LightBurn Demo 20 – Backlash Simulation

0:00 i thought it might be interesting for

0:01 you guys to see how i’m

0:03 checking for backlash and uh

0:06 analyzing the results so that we’re

0:08 getting um so i have

0:10 in our internal version of lightburn

0:12 this is the debug build

0:13 um we have a switch here in the preview

0:15 that turns on the backlash simulation

0:18 and so if i rewind all the way back to

0:19 the beginning and we zoom in on the box

0:23 so this is showing my it’s i think it’s

0:25 set to two millimeters of backlash it

0:27 might only be one i can’t remember

0:29 but it’s huge it’s way larger than you

0:31 would ever have in a real machine

0:33 and then the little red dot is the

0:34 crosshair that you normally see that is

0:36 the position of the laser

0:38 and so if i start going forward

0:41 you can see and if i turn on travel

0:43 moves so you can see that’s where the

0:44 laser started and this is where it’s

0:46 moving

0:46 and as soon as it hits the end of this

0:49 box

0:50 or the edge of the box it’s going to

0:52 start pulling the box with it

0:54 and the middle of the box is what’s

0:57 actually

0:58 being output by the laser so

1:02 we’ll scan forward a little bit and so

1:04 right there

1:05 this is the first point at which

1:06 backlash actually is

1:08 a visible problem if i scan forward

1:11 here we’re getting to the edge of the

1:13 shape that we’re starting to draw

1:15 and now it’s drawing the shape

1:18 and you can see the very very first

1:22 thing that it does

1:22 is change direction in the x-axis

1:26 so it’s now going in the opposite

1:27 direction and still going down

1:30 so at first because it’s still going

1:32 down

1:33 it’s pulling this center of the box down

1:36 with it

1:36 but it’s not moving in the same

1:38 direction in y or sorry in x

1:40 so we aren’t actually moving

1:44 the laser output yet because we haven’t

1:47 gotten to the other side of this box

1:49 so if your laser is a little bit loose

1:52 or you’ve got a little bit of play

1:54 in the gears every time you change

1:56 direction

1:57 you have to take up that slack

2:00 until uh it’s tight again

Video Transcript for LightBurn Backlash Simulation (Cont…)

2:03 and then the the actual laser beam

2:06 starts moving and so that’s what’s

2:07 happening

2:08 here and so this is the path being

2:11 sent by the controller to your motors

2:13 it’s the path that the motors believe

2:15 they are taking

2:17 this is the path that the laser head is

2:19 actually drawing

2:20 and as you get to the next direction

2:23 change which happens right here on the

2:24 x-axis

2:26 um you’ll see again now we’re starting

2:29 to move

2:30 this way through the backlash box

2:33 but the laser head itself is just

2:35 continuing to draw a vertical line

2:36 because we haven’t

2:37 eaten up that slop in the other

2:39 direction and so once we get to the

2:41 other side of this box

2:43 now it starts pulling the laser head

2:46 and so that’s why when you see files

2:50 that have been output by

2:52 machines that are loose they look like

2:54 these weird

2:55 ovals with flat sides

2:58 on top bottom and sides and they’ll have

3:00 strange gaps in them and so on

3:02 and if i run this simulation all the way

3:04 to the end and you look at

3:06 individual shapes in here you can see

3:09 there’s gaps between where it starts and

3:11 ends

3:12 gaps between where it starts and ends

3:15 this one same deal this one same

3:17 you may actually get shapes that are

3:20 quote unquote correct um but it’s just a

3:24 matter of chance

3:25 based on which direction the laser

3:27 starts and ends and which way that it

3:29 enters the shape and exits the shape and

3:31 so on

3:32 so the hide backlash optimization

3:35 if i turn that on and rerun the preview

3:39 you’ll notice that all of these shapes

3:42 now even though they still look like

3:43 hell

3:44 are actually sealed meaning the start

3:47 point and end point actually lines up

3:49 on every one of these shapes if you look

3:51 at it

3:52 um and it does this by deciding

3:56 how to enter the shape and where to

3:58 change directions

4:00 so that the backlash is

Video Transcript for LightBurn Backlash Simulation (Cont…)

4:03 minimized at the start and end points or

4:06 the start and end points have

4:07 the same backlash and that’s probably

4:10 the better way to think of it

4:11 so in this case the laser was moving

4:14 downward and to the right when it

4:17 started this shape

4:19 and so it continues moving downward

4:23 and changes direction because we have to

4:27 um but it ends the shape moving

4:30 downward and to the right so the start

4:33 point and the end point

4:35 actually match they meet up because

4:38 we started the shape moving down into

4:40 the right and if we end

4:41 the shape going in that same direction

4:43 the backlash will match

4:45 and then we head over to the next shape

4:48 and the next shape we arrive

4:50 at however it is that we get there

4:53 and we’re moving slightly upward you can

4:56 see

4:56 we’re going upward and to the right

4:58 again

5:00 so when this shape is finished we want

5:02 to be moving

5:03 upward and to the right in order for the

5:05 backlash to be minimized

5:07 and so again you can see here when we

5:09 get to the end of this shape we’re going

5:10 upward and to the right

5:12 and so the start point matches the end

5:15 point

5:16 we head over to the next shape and the

5:18 process repeats and so in this case we

5:20 were going down into the right

5:22 and so at the end of this shape we’re

5:23 going to go down into the right again

5:25 and so on and so on and so on and

5:29 this just works as long as your entry

5:33 and exit directions

5:34 for a shape are the same you should be

5:37 able to

5:38 not or you should not be able to see

5:41 the misalignment where the start points

5:44 and endpoints meet

5:45 um from everything that i’ve read and

5:48 researched this is correct

5:49 now that’s not to say that i’m perfect

5:52 and

5:52 that this is actually right i could be

5:55 wrong i’m

5:56 willing to accept that i’m wrong but

6:00 this is a reasonable simulation of

Video Transcript for LightBurn Backlash Simulation (Cont…)

6:02 backlash

6:03 and according to my simulation the

6:06 output is correct

6:07 so um until i find a way to prove myself

6:11 wrong i’m going to believe that this is

6:12 working

6:13 and the people that i’ve sent it to so

6:15 far aside from one user who’s been

6:17 having trouble

6:18 um i have all given back very positive

6:21 results and said that the output is

6:29 great