Problem with Inverse Kinematics of Kinova Jaco Arm

[Sj]

Hello,

I have a set of coordinates in space. I am using those as end-effector position inputs for the ikine and ikine6s functions. I normalize the points to lie within [-reach,reach]. 

I have a total of 657 points.

When I input the following (run in a script jaco_graphics) - 

inv_kin = jaco.ikine(T);

I get the following errors - 

Warning: Initial joint angles results in near-singular configuration, this may slow convergence
> In SerialLink.ikine at 139
  In jaco_graphics at 46
Warning: ikine: iteration limit 1000 exceeded (row 1), final err 0.119325
> In SerialLink.ikine at 158
  In jaco_graphics at 46
Index exceeds matrix dimensions.


Error in SerialLink/jacobn (line 62)
            U = L(j).A(q(j)) * U;


Error in SerialLink/jacob0 (line 55)
 Jn = jacobn(robot, q); % Jacobian from joint to wrist space


Error in SerialLink/ikine (line 194)
            J = jacob0(robot, q);


Error in jaco_graphics (line 46)
inv_kin = jaco.ikine(T);

I am not sure what's the issue here.

So I came across ikine6s and tried the following - 

inv_kin = jaco.ikine6s(T);

And I get the following errors - 

Error using SerialLink/ikine6s (line 61)
wrist is not spherical


Error in SerialLink/ikine6s (line 52)
            theta(k,:) = ikine6s(robot, T(:,:,k), varargin{:});


Error in jaco_graphics (line 46)
inv_kin = jaco.ikine6s(T);

Which again doesn't make sense, because Kinova Jaco has a spherical wrist and it's even defined as 'RRRRRR'

Can anyone help me out here?

[hung...@gmail.com]

Just type, 'help ikine' and 'help ikine6s' in command windows.

[Arthur]

Hello,

I do not think that the jaco arm have a spherical wrist. The RRRRRR does mean that all joints a revolute joints.

For a spherical wrist all revolute axes need to intersect in one point. Since there is an offset d5 they do not intersect in one point.

I am sorry if I am wrong.

Best regards 

Arthur