Holonomic issues

[Ether]

Originally Posted by dwodrich

I have validated that the wheels are mounted correctly "x" from the top "o" from the bottom.

I have validated that we are sending in the correct vector in degrees. 0 for straight ahead.

We can move forward with 0 and backward with 180 when motors on one side are inverted. But we have no ability to strafe. The motors are each side are fighting each other.

When we change the motors so that none are inverted. We can strafe side to side with no issues, but we can no longer go forward or backwards.

Check polarity on all the jags.

Anyone have any ideas? seems like we are making a simple mistake.

TOP VIEW of robot:

front
1......2
.........
.........
3......4
rear

numbers in chart above refer to the wheels.
wheel#1 is front left
wheel#2 is front right
wheel#3 is rear left
wheel#4 is rear right


Put the 'bot up on blocks and test it with all six commands (fwd/rev, slide left/right, spin clockwise/counterclockwise) and tell us which direction each of the four wheels is being driven for the different commands.

Here's what you should get:

forward command: all 4 wheels spinnning forward

reverse command: all 4 wheels spinning reverse

slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward

slide left command: exact opposite of slide right

spin right: wheels 1 & 3 spinning forward; wheels 2 & 4 spinning backwards

spin left: exact opposite of spin right

download this PDF file and carefully study the chart at the bottom left of the page:

http://lib.store.yahoo.net/lib/yhst-...lSpecSheet.pdf


~

[dwodrich]

Not at the robot now, but I did that, when strafing both front were turning the same direction and both back were turning the same direction but the opposite of the front.

I am now thinking that maybe the PWM cables are switched front to bakc on one side. I think that would cause one side to fight the other in this way.

and changing the inversion of one side would then fix the side to side strafing, but brake the forward-backward motion.

[Ether]

dwodrich, I updated my earlier post with more detail:

http://www.chiefdelphi.com/forums/showpost.php?p=911502

[Alan Anderson]

Originally Posted by dwodrich

Not at the robot now, but I did that, when strafing both front were turning the same direction and both back were turning the same direction but the opposite of the front.

I know you said you validated the wheel configuration, but I urge you to check it again. If turning both front wheels in the same direction makes the robot go sideways, the rollers are apparently oriented the same way. Same for the back ones. Having them in an incorrect >> orientation could explain the behavior you described.

[Ether]

Originally Posted by dwodrich

when strafing both front were turning the same direction and both back were turning the same direction but the opposite of the front

What you described above is not a valid set of wheel commands for mecanum.

valid mecanum wheel commands explained here


~

[bakketti]

dwodrich - My team is having a similar problem. Left and right slide work but forward/backwards have the front and rear wheels fighting each other:

We are simulating the code on our computers with PWM indicators and a USB joystick just fine. So it makes me believe its a hardware issue. ... Not sure what though.

Keep us posted if you figure out something.

[dwodrich]

left side inverted:
forward command: all 4 wheels spinnning forward - goes forward
***this is what we get
reverse command: all 4 wheels spinning reverse - goes backward
***This is what we get
slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward
***we get 1&2 spinning forward and 3&4 spinning backward - won't go anywhere
slide left command: exact opposite of slide right
***we get 3&4 spinning forward and 1&2 spinning backward - won't go anywhere

When we do not invert one side we get:
forward command: all 4 wheels spinnning forward - won't go anywhere
***front wheel spinning forward back wheels spinning backward
reverse command: all 4 wheels spinning reverse - won't go anywhere
***front wheel spinning backward back wheels spinning foreward
slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward
***this is what we get - slides right
slide left command: exact opposite of slide right
***this is what we get - slides left

I will check tonight when we get to the robot. We must have the left side of the robots PWM cables reversed on the digital side car. We are using the electronics board from last years robot and I know the PWM cables are bundled and zip tied. I also don't remember checking the digital side car end. We called out the PWM location by looking at the label we put on the PWM last year.

Thanks for all the idease! I'll post the solution when we get it figured out, hopefuly tonight.

[Ether]

Originally Posted by dwodrich

left side inverted:
forward command: all 4 wheels spinnning forward - goes forward
***this is what we get
reverse command: all 4 wheels spinning reverse - goes backward
***This is what we get
slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward
***we get 1&2 spinning forward and 3&4 spinning backward - won't go anywhere
slide left command: exact opposite of slide right
***we get 3&4 spinning forward and 1&2 spinning backward - won't go anywhere

When we do not invert one side we get:
forward command: all 4 wheels spinnning forward - won't go anywhere
***front wheel spinning forward back wheels spinning backward
reverse command: all 4 wheels spinning reverse - won't go anywhere
***front wheel spinning backward back wheels spinning foreward
slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward
***this is what we get - slides right
slide left command: exact opposite of slide right
***this is what we get - slides left

I will check tonight when we get to the robot. We must have the left side of the robots PWM cables reversed on the digital side car. We are using the electronics board from last years robot and I know the PWM cables are bundled and zip tied. I also don't remember checking the digital side car end. We called out the PWM location by looking at the label we put on the PWM last year.

Thanks for all the idease! I'll post the solution when we get it figured out, hopefuly tonight.

what do the wheels do when you give a spin command (clockwise and counterclockwise)?

If you map out what each of the four wheels are doing for each of the six commands, you can figure out what wires to swap, or, alternatively, you can fix it in software.


~

[dwodrich]

Originally Posted by dwodrich

left side inverted:
forward command: all 4 wheels spinnning forward - goes forward
***this is what we get
reverse command: all 4 wheels spinning reverse - goes backward
***This is what we get
slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward
***we get 1&2 spinning forward and 3&4 spinning backward - won't go anywhere
slide left command: exact opposite of slide right
***we get 3&4 spinning forward and 1&2 spinning backward - won't go anywhere

When we do not invert one side we get:
forward command: all 4 wheels spinnning forward - won't go anywhere
***front wheel spinning forward back wheels spinning backward
reverse command: all 4 wheels spinning reverse - won't go anywhere
***front wheel spinning backward back wheels spinning foreward
slide right command: wheels 1 & 4 spinning forward; wheels 2 & 3 spinning backward
***this is what we get - slides right
slide left command: exact opposite of slide right
***this is what we get - slides left

I will check tonight when we get to the robot. We must have the left side of the robots PWM cables reversed on the digital side car. We are using the electronics board from last years robot and I know the PWM cables are bundled and zip tied. I also don't remember checking the digital side car end. We called out the PWM location by looking at the label we put on the PWM last year.

Thanks for all the idease! I'll post the solution when we get it figured out, hopefuly tonight.

It was the PWM cables. The ones on the left side were reversed. Drives like a charm now.