and hence wont have my ballon checked by
the army as the regulations would require me to). But in order to have
a ballon as close to the specifications as possible, and the less visible
the better, a small, light balloon would be much better.
jmfriedt@ens-lyon.fr