MoCap Dev Blog

http://www.custompcb.com/

Ok, created the first pass on the 6-DOF sensor module. I need to do a validation check to make sure the schematic for the gyros and the 3.3v are correct. Once I do that, I should be able to start creating the circuit.

Pretty good EE tool.

http://www.diptrace.com/download.php

Someone's mocap system that is close to what I'm building.

http://www.xsens.com/en/general/mvn

Figured out why 3ds Max wasn't reading in the data. Two things were causing problems. The first was the tri axis sensor outputs two values, the voltage of the force and an angle for the force. The angle data is wrong. I took the voltage data and the reference voltage and preformed the math. Once I was doing this process the data was working correctly. The second issue with importing the data into 3ds Max is that max didn't like the data being inserted directly into the scene from a MAXScript file. So, I found that File → Load Animation → .xaf file, is what needed to be used to load the animation. For some reason Max uses this process to load motion capture data.

I built a board and one tri axis sensor and I'm a little stumped as to why 3ds max is torturing the data. It's creating “noise” where there is none. So I started looking into the power supply. I found:

• 280ma propstick
• 7ma tri axis
• 150ma wifi
• total: 437ma
• 2.28 hours at 1000mah
• Energizer XP1000 5v @ 1000mah - $14

So this should satisfy any power issues. That and we can buy multiple and just change them out if the shot runs longer than the battery.

The zilog microcontroller is a pain in the but. I've come to realize that the better products out there are the ones with good help documentation and good samples to go along with the documentation. The issue is that I purchased the zilog and an accelerometer. The accelerometer was designed to work with the basic stamp. This meant that the accelerometer only uses one wire for communications. Well, SPI works by using two wires for the communication. So, I started looking into how the accelerometer works and discovered that the chip they used actually supports SPI communication. So, I added a wire to the input line and connected it to the zilog. Now, that was the easy part. Trying to write the code that supports the SPI functionality on the zilog was a long a torturous journey. The only example the company provides is one that doesn't support this specific processor. I found someone else's code, they had posted it in a forum asking for help, and started working from that. I slowly worked out what parts of his code worked and what parts didn't. It came down to one simple line of his code where he was checking for the SPI buffer to get full? and I removed it and it works perfectly. Now I can start working on the real problem, Motion Capture.