[DWK]

I have been dealing with all kinds of depth of cut inconsistencies since i got this machine up and running. I bought it at an auction 5 years ago and it has sat at my warehouse unused for all that time. It is a 2003 Cosmec Conquest that has been stripped down of all of the "over the top" safety guards and other crap that makes it hard to work on the machine, but once i get dialed in i will put some of that stuff back.

I FINALLY got it retro fitted with WinCNC control software and have been cutting rigid PVC and Polycarbonate for about 2 months.

I know the plastic sheets are not perfectly flat, but when I get to my final profile cuts (which should be ~.005" into the spoilboard) some locations are cutting into the spoilboard much deeper, and some locations have .010" of plastic still left intact.

I thought it had to do with my servos walking or the control software not doing what it is supposed to do but I am pretty sure it comes down to the actual machine vacuum bed frame being out of whack. Here is what I went through this morning trying to figure it out...


I removed the spoil board from the vacuum grid and I placed my 8’ BRAND NEW level (I verified the flatness of the level with all the other ones in the store, and with the 6' magnetic one I already had) in different locations. I found that the table edges around the perimeter were the high spots when I was checking the table in the Y direction (5’ width), so i moved on to checking the X direction. I didn't know where my other feeler gauge sets were located at that time I test the Y axis later in this post.

I placed the level in the middle of the 10’ length and used a .005” feeler gauge along the bottom of the level. I found that in just about EVERY bolt location I could slide the feeler gauge under the level, but then when I got 3”+ away from the bolt the feeler gauge wouldn’t fit anymore.





This is what it looks like under the vacuum grid where all of those bolts attach the vacuum grid to the frame of the vacuum table….





I then placed .030” feeler gauges under the level on both sides of the Y direction (so that the edges of the table would not effect the height of the level) and in the picture below, the level is basically on top of the bolts that hold down the vacuum grid.





You have to zoom in pretty good, but there are two .035” feeler gauges under the level (in the middle 1/3 of the span) that indicate the center bowl that is formed from I guess the metal frame settling over the ~20 years of service. This bowl is similar in most of the vacuum grid, so the center of my table is definitely low compared to the outer edges.


I started to think that this may be intentional by design, so that when the vacuum grid has material on top of it and pulling a hard vacuum, the table will actually rise UP in the middle due to atmospheric pressure pushing it up against it, but that seems stupid in hindsight because you barely have any vacuum pressure when surfacing a LDF spoilboard.


I then checked the diagonals and found the same problems with low spots at every bolt location, and high spots between the bolts.






Now onto the control software so see if it was causing problems as well. I have the Servopack software and USB programming cable so i can see what the servos are doing as well as what the WinCNC dashboard is telling me.

I was told to do the following from the techs at WinCNC...

Seek Limits, which finishes its macro by going to machine Home which is the front left corner of the machine (so left side of the level in the picture above)

G53Z0 - nothing happened because it was already at Z0

I had the servo usb cable connected to the Z axis servo pack, so I am going to tell you the “Angle from Origin” and “Number of Pulses from Origin” that it showed.

G53Z0 - AfO = 297* NPfO = 46,293

I changed the Jog to 1” and moved down 1” - AfO = 187* NPfO = 41,373

I moved up 1” - AfO = 297* NPfO = 46,293


I then ran a ~26 minute engraving session (hovering in the air obviously) that involved A TON of Z action and when it finished the program it naturally goes back to Machine Home in the Z direction. I then did a G53Z0 command (which did nothing) and the servo program showed the same 297/46,293 numbers as before.

Here is a screen shot of the end of the engraving session…






Sooooooo, after all of that, it appears as though the machine hardware and software are working properly, and that my problem is related to the table being deformed.

I have no problem spending the time to run a surfacing program on the vacuum grid, IF you think that is a good idea. My guess is that i would only need to take off .02” of material, but I won’t know until i actually start cutting and see where all of the real low spots are. The table is made of the phenolic so it is quite hard and i don’t want to do anything harmful to my table, spindle or bits, so if can provide some recommendations as to feeds and speeds for my 4” diameter 3 flute indexable surfacing bit I would appreciate it.

Normally I surface the LDF at 12,000 RPM and 300 IPM with a stepover of 25%.


If you guys think there is something else i should do instead to fix this problem then I am all ears.

Thank you for your time and future responses.

Zac