Previously I discussed a first look at the Vexplorer robot. Thanks to the help of the folks at the Picaxe forum, I have finally made some progress in decoding the RC protocol used on this robot. Looking at the microprocessor, it seems that PIN20 is the processed output to the transmitter stage.
In order to tap into the transmitter section, it will be necessary to either solder directly onto Pin20 or somewhere nearby. Since the EM78P458 is a surface mount chip this might be tricky. Unfortunately it seems like it connects directly to a surface mount resistor, so no help there! I will have to spend more time tracing the circuit to see if there is another point where I can bring the picaxe signal into the transmitter section.
Look at the two following pictures:
Apologies aside for the blurriness of the first image, the second image has one set of peaks that are more widely separated in time then the first. Note on the second image, The cursors are showing the wider gap to ease identification.
The remote uses a PPM encoding system with a 20ms frame and 300us pulses. (Note all values are approximate!) The scheme works by varying the spacing between the pulses. At neutral, each of the pulses is approximately 1 ms wide. As the control is moved on way or the other the width of the spacing is varied from 0.75 to 1.5 ms. In this case there are 9 pulses which makes for eight spaces. This means the remote could theoretically control eight channels. Since the remote only has six channels this is somewhat strange! After the 9 pulses with the appropriate spacing there is a mark space to make up the rest of the 20 ms.
Here is a picture of a series of two pulse streams in their entirety with the beginning of the third.
There is an excellent explanation of this scheme at Howard Sullivans site.
The next thing is to get the picaxe to generate this pulse stream. This should be a pretty straight forward adaption of the techniques I used in the Robosapien project. I am not sure if I need to encode the carrier or not? (It seems like the transmitter section might do this?)
[...] First you need to hack into the Vexplorer remote. Open it up and remove all screws so you can expose the back side of the PCB. You will need to cut (destructive) or desolder (Potentially reversible) the top pin of the large IC from its pad. Solder a wire to the small surface mount resistor pictured. You will input your pulse train to the RF section of the electronics at this point. The required pulse train is discussed here. [...]
Pingback by profmason.com » Vexplorer Computer control — January 5, 2008 @ 11:37 pm