|
|
Post by bonylu on Aug 25, 2016 17:39:25 GMT -7
Hi Max,
Thanks for your great project!
My hardware is ready.
Now the first thing is to calibrate accelerate meters.
I can get the output data with python scripts.
But how can I do calibration? Pass the angles value to some variable?
And any picture of install the accelerate meter?
Thanks
|
|
|
|
Post by Max on Aug 26, 2016 20:42:10 GMT -7
First, which board are you using, ramps or fpga?
|
|
|
|
Post by bonylu on Aug 27, 2016 2:34:20 GMT -7
ramps 1.4
1) What is the detail step of calibration ?
2) I design a mounter for GY-521,but dont know the correct install direction, or any direction is correct?
I'm afraid wrong x/y/z axis will get wrong calculation result.
Thanks,
|
|
|
|
Post by Max on Aug 27, 2016 6:51:02 GMT -7
1. Mount accelerometers as parallel to the arm link as possible. If mounted carefully no calibration may be required. Let me know if you still need to calibrate. 2. Please refer to github.com/maxosprojects/open-dobot/wiki/4.-RAMPS where it stated that the connection holes are expected to point to the back of the robot. Same way the stock accelerometers are mounted. |
|
|
|
Post by bonylu on Aug 28, 2016 5:11:50 GMT -7
I take some photos in following url tieba.baidu.com/p/4755408144Do I mount the accelerate meters right? Dose the output angel data reasonable in such status? |
|
|
|
Post by Max on Aug 28, 2016 22:20:25 GMT -7
The data is, obviously, wrong.
Accelerometer on the rear arm link (the one linked to the base) is backwards - the contacts of the IMU need to look down, not up (the holes of the IMU should be where the connector is and the connector - where the holes are).
This is, of course, a soft requirement as the data returned from Arduino to SDK is raw and you can just update SDK according to whatever is the location/direction of the accelerometers on your dobot.
|
|
|
|
Post by bonylu on Aug 29, 2016 17:42:37 GMT -7
I swap the rear arm accelerate meter direction. The output end effector height is positive while original data is negative. So it should be ok now.
1) Next question is that is there any python function I can control one step motor separately to make sure its rotation and connection are correct.
2) I use harmonic-wave reducer of 1:50, so I set the variable as 50, but I heard strange sound(gagagaga) when motor running. If use 10, no such sound. Do you have any clue?
|
|
|
|
Post by Max on Aug 29, 2016 17:56:32 GMT -7
1. Steps() in DobotDriver. For joints it expects values produced by stepsToCmdVal() or freqToCmdVal(). See example-driver.py
2. That is too much reduction. Although it's possible do that on Arduino Mega in expense of timers, but you'd have to do that yourself. The problem is that stepping for all three motors is driven by one timer and the faster you make it go, the less accurate the stepping signal is generated on arduino to the drivers. The options are to accept it going two-three times slower, change the reduction ratio or change the firmware and rewire RAMPS a little bit to utilize separate times for motors. If the latter then you'll have no more timers to do other things. That's Arduino Mega. If you check firmwares for arduino for 3D printers, they all have direct drive without any reduction.
|
|
