Lines in zig-zag

[c0depr1sm]

A zig-zag like this, is due to elastic behavior of the sled to triangular link to chains to beam to workspace.

It is more important at high center position where chain tension is higher and chain angles lower when the router gets closer to top beam.

I would recommend to make sure the beam does not flex. To observe that, attach a laser pointer on the top beam, pointing on a wall ideally several feet away. When moving the sled from top to bottom, any horizontal beam movement toward workspace center indicates a horizontal bowing where beam tips get closer.

Second, move the laser pointer to a nearby steady position and point it to the top beam tip face. Then move the sled from left to right along the bottom of the workspace. This reduces the previous bowing and shows the horizontal downward flex of the tips as the sled weight gets closer under.

Then the triangular link and chain clamps may flex too.

How stiff you need all of it depends on how much vertical force your router bit produces. When going from left to right, a vertical lift is induced, the other way around, it is a downward force. The deeper the router bit, a dull bit, a fast feed rate generates a stronger force.

Now this type of big wood chips is not homogenous, and that is likely to create vertical force variation along the path, creating a zig-zag.

[Mario_Giacummo]

Thank you, i am going to follow what you said, amlng others things

[Mario_Giacummo]

OK, I dont know what more I can do

The ring sled looks fine… the back plate is very smooth with just 3 little holes, the bearings the best I can buy, and as i can see everything is fine… what do you think?

New video by Mario Giacummo

I cut a 3 mm plywood pattern… very simple.

the dxf file is this 1-cuaderna.dxf (38.7 KB)

and the gcode generated, this one

1-cuaderna-d2g.ngc (10.3 KB)

the router I am using is a Makita https://www.makitatools.com/products/details/3709

and the results are this

May be this result is from the very little power the router have?

I dont know what to do, buy a new router or leave the proyect… I am beginin to be tired ;o)

Regards.

[Gero]

The g-code looks fine now. You still have a zig-zag in the dxf, but in so samll (~0.2mm) steps that it should not show. What does the inside of your ring look like? Do you have ‘moon-like’ craters on the inside of the ring?

Edit: I see you are going down with 2 steps, 1,5mm each cut. Are you stopping the g-code and manually adjusting the z-depth? What is the bit-size you are using?

Edit2: Is your sled balanced correctly? If not it could be tilting.
Balancing the sled. Move the sled the the middle of the sheet. Pull it away from the sheet with 2 fingers on the chains near the ring. Look at the angle from the side. Raise or lower the ring until the sled top is tilting forward towards the sheet. #4 in the picture.

[Mario_Giacummo]

Gero, you are everywhere !!!

I cut the ring from a 1.5 or 2 mm sheet… I think 2 (I am at work now), and the inside wall (that of 2 mm) is like a mirror, the outside is rusty, but i dont care by the moment.

the balance with the router is like nr. 4 … I think is ok this… and the total weight of the sled with router it could be around 9 k … 20 pounds so so.

I dont have a z motor, so when the sled arrive to the start point, the code stop, and I rise or pull the router…

I did the cuts with 2 bits, one 6 an the other 1 mm in diameter, the big one have two stright cutter blades, the small one is an in-out cutter… the errors are quasi the sams in boths

1 mm

6 mm

[Mario_Giacummo]

something new; this image have the same pattern draw horizontaly and at 45 degrees… at 45 degrees its quasi perfect, bot horizontaly it have bumps… the circles are perfect, the lines have a litle bup… the ellipse have a very small bump in the upper left side and the bottom curve looks ok… what could be wrong?, the sled run smoth, may be the iron ring could have so small bumps that i don feel?.. any suggestion? may be its something i think is fine but not…

[Gero]

I am pretty sure it is mechanical. But who am I? Just some crazy old guy, melting in the heat of Bahrain.
With the pen you have excluded the power of the router (Makita is good), and going in too deep and too fast (F600 first .nc and F400 second). Could it be the workflow Qcad->dxf2gcode? I doubt it (looking at the second .nc file), but test it by drawing a part (same size as your surfboard? part) with a long straight horizontal line (only 1 segment per side).
I think it is mechanical, the 3 places I would look and carefully monitoring during the cut:

• Are the chains on the sprocket side feeding without catching the teeth with a side. Even if the chain does not jump, it can slightly pull the sled up if it catches the tooth a bit. If a chain rides on a tooth, it either falls off or back into place and this would cause more visible zig-zag and sound. But perhaps a small friction could cause what you are seeing.
• The ring. Watch the bearings on the ring with a marker in your hand and mark the spots on the ring when the sled goes off the line. This way you can find out if the curves happen always on the same place of the ring. Drawing the same part 2 times at the same place could also help to tell if the curves are on the same spot or moving.
• The chain mounting points on the sled. Can the chain move freely in the last link attached?

Kind regards, Gero

[bar]

@Gero

I am pretty sure it is mechanical

I agree with @Gero

@Gero

The ring. Watch the bearings on the ring with a marker in your hand and mark the spots on the ring when the sled goes off the line. This way you can find out if the curves happen always on the same place of the ring. Drawing the same part 2 times at the same place could also help to tell if the curves are on the same spot or moving.

This is brilliant.

[Mario_Giacummo]

it seems what Gero say, I am going to confirm it soon.

[Mario_Giacummo]

Hi,

I remove de ring sled, and make an other triangular linkage… but the zigzag problem continue like with the ring linkage, exactly… so, the problem is not the linkage, its something else… but what?

I draw some lines and the horizontals show the problem more than the verticals and inclined ones… even more, the lines on the top of the frame are worse that thats on the bottom.

but there is something else and that is that the errors are reproduced exactly in the lines on the top… on the bottoma not so much

I am thinging that the origin of the problem could be the sprockets, with teeth to long… so they drag the chain under the escape line a little; i shorten them a little, but the simptons persist,.

May be the sprockets are not perfects?

the lines at the bottom looks better, but bad too

the diagonals are like the bottom ones…

could it be the sprocket?.. when i boght the motors i thought they come with them…

[Orob]

That looks like the router isn’t centered on the sled, the sled is sticking to the surface and as it moves across, it spins a little and the offset moves the bit off center, so the line is not straight / smooth. If the chains are catching on the motor sprockets, then the beam is not aligned properly with the height of the ring for the chains to be parallel to the surface of the cutting piece.

[madgrizzle]

@Orob

That looks like the router isn’t centered on the sled,

Good thought, that could quite possibly be it. My experience is that horizontal movement tended to be more jerky than vertical movement and if the center of the router is offset, I think that could produce what’s shown.

[Mario_Giacummo]

I dont know, but this thing is making me crazy… and now …when I move the sled 25 cms horizontaly it do the folowing

it go slowwwwli down, and when it is near to reach the end, it go up what it go down when moving… this is magical… I calibrate this machine 29 times… oh!!! may be the bit is not centered… lets see

I knew before buy this things that build it need a little work, little… my Good!!! …

one more thing, I begin to use web control to save steps in the garage… I am using a linux machiene with ubuntu… when i try to update de holy firmware (may be the Holy save me…) it said everything is ok, but when i try to move the motors they dont answer any command, everything is dead. I load the stock one and everything run… run, but run as you see above.

PD: the second photo show a raged surface, but not, I put a teflon plate there to make it move smooootthhhhh…

[Orob]

So is the sled getting caught on something or is the chain getting caught? It looks like it is moving left and riding up and then dropping down like part of the sled is catching and rolling past. I have seen similar marks when the frame piece that holds the work piece was out too far and the sled was moving along the bottom edge and caught it. I also saw this when the z axis cable hooked a cut section and stopped the whole system. Something is not allowing free movement. Are any screws or edges on the bottom of the sled sticking out?

[Gero]

This is something you should be able to see if you are watching . Hope you do not leave the Maslow alone when it is cutting. Either a tooth of the sprocket catching, or something with the ring. Sorry for the delay, Post-office is only allowing 3 people inside the same time and standing outside waiting with up to 50°C and a mask is a suicide mission. Will get the parcel with the ring kit shipped soon.

[Mario_Giacummo]

…i am working over a styrofoam surface, not z axis, no screws, nothing…just the foam snd the sled.
I am walking to the garage now to reset evrything i can…show you results later
Dont worry gero, yoy just did too much, and i dont beleave the sled support be the problem

Show you new wanderfull results soon

Thanks, regards

[Mario_Giacummo]

Each step is worst…
My cnc (the only thing I boght from maslow are the motors and a electronic plate, nothing else, everything else, I make it at home) support is a steel frame made of 1-1/2" square tubes, with a 21 mm 8’x4’ (2440x1220mm) osb table support over it, and 4 put in and out backs supports to keep it firm.
The distance between motors is 3262.5 mm,
The distance from top work area to upper motors line is 470 mm
The chains have 10 feets each one (3048mm).
I make to kind of sleds supports, a ring and a triangular linkage, now I am traying to calibrate the triangular linkage. the distance or radious of the arms are 108.5 mm , exactly.
The router is perfectly centered, there is a photo above.

Bien… I went to the garage, connect my computer, run webcontrol, and the first thing i did was wipe everything; where it said WIPE, I click the mouse.
After that, I load the the Stock Firmware (1.26)… ok… that ru ok.
then i setup again, the Maslow basic Settings.

In the Advance settings, the only thing I change where the teeth to 11

Now QUICk configure…nothing new… just linkage system…

Everything OK… now I wen to TRIANGUAR CALIBRATION

And here I begin to see lot of strange thing, the velocity of the sled is incredibe slow…when it reach the first corner, the motors didn’t stop making sound, they dont move, just sound…ok, they make the first cut… it go to the second and I saw the code and UOP!!! look…

Sending: G21
Sending: G90
Sending: G40
Sending: G0 Z5
Sending: G0 X0 Y0
Sending: G17
Sending: G91
Sending: M3
Sending: G0 X-966.0 Y+368.7
Sending: G1 Z-7 F500
Sending: G1 Y-25.4

there is a 25.4 mm to cut, but when i measured the cut… look

may be you cant see it, but it said 32 mm

then the code said…

Sending: G1 Z7
Sending: G0 Y-686.6

but it move 780 mm down

then said

Sending: G0 X1906.6 Y737.4

but it move…

and look the horizontal line…

anyway… it did all the work and when it try to return home…

Sending: G90
Sending: G0 X0 Y0
Client disconnected
[2020-08-20 17:03:21,927] ERROR in app: Exception on /triangularCalibration [POST]
Traceback (most recent call last):
File “site-packages/flask/app.py”, line 2292, in wsgi_app
File “site-packages/flask/app.py”, line 1815, in full_dispatch_request
File “site-packages/flask/app.py”, line 1718, in handle_user_exception
File “site-packages/flask/_compat.py”, line 35, in reraise
File “site-packages/flask/app.py”, line 1813, in full_dispatch_request
File “site-packages/flask/app.py”, line 1799, in dispatch_request
File “main.py”, line 421, in triangularCalibration
TypeError: ‘bool’ object is not iterable
2089.0
2086.0
208.0
1.0
Previous machine parameters:
Motor Spacing: 3262.5, Motor Y Offset: 478.7, Rotation Disk Radius: 180.5, Chain Sag Correction Value: 27.458973
Iterating for new machine parameters
N: 1, Motor Spacing: 3262.5, Motor Y Offset: -0.0, Rotation Disk Radius: 180.5, Chain Sag Correction Value: 27.458973
Chain Error Cut 1: -367.8217, Chain Error Cut 2: -16.6809, Chain Error Cut 3: -488.8277, Chain Error Cut 4: -142.8079, Sled Drift Compensation: 0
N: 2, Motor Spacing: 3262.5, Motor Y Offset: 183.911, Rotation Disk Radius: 172.16, Chain Sag Correction Value: 27.458973

… a lot of lines like this and…

Chain Error Cut 1: -23.2281, Chain Error Cut 2: 11.5175, Chain Error Cut 3: 17.5269, Chain Error Cut 4: 40.2803, Sled Drift Compensation: 0
N: 13, Motor Spacing: 3262.5, Motor Y Offset: 699.385, Rotation Disk Radius: 303.553, Chain Sag Correction Value: 27.458973
04.064, Chain Sag Correction Value: 0
Chain Error Cut 1: -0.0, Chain Error Cut 2: 0.0, Chain Error Cut 3: -21.0307, Chain Error Cut 4: 21.0307, Sled Drift Compensation: 58.1683
Machine parameters could not be determined
‘bool’ object is not iterable
Client disconnected

I have to wipe the motors too?

I did a lot of things in my life, lots difficults, lots easy, some so so… three sailboats, a pietenpol air camper, software servers, programs, gardens, cooking, a very small “normal” cnc… but this thing is realy mind blowing!!!

[Mario_Giacummo]

The sun is beginin to shine…
in short, the problem with the zig zag lines where the sprockets, I bought a new pair of them, and now the lines are stright everywhere… measuring the old ones, the distance from the bore border to the bottom of the chain contact, in some places had a difference of 0.4 mm… I thought 0.4 mm make no problem, but yes, it did.

by the way this post/problem could be close… but i have new ones :0)))

regards

[Metalmaslow]

glad you figured it out. I was going to recommend changing to the standard sprockets with 10T instead of your 11T ones.