BS2 architecture underwater ROV
ceruleanplains
Posts: 11
Figured I'd post this to show my progress with my underwater vehicle to date.
I'm using a BS2 architecture to monitor a PS2 joystick and then provide outputs to a custom 3-motor h-bridge. I'm in a spiral development cycle that just completed cycle 2.0 where 1.0 was just simple DPDT switches manually controlled.
This experiment demonstrates that the BASIC STAMP 2 has the ability to also do the PS2 control, and contrary to some sources, it handles the left analog joystick fine.
This video shows the debug screen to demonstrate the range of values
[noparse][[/noparse]youtube] http://www.youtube.com/v/Ay0KCS-3RaA&hl=en&fs=1 [noparse][[/noparse]/youtube]
This image shows the BS2 setup hooked into a LCD screen to show the values while not hooked into the computer and controlling the motors.
[noparse][[/noparse]img width=500 height=375]http://www.ceruleanplains.com/images/PS2controller/20090713_System.jpg" border="0">
The video to demonstrate the setup:
[noparse][[/noparse]youtube] http://www.youtube.com/v/zOx66FrmetM&hl=en&fs=1[noparse][[/noparse]/youtube]
This is the operational version at poolside. Note I forgot to turn on the MCU before starting control, then I turn it on and it works fine.
[noparse][[/noparse]youtube]
[noparse][[/noparse]/youtube]
Here Inky is struggling against the tether to pick up a ring on the bottom.
[noparse][[/noparse]youtube]
This time it does better after it's been straightened and the retrieval hook is tightened and lowered.
[noparse][[/noparse]youtube]
[noparse][[/noparse]/youtube]
You can note that I have a stability problem. Both from the tether's tension on teh vehicle and also the issues associated with a misaligned CG before ring retrieval (nose up) and after (nose down). I propose to correct for thsi with a 2nd vertical motor forward used to primarily control pitch and also assist with thrust. This is a delicate business of balancing the CG unfortunately and I may with the new weight push the CG a bit to far forward. So I'm curious if anyone has any thoughts on the matter.
At current my build 3.0 is to redesign the frame a bit to hold two vertical engines and install an accelerometer, an onboard camera, LEDs, and a compass for semiautonomy. The objective of this variant is that the robot will detect the ring and alert the operator with a strobing LED. The robot will also have finer attitude control by using control guidance to PWM the motors to maintain a heading. I'd also like to have it perform dead-recocking manuevers by integrating the accelerometer. If possible, I will also attempt to do some retrievals autonomously - but this is highly unlikely for this build.
Suggestions and comments desired!
I'm using a BS2 architecture to monitor a PS2 joystick and then provide outputs to a custom 3-motor h-bridge. I'm in a spiral development cycle that just completed cycle 2.0 where 1.0 was just simple DPDT switches manually controlled.
This experiment demonstrates that the BASIC STAMP 2 has the ability to also do the PS2 control, and contrary to some sources, it handles the left analog joystick fine.
This video shows the debug screen to demonstrate the range of values
[noparse][[/noparse]youtube] http://www.youtube.com/v/Ay0KCS-3RaA&hl=en&fs=1 [noparse][[/noparse]/youtube]
This image shows the BS2 setup hooked into a LCD screen to show the values while not hooked into the computer and controlling the motors.
[noparse][[/noparse]img width=500 height=375]http://www.ceruleanplains.com/images/PS2controller/20090713_System.jpg" border="0">
![http://www.ceruleanplains.com/images/PS2controller/20090713_System.jpg"](http://www.ceruleanplains.com/images/PS2controller/20090713_System.jpg"){kind=link}
The video to demonstrate the setup:
[noparse][[/noparse]youtube] http://www.youtube.com/v/zOx66FrmetM&hl=en&fs=1[noparse][[/noparse]/youtube]
This is the operational version at poolside. Note I forgot to turn on the MCU before starting control, then I turn it on and it works fine.
[noparse][[/noparse]youtube]
[noparse][[/noparse]/youtube]Here Inky is struggling against the tether to pick up a ring on the bottom.
[noparse][[/noparse]youtube]
This time it does better after it's been straightened and the retrieval hook is tightened and lowered.
[noparse][[/noparse]youtube]
[noparse][[/noparse]/youtube]You can note that I have a stability problem. Both from the tether's tension on teh vehicle and also the issues associated with a misaligned CG before ring retrieval (nose up) and after (nose down). I propose to correct for thsi with a 2nd vertical motor forward used to primarily control pitch and also assist with thrust. This is a delicate business of balancing the CG unfortunately and I may with the new weight push the CG a bit to far forward. So I'm curious if anyone has any thoughts on the matter.
At current my build 3.0 is to redesign the frame a bit to hold two vertical engines and install an accelerometer, an onboard camera, LEDs, and a compass for semiautonomy. The objective of this variant is that the robot will detect the ring and alert the operator with a strobing LED. The robot will also have finer attitude control by using control guidance to PWM the motors to maintain a heading. I'd also like to have it perform dead-recocking manuevers by integrating the accelerometer. If possible, I will also attempt to do some retrievals autonomously - but this is highly unlikely for this build.
Suggestions and comments desired!