Let's Make Robots!

Decision on RF video link form robot to handset

This afternoon I finally realised, that transmitting baseband video via two XBee transceiver modules with their maximum data rate of 250kbps is simply not feasible, or well not to an acceptable quality. I read some chapers of Adrew LaMoth's book on the Hydra Game development plattform and some stuff on Wikipedia in order to understand the basics of NTSC and then I made a small calculation:

  1. NTSC video bandwidth is 4.2 ... 4.5 MHz. You need this even if you want to transmitt a black/white signal, because within this bandwidth you are able to control the luminance of each line on TV. As each line is drawn in 52.6us, there are only 52.6us * 4.5 MHz = 234 possible luma chances available on each horizontal scan. Not much.
  2. So we need to stick with the 4.5 MHz. If we want to sample this signal the minimum sampling rate would be the double frequency, so f_sample is 9MHz. In other words, there are 9 MegaSamples per second (MS/s).
  3. In case we do a simple 3-bit-quantisation for the magnitude of the baseband video signal then each sample consumes 3 bits, which have to be transmitted on out projected serial transmission line to the receiver. Hence, there are 9MS/s * 3bit/S = 27 Mbit/s which are simply a bloody 27000 kbps. Compare this number against the data rate of your preferred transceiver device and you feel sick...

At this point I noticed that my initial idea of having live video transmission from the robot onto my handset simply would not be feasible. For the geeks here - you can do the calculation above in reverse mode. If you have 250kbps available (like an XBee), then you'll end up with about 2 possible luma chances per TV line. Hahaha...

After that I had a beer. After the beer, I had an idea: are there any faster transceiver modules on the market available? Yes there are. e.g. this one with 11Mbps using SPI interface or this one with 2Mbps. Well this is way beyond the required 27Mbps.

After that I focused on ISM-band transceiver modules for 5.8GHz like this one (1Mbps) or this one. But they all do not come up with a nice data rate. Finally and more accidentally I got stuck here, where I found the solution for my way ahead.

I simply was so focused on digital video transmission, that I forgot the rather classical analogue way of transmitting videos. Following the above link you can find a number of dedicated transmitter and receiver modules for analog video transmission in ISM 5.8GHz band, made by a company named AirWave. They are all quite cheap and a European distributor is located in the UK. In order to get myself a clear overview I compiled a small sheet for comparing the devices, which is attached to this post. I also made some investigations on the legal aspects of using such devices here in Germany.

I will now order at least two Rx and Tx modules form the UK and try to get in touch with them.

AirwaveTechnologies_Overview.pdf199.38 KB

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Hi AnTenNnA,

I have used an AirWave Tx/Rx (630/634) setup, with a head-mounted display (node/22434).  It's been okay for short distances but not so good for flying videos.  The 630/634 setup does not handle 2.4GHz interference well.  I'm looking at RangeVideo (http://rangevideo.com/) for an upgrade.  I think part of the problem is the low transmission power, and part is a lack of good filtering or channel-switching when interference is detected.  Good transmission power and good handling of interference seems desirable in a video system.  Hopefully the 660-series/680-series is better than the 630-series.

Best of luck!



Thanks for your comment. I hope that the 5.8GHz modules will do better. I have heard about the interference problem of the 2.4 GHz devices, too. In any case I want to create a custom PCB layout. This opens the opportunity to integrate a printed antenna on the PCB. Maybe my access to professional RF simulation software can help. For now I have ordered the modules in UK. In the meantime I can work on the theoretical RF stuff.