A little ticked off

[DocWheeler]

I tried to make a name-plate using the two-mpc method and had a problem with the first mpc not carving in the center of the sled/board, plus the first pass was too long. I went ahead and filled/glued the routed area with the second wood type (maple) and ran the second mpc to see what would happen - picture #1.

I then re-created two more mpcs on another card and found that the first of the two mpcs carved off-center again - ruining another piece of purple hart! (second picture).

Picture three is a closeup showing the first pass going too far to the right (actually it wasn't too far to the right, I just thought it was at this time) by about a quarter inch and the carve region being off-center of the board and sled.
(The pencil is in the center of the board and sled)

Is there a known bug in the firmware? Does anyone know what the cause is?
Problem one - first pass too long.
Problem two - wrong location.
Attached is the last initial mpc that carved as shown in pictures two and three.

[lynnfrwd]

Did you use tape?

[DocWheeler]

Lynn,

The problems are in the "Y" direction, but yes I have tape on the bottom of the sled.
The second mpc has the same board size as the first mpc and it carves fine.

[alan.galbraith]

Shouldn't the sled have solid wood around the carve area so that the sensors read the sled properly?
I'm probably wrong but it's the way I do it.

[jab73180]

Hey doc. This is interesting cuz I am having troubles like this. I was thinking it was a bad mpc and/or memory card.

[SteveEJ]

Hello Doc.. Nice to see ya..
My theory: The contrast change in colors between the darker Purple Heart and the lighter wood in the sled may be causing a sensor 'Glitch'. If I remember right, sensor checks compare differences between light and dark surfaces. Did you try the same carving on a cheap piece of wood that is light like the sled? or Darkening the sled with some dark paint that will contrast match the Purple heart?

Just some ideas.. Let me know what you find!

[DocWheeler]

Lynn,

If you were talking about the "Y" direction for tape, sorry for my not understanding.
Measuring the width is a controversial subject. I have found that if the sled side-rails are 3/4" wide or less there is no need for tape across the sled (or rails if you are just using them without a "sled" bottom).

I will try it with tape although it is very confusing that the mpc with the lettering worked correctly. I watched it measure the width in both places and it seemed to do it correctly.

Steve, I'll test that!

Notes about the pictured sled:
It is 20" X 11 3/8" with a 3/4" bottom (because that was the size of the piece of plywood I found).
The rails are 3/4" X 3/4" X 20" long.
The end fillers are just 1 1/2" in the "X" direction since that is all that is necessary for the "X" measurements and the roller to have contact and the sensor to see for the first "Y" measurement if you want to use it. (note that it is not necessary for the sensor to see anything between the first "Y" edge and the second 3/4" rail, it will go to the back of the machine and "re-find" the outside of the second rail - the software has worked that was since version 1.119?).
By not using 3-4" fillers on the sled, I can use bigger pieces of wood in the sled. Note that there are a bunch of screw-holes in the bottom board - used it for a lot of carvings including the purple heart tea box I made (four sides and two tops - never used tape in the "Y" direction).

Attached are the first two mpcs that I used for the first failed project, they are a little smaller than the mpcs for the second attempt (different pieces of wood found).
I also attached the fourth mpc that was never carved so that all four mpcs are posted.

Note that the first mpc carved in the wrong place and the first pass was too long, just like the mpc I posted above. The second mpc positioned itself correctly - same board on the same sled using the same commands.

[DocWheeler]

First let me say that I carved the pattern #3 without a glitch.
Second, I would like to add that I gave false info on the sled, the leading edge is 2" rather than 1.5" in the X direction.
Third, I would have sworn that the machine measured the width after touching the bit-plate.

So, what did I do to make it work? I put tape across the sled at the point where it does the bit-to-plate which didn't do diddly since it didn't actually measure there like I thought it had before.
After it carved correctly I set it in again and it did it correctly without the tape!

I think the machine has contracted "some-timers".

Incidentally, my claim that the first pass went too far to the right(back of the machine) was incorrect. It started in the right place and simply went too far to the left thus moving the region to the left.

Sorry to have taken your time with this since it really is not resolved.

To clarify the width measuring that I use on this sled, I set the sled to where the roller just catches the leading edge of the sled. There is light colored wood there for the sensor to see all the way across the sled. However, it works as well for me to have it not see the light colored wood at all be inserting the sled a little further so the sensor is over the opening in the sled (admittedly, I haven't done this with the latest firmware).

I'll post the finished project here when it is completed. Hopefully the area carved incorrectly won't show up too much, I packed it with glue and sawdust hoping that it carves OK.

[SteveEJ]

Gremlins.. I hate them. Glad you got it working Doc! Looking forward to a picture of the final product!

Cheers..

[Digitalwoodshop]

Ken,

Back in the shop today after a Field Trip Yesterday updating the Gravity Depot Life Member donor board.... A Fun day visiting relatives on my old hometown.

I agree that problems detecting the actual sides of the sled could be a problem.... Even detecting the inside edge of the sled could have been a possibility and if you have any Restraints set like yellow numbers or center on board..... That can confuse the issue if the sled is measured wrong in Width.

So then since the Y Encoder is SO Important in this process and I have posted my picture of MY Y Encoder with Sawdust inside. Plus there is always that black dust inside the Encoder from the magnets. Clean the encoder disk and slot.

Then there is the question on when is the last time Y Gear Box Bearings were changed..... I have had the Y to skip teeth in the process of cutting... Over 450 Cut Motor Hours and the Y Bearings should be replaced and the shaft and big gear if worn... If you got a CLICK at the Bit Plate then the Gear Shaft is WORN and Bearings are already BAD.... 5 x 11 x 5 MM 2 each.

I recommend a strip of masking tape over the open area of the sled and remove it before actual cutting.

Let us know....

AL

[want2learn]

I had a similar problem and it turned out there was a short happening with the small four wire harness in the board sensor in the z truck, I had a wire pinched and it was shorting out , making my bit plunge into the board

[liquidguitars]

I recommend a strip of masking tape over the open area of the sled and remove it before actual cutting.


"For the love of Roddenberry" the Idea that a open sled needs tape on the top or the sidewall needs to be .75" or the CW will not measure correctly shows we are not using the old load under the 4" tail trick. All you need to do is to build a sled with the correct 4" tails and load the sled tail under the Board sensor.

Doc, Their was a firmware error when cutting the B side but it only showed for me if you used " place on corner" it should be fixed in the lastest build.

44911

[DocWheeler]

Thank you all for hanging in-there with me on this.

As I hope that I stated clearly, the problem was that the start of the carve was in the correct place (so the width measured correctly), the problem was that the first pass went too far to the top (keyboard-side) to cut out the region.
The three or four inch waste of "X" movement or tape was really not a part of the problem.
I will check the encoder but doubt I will find a problem since the region(S) carved smoothly.

This machine kept me from making some Christmas presents by giving me sensor errors consistently for a few days until I gave up and waited two months before trying again, at which point it worked perfectly!

[DocWheeler]

I hurried up to finish this yesterday since I was going to a meeting near where Michael's office was. It was late when it was done so I had to use a flash which caused unwanted effects.

The first photo is of the back where you can see the area that the machine first attempted to carve the region in the wrong place. (I filled it with sawdust and glue before carving the last mpc)

The second shot is the front, the third is the bottom, and the last one is a closeup showing the boo-boo.