Deformation along to the top... Normal?

20171127_194119

Overall I’ve had pretty decent results with the Maslow. However, As I’ve started to test cut across the top of the frame I’ve had some deformity issues. Is this normal? If so, is there a known region of good space across the platform? I can always Cut down the sides and tops and move them to the middle if needed. Or is there a way to calibrate this problem out?

Any solution is fine with me, I started my testing with a cheap piece of plywood and a junky box design for testing. I’ve also ordered a linkage kit, though i’m not sure if that would solve the issue or not.

Thanks in advance for the help, you guys are great!

I generally planned my cuts to avoid a foot at the top and sides before beefing up my frame, haven’t re-tried those areas yet. @Bar seems to get good cuts all over his stock frame judging by the recent update videos cutting furniture. He’s got the touch.
That said, I wonder if one of the chains jumped a tooth during the cut in your picture. There’s a lot of tension when cutting at the top of the frame, and things like bowed plywood or a slight angle to a motor mount or a saggy slack-side chain can cause a link to ride up on a tooth. It’s too fast to see (Bar filmed it once, it jumped between frames), you just hear a loud pop and everything seems to go on as normal, but the cut is off from there. If you’ve painted the top links during the chain calibration, you can check for a jump when the sled is at 0,0 - the paint mark on one chain will be a whole link farther from it’s motor.

3 Likes

I’d bet that as you are cutting across the top, the frame is warping a bit, and
that is causing the errors you are seeing.

My frame seems pretty solid a jumped tooth sounds about right given the cut change. Maybe a stronger tensioner is in order?

1 Like

check that the chains are parallel to the workpiece and not angled as they get
to the sprocket.

2 Likes

More tension would cause more jumps. Try some of these things:

  • Check for warping as David suggests - I stretch a piece of dental floss across the top from one sprocket shaft to the other, just tight enough to be straight. If the chain tension bows the plywood or the arms or twists a motor mount, the floss will sag. Sight across the plywood from the side to see whether the plywood is flat or bowed. Thicker, straighter plywood or braces and stiffeners might be needed.
  • Look for whether the chain is parallel to the surface - might need to choose a different mounting hole at the sled end, or change the position of the motor mounts.
  • Look at the motor mounts, are they still firmly attached? I had one twist because I hadn’t used the full number of screws mounting it.
  • Look at the slack-chain handling arrangement when most of the chain is on the slack side - is the tensioner keeping the lead to the sprocket fair? Some folks have arranged fair leads to make sure the chain wraps around more than 45 degrees of the sprocket and that it doesn’t catch on the side of the teeth as it meshes.

That’s a list of some of the things that I’ve had cause disappointing cuts much like your picture. I blame the builder, uh, myself :wink:!

3 Likes

So once you jump a tooth, is that an automatic required re calibration?

Every time it happens to me that is my course of action. Well, that is after the normal round of cursing… >_>

Not sure if someone else has come up with a method of resetting it without recalibrating the chains. Maybe if you marked the chains when the sled is at 0,0, you could move the chain to the marked position to recover it? I haven’t tried it, but it may be a faster solution to the problem if it works.

That’s almost right. During the automatic chain calibration, after the chains have done the first move, put a bit of paint on the link at the top of the sprocket. Then finish up. Now when recalibration is needed, put the marked links back to the top of the sprockets and use Advanced/manual chain calibration. That tells the Arduino where your chains are, at the calibration position. Now use Return-to-center and you’re back in business.
If the sprockets need to be rotated to put a tooth at 12 o’clock, use the first step of the automatic chain calibration and then bail out before the step that runs the chain out, the manual calibration lets you move the chains instead of waiting for the motors to do it.
There is a reason to mark the chains at this step instead of at 0,0, this way the paint marks will be true tlregardless of any changes you make to your frame or sled. If you instead paint the link at 0,0 your sled won’t be in the same position if you change to one of the pantograph attachments or if you change the geometry of your frame. Through all the changes to my frame and sleds, the same painted links on my chains let me use the manual calibration so I can get back to cutting more quickly :smile:.

3 Likes

The ring performs WAY better all the way at the top of the sheet than the old attachment method (I promise I’ll have cheap kits available soon!).

The ring is better at the top for two reasons:

  1. The calibration is much simpler, and because of the way the math works out with the ring any calibration errors are smallest at the top whereas with the original chain attachment method any error is magnified at the top

  2. The ability of the chains to rotate out to the sides reduces the tension at the top. In essence the sled experiences the same forces as if it was maybe 6 inches lower, but because the tension is exponential as you approach the top it makes a big difference.

Having two distinct paths like that makes me think this is a case of a skipped tooth. I kinda thought we had that issue locked down with the addition of the extra chain idler that keeps more chain around the sprocket. I don’t think I’ve seen a chain skip in months, are we still seeing this happen often?

4 Likes

The only time I’ve had it happen when the chain attachment points are not in the same plane as the motor sprockets. It doesn’t need to be exact to prevent skipping, but when I first got my linkage set-up it would skip at the top and to the left and right of the bed where there is a lot of tension on the chain and when the attachment points are close to the motors.

1 Like

I’ll gladly drive to Portland to pick one up! I live in the Dalles.

2 Likes

Near the end of a long cutting session, I was trying to cut a piece near the top of the board and all of the sudden, the unit dropped maybe 1 cm? or so and continued to cut. I thought the linkage kit had snagged (my C bar gets a little twisted and rubs against the top and bottom of the two D bars) and set itself right. I had just cut out an identical piece below it and when you place them one atop the other, they actually look very, very similar. These are big L-shaped pieces (back and seat of a chair). However, the cut was lower on the board than it actually should have been. Below is a photo. The top track was the original cut and where it ends is where the drop occurred. The bottom track of the top piece should not have road over the top track of the bottom piece (there should have been a very small gap between them).

My questions are 1) what error should we expect from jumping a tooth and 2) how do we avoid it. I have a final frame built like the instructions say (though I added a 2x4x10 between the upper parts of the arms) and use a linkage kit.

The distance of error will depend on how many teeth jumped. Usually it’s only one link however it is conceivable that it could have jumped two. That does not look like a linkage snag to me, but I could be wrong. Making sure that your chain idler is in the right place could help with skipping teeth.

If you are experiencing too much friction where the linkages cross check first that your motors are on the same plane as your linkages (or close to it, doesn’t have to be perfect). This will help reduce linkage friction (if there is any) and it will help reduce tooth jumps. The other thing I would do is sand both sides of bar C (not the ends, just any part that passes through D). This will help to make C thinner and increase the tolerance between C and D.

2 Likes

The idler is placed similar to what’s in the pictures on the website. I don’t think I could put it any further “inward” to ensure provide more contact with the sprocket.

With regards to C, one edge rubs the bottom of the top D bar and the other edge rubs the bottom of the bottom D bar… so it looks like its twisted/tilted. I don’t think sanding would solve that. I wonder if I might need to shim one of the connection points to the sled to adjust its height (i.e., not all standoffs + washers are at the exact same height).

1 Like

Those bars are only under tension and the most force they see with a 20 pound sled is about 17 pounds (that’s with the sled at the top center where you have ~33 pounds of force on each chain). While not ideal, you could remove quite a bit of material from C without causing problems… If the plywood itself somehow warped and the rubbing is not due to how it is mounted I’d be happy to send you replacement parts. Shimming is a good idea as well.

Just let me know if you need anything from me!
-Logan

1 Like

About the only way the chain will jump (and a sudden drop of the unit does sound
like a chain jump) is if the chain isn’t feeding cleanly into the sprocket.

do you remember if it was moving towards or away from the motor when it jumped?
that will help you know which side of the sprocket was probably the problem.

On the sled side, if the chain is parallel to the workpiece, there is pretty
much nothing that will cause it to jump, but if you have it too close or too far
away from the sled, the chain will approach the sprocket at an angle and can
climb on the side as it’s wound in under heavy tension.

on the other side, the idler was added to help avoid this problem

can you show us a picture of your motors?

1 Like

It was moving towards the left motor (and away from the right motor) but it was also on a downward trajectory as well… about 45 degrees toward the bottom left corner. The cut was not near the left edge of the board… approximately 20 inches left from center. I wanted to do the cuts in the middle for better accuracy.

I measured the offset of the motors to the chain mounting point. The chain mounting point is at 3.125 inches from the workpiece and the motor is 2.75 inches… So 0.375 inches. The workpiece is 3/4-inch plywood and would be worst case. I don’t expect to cut thicker than that.

Below are pictures of the motor mounts. Maybe I could move the motors back a little, but the idler really can’t move forward any.

I also checked the mounting heights of the linkage and I think one is maybe (maybe) 1/8-inch too low… not sure and will measure better when I have better light in the shed.

2 Likes

If you can try doing this multiple times and see if it reliably skips, you can try getting a movie of the motor as it skips to see exactly what happens (you can have the Router out from the wood so that you aren’t cutting anything, or skip having a bit in the router entirely)

I’ll give it a shot sometime this week. It did “reliably” cut 6 passes in a row after the chain slipped. Everything looks ok with the motor mounts? The picture was with the sled in center position.

It’s also conceivable that the sled got hung up on on edge (I need to round out my sled’s edge more) and got off kilter and that led to the chain slip. That blue painters tape is what I use to keep edges down if needed.

1 Like