What might cause this?

[Mike Wenman]

Still working on my homebuilt system..it has been one problem after another...bad motors, bad power supply, racking of the gantry, oval holes when trying to drill and the list goes on.

Anyhow, I think I finally got everything tightened up and clamped a 3/4" x 6" x 14" piece of mdf on the table and launched Mach3. Thought I would try something simple for a test cut and ran the Text wizard and just did a simple 2" high bold letters gcode that said TEST CUT". Nothing fancy for sure. put a 1/4" 60deg V-carve bit in the router, set my 0 ref point for X, Y, and Z. turned on the router and dialed in a 16000 rpm and started running g-code on mach 3.

Ok..first the router moved to start the T, lowered Z down to -0.125 where I had defined it..cut the T, raised up to .5..moved to E, lowered to -0.125 and did that cut. after raising up to .5 and moving to the S position, instead of Z lowering to -0.125 as the code showed, it instead plunged straight down through the mdf and into the spoilboard.

I did emergency stop..raised Z up...and re-zeroed. checked everything and with it looking good, started up again. exact same thing happend..on the 3rd letter, Z axis plunged full hilt.

so, any ideas what would be causing this anomoly?

Mike

[Guy Mathews]

Still working on my homebuilt system..it has been one problem after another...bad motors, bad power supply, racking of the gantry, oval holes when trying to drill and the list goes on.

Anyhow, I think I finally got everything tightened up and clamped a 3/4" x 6" x 14" piece of mdf on the table and launched Mach3. Thought I would try something simple for a test cut and ran the Text wizard and just did a simple 2" high bold letters gcode that said TEST CUT". Nothing fancy for sure. put a 1/4" 60deg V-carve bit in the router, set my 0 ref point for X, Y, and Z. turned on the router and dialed in a 16000 rpm and started running g-code on mach 3.

Ok..first the router moved to start the T, lowered Z down to -0.125 where I had defined it..cut the T, raised up to .5..moved to E, lowered to -0.125 and did that cut. after raising up to .5 and moving to the S position, instead of Z lowering to -0.125 as the code showed, it instead plunged straight down through the mdf and into the spoilboard.

I did emergency stop..raised Z up...and re-zeroed. checked everything and with it looking good, started up again. exact same thing happend..on the 3rd letter, Z axis plunged full hilt.

so, any ideas what would be causing this anomoly?

Mike

If it happens in the same place each time it is more then likely your vectors. A stray sliced curve sitting below Z for example. Look at your code and find the third letter then look at the movements.

This is a good reason for always working in the positve quadrant until you get used to the machine and the software. With your text elevated at the height of your material there should be no -'s in the code when posted unless you set cut depth deeper then material thickness.

If you would like to email me the code I can read it for you to make sure that the code is up to snuff.

Guy

[Joseph B. Chritz]

Hi Mike,

You’re going to have dig into this a little deeper. The program code said to go Z minus .125, the material was 3/4 thick, and it went through the material, so it went at least 5/8 deeper than it was supposed to. One possibility is that when the Z axis rapids up the steppers can’t keep up and skip steps. If it only raised part of the .5 up move the second letter would be deeper than the first and the third could be even worse.
Try raising the Z axis above the material and see what it wants to do after the third letter. Keep an eye on the Z axis read out to be sure it's correct and see if the error is repeatable, it could be a ploblem with electrical noise or something mechanical.
This should give you a clue where to look for a fix.

[Michael Schwartz]

check that your z origin point is set to the top of the material, if your zeroing the Z axis to the top of the workpiece.

[AL Ursich]

When I was working at the Sony Picture Tube Plant in San Diego we would use a little trick to help troubleshooting the Fanuc Robots. We would set up a video camera to watch the function close up. When you run the Text you are trying to look at the display and the Z at the same time and sometimes missing some of the details of the error. Let it be another set of eyes... Plus shooting the display then a second time shooting the Z and sending the Video to Guy to look at could be helpful.

This could simply be a case of the Z not balanced. When it tries to retract the motor tries but does not back up far enough and down again on the first letter, part way up, down on the second letter, part way up, down into the table on the 3rd letter......

Guy has a GREAT Shop !!!!

Good Luck,

AL

[Bob Savage]

Hi Mike,

You’re going to have dig into this a little deeper. The program code said to go Z minus .125, the material was 3/4 thick, and it went through the material, so it went at least 5/8 deeper than it was supposed to. One possibility is that when the Z axis rapids up the steppers can’t keep up and skip steps. If it only raised part of the .5 up move the second letter would be deeper than the first and the third could be even worse.

Take a good look at this one. I had an issue like this where my Z acceleration was set to high and the Z0 point of reference was being lost. It wasn't by 5/8", nonetheless, I was digging about .30" deeper than I was supposed to be, ruining the part.

[Gerry Grzadzinski]

When it tries to go to deep, you need to see if the Z DRO shows the correct location. If it's going deeper than the DRO says, then you're probably losing steps.

[Bryan Cowan]

Take a good look at this one. I had an issue like this where my Z acceleration was set to high and the Z0 point of reference was being lost. It wasn't by 5/8", nonetheless, I was digging about .30" deeper than I was supposed to be, ruining the part.

I, too, had this problem. Wound up causing damage to the Z-drive on the mother board and a few broken bits.

[James Leonard]

Hi,

You can visually check your toolpaths with NCPlot free. The isometric view can be rotated to show the tool depth. This is a totally free tool, I recommend it to all of my customers.

http://www.ncplot.com/ncplotfree/ncplotfree.htm

Also, check the calibration of your Z axis. Put a dial indicator in the spindle and check if a commanded .250 move is really .250. Then do a repeated series of Z axis rapid UP then cutting feed down to see where your Z axis starts losing steps. Then set a maximum Z rapid of 20% less than that.

-James

[Mike Wenman]

Was definitely missing or even gaining steps, but while doing some looking found 2 things that were causing problems. First, the 2 bolts I had to hold the anti-backlash nut to the mounting point had come loose and I had approx 1/8 - 1/4 inch play depending on whether the Z was being raised or lowered. Secondly, and I ran across this simply by accident when reading another forum, another person was having a similar problem where his anti-backlash nut was actually binding on the lead screw. Watching mine and listening while jogging Z, I had this too. His solution was good old WD40. once I did that and ran it a jogged a few cycles, my Z is now running nice and slick. was even able to re-adjust the velocity in Motor Tuning in Mach3

FINALLY was able to make a successful cut...a few hiccups, but that was simple errors on my part.

Thanks for help and info.

Mik,e

[Gerry Grzadzinski]

A light oil will last longer than WD40. I use air tool oil on my screws, and it works great.

[james mcgrew]

we use 3n1 oil