Depth of cut consistently off

[dlcw]

I've encountered a Z-axis problem that I can't seem to make go away. I'm running a Buddy BT32 using a PC router. The spoiler board has been resurfaced. All the electronics have been shut down, unplugged and restarted. And the problem remains.

I'm cutting a 3D model from Vector Art. I've setup the model to be, at most .46" at its deepest point in to the material. In Partworks 3D I established the top surface of the materials as the z-zero location. I've gone through the rough-cut SBP file and seen the Z-axis commands and they are correct.

Here comes the part that's driving me crazy. I've run the same SBP file 3 times on an MDF sheet (don't want to ruin perfectly good wood) and the z-axis actual depths are different (but consistent) from the commands in the SBP file. All electronics were turned off and unplugged between each test. X,Y,Z axis points were set to the center of the material at the surface of the material.

The table below has the SBP file depth command, the actual cut depth, as measured by dial calipers, and the depth difference.

SBP file Actual depth Delta
-0.125 -0.15625" -0.03125"
-0.25 -0.3125" -0.0625"
-0.375 -0.46875" -0.09375"

It seems that the deeper the programmed cut, the greater the delta in depth of cut.

I reset the z-zero to 0 at the surface of the material I'm cutting, in this case the MDF sample board. My process is to use the keyboard control to lower the router until the tip of the bit just touches the surface of the material. Then I use the menu option to set the z-zero. I get confirmation of this action by the fact that the Z setting on the ShopBot control program resets to 0.

If any one has suggestions on eliminating this problem I would greatly appreciate it. I can send SBP files if needed.

Thank you,

Don Thomson
Diamond Lake Custom Woodworks, LLC
www.diamondlakewoodworks.com (http://www.diamondlakewoodworks.com)

[erik_f]

I know at some point the updated the post processors. Maybe you want to check which version you have for your software.

[erik_f]

VERSION 3.5.10
-Torch Height Control enabled for V201 and later Controllers
-Minor fix: On a few alpha tools with reported variability in electrical service voltage we found occasional lags in the driver's switch from Move to Jog resolution. A slight change was made to the timing of the resolution signal to eliminate any possibility for this error.
-Fixed a positioning problem that showed up infrequently in 3D contouring.

[richards]

Your actual cut depth is 1.25X the commanded depth. Have you checked your unit values (VU command) to see if the Z-axis is correct? Try dividing the Unit Value for the Z-axis by 1.25 and see what happens.

[myxpykalix]

I know you said all electronics turned off and unplugged between tests however, not knowing how a buddy is wired let me ask this.
When i was having a "hangup" problem where i couldn't get it to unhang itself after multiple reboots of control box and computer I went over and cut the breaker off dedicated to the bot and that erased the settings that are somehow kept in some processor and I wonder if you have completely cut it off from a power source so as to erase those settings. Possibly that might be your problem?

[dana_swift]

Don- is your buddy a standard or alpha?

D

[dlcw]

Dana,

I have the Standard model.

[dlcw]

Jack,

After everything has been properly turned off I remove the power from the power strip that everything is plugged into. Basically I pull the plug out of the wall.

Don

[dubliner]

Are your Z V wheels moving freely. Disengage the motor pinion & move the Z up and down, any binding spots?

[dlcw]

Mike,

I checked the unit value for the Z axis and it is 2291.8310. When divided by 1.25 it equals 1833.4648. The Base coordinate offset for the Z axis is -3.595.

Don

[dlcw]

Neville,

Yes, the Z V wheels are moving freely. The router moves very smoothly up and down without any binding spots or lurches from binding.

Don

[richards]

Don,

I've only seen the PRS Buddy at the Maker Faire in California. It is an impressive machine.

Let's go over some Unit Values. My understanding is that the standard model uses Gecko stepper drivers that require 2,000 steps per revolution.

A 20-tooth gear has a pitch diameter of 1-inch. It travels 1 X 3.14159 inches per revolution, or 636.620 steps per inch. With a 3.6:1 gearbox, it needs a unit value of 636.620 X 3.6 or 2291.83.

A 25-tood gear has a pitch diameter of 1.25-inches. It travels 1.25 X 3.14159 inches per revolution, or 509.296 steps per inch. With a 3.6:1 gearbox, it needs a unit value of 509.296 X 3.6 or 1,833.465.

Count the teeth on your Z-axis's spur gear. If might be 25-teeth.

At any rate, try changing the Unit Value for the Z-axis to 1833.4648 and see what happens. If you get correct distances, then that is the correct number for your machine.

[dlcw]

Mike,

You nailed it. It appears that the Z-axis gear is exactly the same as the X and Y gears at 25 teeth. I changed the Z unit value to 1833.465 and the cut depths are exactly on the money.

I guess I shouldn't have trusted that the Bot was ready to run from the factory as advertised. A little due diligence on my part was in order.

I guess that as a word of advice for new ShopBot buyers, check all the settings before going through the headaches I've had to endure. It will save you a lot of time, patience and MDF....

Thank you to all for the suggestions on getting this problem worked out.

[erik_f]

Its possible to load the wrong set of unit values. Or its possible they put the wrong pinion on your Z. Glad you got it worked out. Owning a CNC is like owning a classic car...it seems there are times when you really get to enjoy it...but there are also times when you get to learn the joy of trouble shooting.