Why the circle drawed by "G2 X390 Y-280 I5 J0" not round?

Typically the X axis joint would be 0 and the Y axis joint
would be 1 so the following errors would be:

joint.0.f-error
and
joint.1.f-error

(both are pins)

This is joint0 and joint1 f-error figure. It seems they are overlapped.Do their following errors has a problem?

But when the speed almost to be fastest that time, it seems has a clearly difference. 

 does this following error normal? Can this represent there is a position feedback from drive?

and ddt is the meaning of machine unit?

The following error is not meaningful unless the drive is
providing the actual feedback position

This hal statement connects LinuxCNC joint 0 position command to the drive:

net Xpos joint.0.motor-pos-cmd => lcec.0.0.TxPositionCommand

This (disabled) hal section looks like it attempts to provide the joint 0
feedback position to LinuxCNC

#net xpos-cmd joint.0.motor-pos-cmd => x_scale_cmd.in
#setp x_scale_cmd.gain [JOINT_0]PSCALE #invert of above
#setp x_scale_cmd.offset 0
#net PosCmdX-Scaled <= x_scale_cmd.out

#net RawPositionFeedback0 lcec.0.0.RxPositionFeedback => x_scale_fb.in
#setp x_scale_fb.gain [JOINT_0]PSCALE_INV #invert of below
#setp x_scale_fb.offset 0
#net xpos-fb joint.0.motor-pos-fb <= x_scale_fb.out

Isn't  this sentence the statement that provide the joint 0 feedback position to LinuxCNC?
net Xpos_fb lcec.0.0.RxPositionFeedback => joint.0.motor-pos-fb

looks like i sampled the the joint 0 feedback position to LinuxCNC, it's xpos_fb
     

Isn't  this sentence the statement that provide the joint 0 feedback position to LinuxCNC?
net Xpos_fb lcec.0.0.RxPositionFeedback => joint.0.motor-pos-fb

Yes, but that was commented out in your original hal file.

Looks like the X axis drive is poorly tuned
(not following the commanded position)

I provide the joint 0 feedback position to LinuxCNC through this sentence that net Xpos_fb lcec.0.0.RxPositionFeedback => joint.0.motor-pos-fb, and I didn't comment out it.

What's the meaning ddt  on bottom in halscope UI.

I find Xpos_fb's  waveform lags Xpos's waveform 50ms, but y axis hasn't it, what's the reason of this problem generally?
 

curran209 wrote:

What's the meaning ddt  on bottom in halscope UI.

It's the rate of change of the signal at the cursor position. ddt = dX/dt.