Marlin / Anet - Skew Compensation not working

Hello,

I've spent a lot of time on this site and others trying to correct this.

I printed the test print as described in the Marlin firmware (see pic), I then inputted all the measurements (after doubling them) into the firmware.

I un-commented what I believe is supposed to be:

#define SKEW_CORRECTION

#if ENABLED(SKEW_CORRECTION)
// Input all length measurements here:
#define XY_DIAG_AC 319
#define XY_DIAG_BD 318
#define XY_SIDE_AD 199

I also made sure that #define XY_SKEW_FACTOR 0.0 was commented out. (i did try with the skew factor with the other lines commented out and that did not work either).

I also tried with the following both commented and un-commented.

#define SKEW_CORRECTION_GCODE

When I upload the firmware, with the #define SKEW_CORRECTION_GCODE and #if ENABLED(SKEW_CORRECTION) I can go to the terminal and type M852 and it will return a value (in my case 0.00322, which i double checked with the formula on wolfram).

It seems like octopi/the board is seeing the skew correction but it does not change my prints.

I've also tried hard coding it in the print file, also no results. I even incresed the value by 300% to see if there was a difference and there was not, however typing M852 in the terminal would bring up the increased value.

Any help would be greatly appreciated.

First thing I see is your diagram show you measured

XY_DIAG_AC
XY_DIAG_BD

but then you measured XY_SIDE_CD not the required XY_SIDE_AD (ie bottom of your diagram, not the right hand side.)

so the calculated AB will be all wrong... leading to a wrong XY_SKEW_FACTOR

My bad... the drawing is wrong... the 99.5 is AD not CD

Also multiplying * 2 is bs. you don't need to do this.

But it is irreverent. Doesn't change the results, just wastes your time.

I was following a tutorial and they mentioned to double it...but you;re right it would not alter the results.

Any idea why it keeps printing without the skew correction?

Not really. it all looks ok...
how many steps/mm on your X and Y axis?

As it only going to move 0.322 mm at the end on a 100mm object

most have 80 steps/mm so that is still 25 steps so should be noticeable, but if your steps/mm is much lower, perhaps your machine doesn't have the resolution.?

Try a test

Home the printer
send the printer a G1 X10 Y10 Z10
then start to watch your X axis for signs of movement. (perhaps add a tape flag on your stepper motor)
send the printer a G1 Y100
Did X move at all?

also try
Home the printer
send the printer a G1 X10 Y10 Z10
then start to watch your Y axis for signs of movement. (perhaps add a tape flag on your stepper motor)
send the printer a G1 X100
Did Y move at all?

I don't know which axis the compensation is applied to, but one of these two should move the other axis also

Thanks for the ideas... i did as you mentioned. In either case there was no movement on the axis that was being tested.

Looking at my picture, I'm thinking that maybe SKEW isn't what i want. Is there a scale option? My prints are straight but always elongated, perhaps i'm approaching this wrong.