Unwanted Beveled Edges or Stair Stepping

[liebhaber]

I've had 2 carves where this has become a problem. Please see the pictures below.

On the sign, the left half of the cutout is 90 degrees and perfect, the right half of the cutout is stair stepped with a beveled edge. I have included the mpc.
On the letters, all were good except for the S which had the same effect (sorry no pictures but same exact problem).

The sign was 1-inch whitewood, the letters 1/2inch baltic plywood. All set to .125 max pass during the cut.
587715877558776587775877858772

Belts are good, head pressure is good, roller is good with o-rings present, etc... The 1/8 cutting bits were different too, the sign had a QC bit and the letters had a brand new right out of the box Carvetight bit. So it did it with both the QC and carvetight chucks.

Any ideas? Am I missing something simple? It's only the cutouts, patterns are perfect.

edit: oops, I guess I attached a few other pics too :)

[bergerud]

First question: Did you stay under the rollers during the whole operation?

[Digitalwoodshop]

This is a case of Pilot Error as in the Board has lost contact with the Brass Roller in length of the board. Tips and Tricks on the LHR page will give you lots of info on this and other great stuff to know like the 7 inch rule of wood and scaling.

The Masking Tape Trick can help too... Masking tape on the board edge where the Brass Roller runs and makes a rack and pinion out of the tape. PLUS a rolled under belt at the brass roller side can raise the board off the brass roller too.

As you board came out from under one roller it tipped up and the board lost contact with the brass roller and the computer will keep moving the board until the brass roller is happy. This is what causes stair stepping on one end and the board being cut off in a full depth cut on the other... That can SNAP that $40.00 bit easily...

I set up my projects in Designer with the 4 inch blank area on the right side of designer and that lets the board stay under the roller all the time without selecting to stay under rollers. I use a longer board than required to keep the back under the roller and cut the project off when done.

I even tape blank wood on the end of my blanks to make more frugal use of wood.

I also use Place on END or Corner ALL the time... So I know my project starts 4 inches from the right side of designer.

Place on CENTER measures the board TWICE and is a waste of Servo Time.:roll:

Good Luck,

AL

[bergerud]

To add to the list: I have had stair stepping exactly as shown caused by the board slipping on the sandpaper belts a little each time as the board had to push up the rear roller. The board was always in contact with the brass roller and was always under the front roller but would come out from under the rear roller. This was during a mult pass cut out and each pass was shifted the same small amount. Make sure you stay under both rollers during cut outs.

[liebhaber]

Ok, I had 5" extra on either side for the letters, the board was under rollers the whole time. Masking tape was on the board, 2 layers. On the sign, there was about 6" extra on either side. No scaling.

[DickB]

Warped boards can also cause a problem. I've had problems with boards that were cupped just a bit if I oriented the board with the protruding part down. Flipping the board over got it to track.

Another possibly is dust in the optical encoder of the tracking sensor. Not too hard to remove it and check. First time I had this issue I was shocked at how much dust was in mine. I don't know how it worked at all.

[bergerud]

Well, somehow there is an x tracking problem. Check everything! Head pressure. Head level. X gears. Excessive friction with the sandpaper rollers. Maybe Al has more.

[henry1]

I had that problem and it was the incoder on the Z truck ,incoder dirty or slip on the shaft

[liebhaber]

Ok I will try to clean everything out again and see what happens.

I just did a smaller carve and it came out perfect, seems to only be the larger cutouts.

[Digitalwoodshop]

Another WAG... Wild AL Guess.... I mentioned in the other post... A Rolled under belt... Cut Motor one at the Brass Roller... When the belt splice is rolled under double thick it raises the board off the brass roller.... I tear a strip off the belt all the time... As the belt creeps toward the brass roller when I do allot of 6 inch board wearing half of the belt causing the creep.

AL

[liebhaber]

Ok, cleaned everything out, belts are trimmed from previous maintenance, head pressure 77#... Cleaned board sensor, rollers, brass roller, etc... Board sensor over 100 on this board, nice light pine.

Ran another project with 2 large (12-13") and 3 small (4" square) cutouts. All the small cutouts were perfect. All had 90 degree edges on all sides, no problem.

Again both large cutouts had the "stair step" bevel effect. One cutout was placed with 8" on one side and over a foot on the other side, and the other was 4" from the edge as I used place on corner per Al. Under rollers the whole time. Carving was spot on perfect, no issues. Just the cut out keeps being a problem, same 1/8" or so bevel on the left side of every cut out (as looking from the keypad). The right side is spot on perfect. If I cut out a large cross type design, all left edges have the bevel. On a circle, the bevel starts at the midpoint and goes along the left side. (I've tried variations of these from other signs).

To add to this, my board measurement is a consistent .653 or so off on the short side (I can't remember the decimals). Would this cause the issue? None of the carvings have any hard anchors in the mpc. If I try to bring the measurement back in calibration I get an out of range error. I just live with the length issue.

Attached cutout MPC (there was carving on it that I deleted), and pics.

Any more ideas?

[DickB]

Did you disassemble the tracking roller to clean the optical encoder?

[liebhaber]

I did not, just cleaned the outside. That's just another items on the to-do list.

I'm just baffled as to why the issue seems to occur only with cutouts greater than about 6-7" in length. All of the small ones are just fine.

[DickB]

If it were me, that would be on the top of my list.

If it is a dirty encoder, it may be marginal. The nature of the problem might be that one of the slots is partially blocked, so sometimes it is detected and sometimes not. So over a short distance odds are it works or mostly works, but over a long distance there is more opportunity to misread. Also, on carving the board moves only in one direction, and if the slot is misread the board will move only a fraction farther than it should and you probably can't tell. But on multiple passes when cutting it of course moves back and forth and the problem is much more noticeable.

[DickB]

Another thing to check is warped boards. Not sure if I'm seeing it in your photo, but it looks like your board may have a slight cup to it. This could be enough to loose traction with the roller.

58889

[Digitalwoodshop]

Since this is the X Axis that is the problem....

A few more thoughts.....

Do you have Dust Collection? Is sawdust building up on the board?

Is the 1/8 inch bit very old? Can you try a new bit... The TIP may be dull since that is doing the cutting.

Have you ever removed the X Gear Box plastic Cover and inspected all the teeth? Could be one cracked and doing a Raster Carve it is OK but doing a Cut Path is under more X direction stress and slips one tooth...????

The 2 screws under the machine going into the 2 smooth vertical rails on the keyboard side of the machine. Are they Tight...?

Since the 1/8th inch Bit is in a Bit Holder.... IF you grab the tip and tug in 4 directions, ANY movement OR any BB Marks in the Bit holder?


AL

[liebhaber]

I will clean out the z encoder tomorrow when work permits. The warping you see in the photo is looking through the cutout to the board below sitting bowed on a sawhorse.

The 1/8" bit is brand new as of the project right before this one. The teeth on the x-axis are all intact, I check right before this carve when cleaning the machine. I have the ringneck DC system and the board is generally sawdust free. The chuck is carvetight and the bit is very securely seated. No movement whatsoever.

I'm thinking it has something more to do with the sensors/encoders since the board measurement is so far off (2" if on factory default) and only the longer cutouts with longer time for encoder/measurement error are affected.

[DickB]

X encoder.

[liebhaber]

Just to make sure i got it right when cleaning, the X-encoder is actually in the Z-motor. Is this correct? If not can you please point me in the right direction. Tried a search but nothing specific comes up.
Thanks

[DickB]

Take a look here: http://store.carvewright.com/product.php?productid=16207&cat=254&page=1
The encoder is in this part, the board tracking sensor ("the brass roller"). Instructions on how to remove it are to the right of the photo. The optical encoder is to the right in the photo. You need to remove the cover and blow it out - gently - with compressed air (this step is not in the instructions).

[badbert]

If your tracking sensor has an O-ring on the brass roller, try carefully removing it, and checking it again... If it does not have an 0-ring try purchasing the correct one from LHR and try that. The diameter of that sensor "is" the calibration. By adding or removing the o-ring it effectively changes the diameter of the sensor. Similar to changing tire size on a car. The speedometer has to be re-calibrated to compensate for the bigger/smaller tires.

[liebhaber]

I cleaned the x encoder out, the z encoder, and everything else. I literally tore the machine down to a bare frame, and removed every speck of dust in every compartment and sensor I could open.

After that I did a board measurement test and on a board 13 inches long the measurement was perfect. When I put in a 36 in board the measurement was almost 2 in short.

I have the o-ring. I will see what removing it does.

[bergerud]

If your tracking sensor has an O-ring on the brass roller, try carefully removing it, and checking it again... If it does not have an 0-ring try purchasing the correct one from LHR and try that. The diameter of that sensor "is" the calibration. By adding or removing the o-ring it effectively changes the diameter of the sensor. Similar to changing tire size on a car. The speedometer has to be re-calibrated to compensate for the bigger/smaller tires.

I do not think that is true Bert. The O ring compresses to the diameter of the brass roller and does not change the effective radius. It would if the O ring was too fat. Wrong O ring. To convice yourself that this is true, you only need see the brass roller tracks on the under side of the board.

Knowing this begs the question: Just what is the O ring for?

liebhaber: Bert may be onto something. If you have too fat an O ring, the machine would measure the board short.

[chief2007]

O ring is for added traction to help ensure the brass roller is tracking properly, especially on the harder woods that the brass roller may not grab into, thus the big reason Al says "masking tape" and I totally agree with that statement.

[liebhaber]

With the o ring I get consistent roller marks on the masking tape all along the board.

Should I just replace the roller? It seems like all short measurements are good but once it has to measure a longer piece it gets confused. I tried calibration but its so far off I can only get within about 5/8" of actual size.

Every time it measures a longer board the measurement is .05-.2 off from the last measurement. So the same 36" board may get 34.82 34.91 35.02 or 34.62, just depends on how the machine feels.

Length measurement only, width is good.

[lynnfrwd]

Havent been following thread and didnt read earlier posts, but check heights of your outfeed trays and any additional outfeed roller support.


Sent from my iPhone using Tapatalk

[fwharris]

What belts do you have? Sandpaper or rubber...

[bergerud]

What is your default x calibration number? Is it around 890?

[badbert]

My refurbished "A" came without an O-ring. Trying to improve traction, I installed one. After experiencing many calibration (wrong length issues) I removed. After removing it, the Calibration is almost exact. I always use masking tape and I have never had a problem. Everything from Oak, walnut, Maple, to MDF, no tracking issues. My 48" sled measures out at 47.974.

[bergerud]

My refurbished "A" came without an O-ring. Trying to improve traction, I installed one. After experiencing many calibration (wrong length issues) I removed. After removing it, the Calibration is almost exact. I always use masking tape and I have never had a problem. Everything from Oak, walnut, Maple, to MDF, no tracking issues. My 48" sled measures out at 47.974.

I remember that. The O ring you were using was too fat. It did not compress to the diameter of the brass roller. The effective diameter of the roller was too large (and probably inconsistent). That, however, was not the genuine O ring was it? I and I think LHR does not really have a good explanation for the purpose of the O ring.

I have another bone to pick. Everyone keeps saying to put masking tape under the board edge to give the roller traction. Well, the roller rolls with very little friction and all of my boards always have roller tread marks. There is no way that roller is going to slip on a straight board. The problem is that the board slips on the belts while, for example, trying to push up a roller and the machine gives a tracking error. This is not because the brass roller slipped but because the x encoder and the brass roller disagree. The board slipped. It is the underside of the board which needs the masking tape for better traction on the belts. If one puts masking tape on the edge of the board, the problem is fixed and it is easy to think it was the brass roller which had caused the problem. But I do not think so. The tape on the edge has stopped the board from slipping. The roller never slipped. I challenge you guys to test this. The next time you get a tracking error, put a couple of strips of tape in the middle of the board instead of the edge.

[badbert]

I remember that. The O ring you were using was too fat. It did not compress to the diameter of the brass roller. The effective diameter of the roller was too large (and probably inconsistent). That, however, was not the genuine O ring was it? I and I think LHR does not really have a good explanation for the purpose of the O ring.

Correct! :) I use masking tape on the board for traction with the sandpaper. But I do run it along the edge to improve the surface of the edge.

[liebhaber]

Sandpaper belts, in good condition and not rolled up. Carefully inspected when I removed them for a deep cleaning of the machine. Default x is around 890, but I'm not around my machine at the moment.

The rubber o ring does seem a little large. This is a used a machine (craftsman) but only had about 15 hours on the motor when purchased. When I opened up both the x and z encoders there was extremely little to no dust and everything looked brand new.

Perhaps this is an old o ring that is causing problems?

[bergerud]

The main question is: Do you see brass roller tracks on the board? If you do then I think you have a defective board sensor. Maybe dust, maybe cracked.

If you do not see those tracks then somehow the roller is not contacting the board properly. Take off the O ring and try it. See if the brass roller is up proud of the plane of the belts by about 1/8". (Does the roller push against the board?) Does the board get lifted up off of the roller during travel.(Lynnfrwd's suggestion of high out trays, makes sense if longer boards are worse.)

[liebhaber]

Ok, to close out the thread:

I removed the o-ring and ran 3 test measurements on the 36" board. They all came out 35.92 on default calibration (I reset it to factory just to ensure it wasn't throwing off the measurement).

Close enough for me. I guess the o-ring was just too big.

[bergerud]

It may also mean that the brass roller is not being pushed hard enough against the board. There is a piece of rubber under the tracking sensor plate which pushes it up. Make sure the roller is being pushed hard against the board.

[DickB]

+1. That's an easy part to miss.

[liebhaber]

The roller is actually fairly rigid. So much so that I have to put positive pressure on most boards to get them to sit flush when putting them in the machine and aligning them. Sometimes the roller pushes the board up and over the keypad side rail while I am adjusting the sliding plate.

edit: It has the appropriate "cushion" to it, it's not stuck in the up position.

[badbert]

Ok, to close out the thread:

I removed the o-ring and ran 3 test measurements on the 36" board. They all came out 35.92 on default calibration (I reset it to factory just to ensure it wasn't throwing off the measurement).

Close enough for me. I guess the o-ring was just too big.

Good news!

[Digitalwoodshop]

Good job !!!!

And for later users reading this thread... Sometimes the Sand Paper Belt under the Cut Motor drifts toward the brass roller and rolls under. This is caused in my machine when I cut many 6 inch wide boards and half of the belt gets worn more than the bit plate half. As the belt on the keypad side wears, I believe it gets slightly bigger and the taper causes the belt to drift like our belt sanders when we tweak the adjustment. So when the belt is rolled under the DOUBLE Thickness can raise the board OFF the BRASS ROLLER.... Especially when the SPICE hits....

That is why I rip the strip off my brass roller side as seen in the earlier pictures. When I was cutting mostly 11.25 inch wide boards I had NO Drift Problems... It was only after the Craft Show stuff Died and I was cutting more 6 inch wide boards than anything else.

And that Rolled Under Belt is VERY HARD to see... Only after you rip the belt is it seen. :razz:


At least that is my Theory....

AL

[badbert]

It may also mean that the brass roller is not being pushed hard enough against the board. There is a piece of rubber under the tracking sensor plate which pushes it up. Make sure the roller is being pushed hard against the board.

I have never had track marks on my board. I can push my sensor down fairly easy. Maybe I am missing something...

[liebhaber]

So there I was...

Ready to carve, and I got a Y-Belt slip when the truck went to the bit plate. Went to go tighten the Y-Belt tensioner (per another thread) and as soon as I even wiggled the screw it came completely out of the tensioning assembly. The threads in the hole are completely stripped, and I've never even touched that part before! It won't even stay in at all (with/without tension).

Oh well, down AGAIN until I get that part. I'm starting to think the 15 hour machine I was sold from a member of a another forum was actually a 1500 hour machine with a new computer. Almost every screw hole on my machine is stripped. Half of the bottom screws won't stay in.

[bergerud]

I have never had track marks on my board. I can push my sensor down fairly easy. Maybe I am missing something...

Check out the "Board tracking sensor damper spring" on page 3.

http://www.carvewright.com/assets/service/Service_instructions/CarveWright_service_replace_board_tracking_sensor. pdf

[Digitalwoodshop]

Normally when you get that warning it is because the Bearings in the Y Gear Box are SHOT. The big plastic Gear is leaning toward the center of the machine and making the belt loose and the plastic gear is rubbing on the U shaped frame that the Y Gear Box Sits in....

They are the Two 5 mm x 11 mm x 5 mm or a Standard 685ZZ Bearing. ZZ is metal Shields.


OR one of more of the Y Roller Bearings are frozen. Easier to see when you remove the Y Motor and roll the head.


IF the big Y Gear Shaft is WORN Thin they sell them at LHR. See my old invoice.

Good Luck,

AL

[liebhaber]

The tensioning screw literally fell out of the hole, no force required. The hole is so stripped it won't hold the screw at all. I'm pretty sure it's the screw that wasn't holding the tension properly if it was that stripped/on the verge of failure.

I have also done a pretty extensive forum search and a lot of the Y axis issues related to the truck clicking or "slipping" when going to extend the bit plate are either wires in the way (not in this case) or the tension was off and need to be re-tightened.

[chief2007]

The tensioning screw literally fell out of the hole, no force required. The hole is so stripped it won't hold the screw at all. I'm pretty sure it's the screw that wasn't holding the tension properly if it was that stripped/on the verge of failure.

I have also done a pretty extensive forum search and a lot of the Y axis issues related to the truck clicking or "slipping" when going to extend the bit plate are either wires in the way (not in this case) or the tension was off and need to be re-tightened.

Think you have the problem, your talking the screw on the top of the tensioner correct?

Might try a self tapping screw as a replacement to get it fixed.

[henry1]

you would need to change the tentioning screw and and the part that screw in (Y-Belt Tensioner Assembly MWM A2010

[liebhaber]

Just ordered the assembly. It was the actual allen head screw with the spring that no longer will stay in the hole in the tensioner assembly.

I didn't see a link to order the tensioning screw, either it comes with it or else the one I have will work I hope.

[chief2007]

Just ordered the assembly. It was the actual allen head screw with the spring that no longer will stay in the hole in the tensioner assembly.

I didn't see a link to order the tensioning screw, either it comes with it or else the one I have will work I hope.

If it is missing the Allen head screw and spring, pm me, I have a spare.

[liebhaber]

Installed the new Y belt tensioner assembly and it works like a charm!