V Carving/Z depth consistency issues

[johnm]

Folks -

Yeegads... More Z trouble! Okay, Rose is creating a round picture frame as a surround for stained glass, so it can be viewed from either side. No real "front" - same carving on each side.

We are milling the frame parts out of 1/2" mdf, and this is only the second job since I've flattened the table. I had each half of the frame panels screwed down next to each other, and they came from the same sheet of mdf - yet when we v bit carved the decorative elements on the face side of each half, one side carved deeper than the other.

We set the final cutout depth at .520 and I checked the depth cut into the spoilboard and it was the same beneath each piece. The thickness of the mdf varied by less than .005...

We are using a 1/2d 90d Vbit for the detail work. I suppose that the bit *could* have slipped to give me a cutline that was precisely twice as wide as I needed, and slipped no further for the duration of the file. If that IS the case, then someone PLEASE suggest some lottery numbers to me ASAP.

I have included a pic showing the faces of the two frame halves side by side and you can clearly see that the carving is deeper/wider on the left portion of the frame. We ran one set of files for both frames, using the 2D offset feature. (Thanks for that tip, Bill Young!)

Now, for the bonus question... The width of the v-bit carving centerlines between two parallel lines .0275 apart, which is narrower than either of the carvings we got. The carving on the left side is about 1/8" and on the right, 1/16...

Our spoilboard consists of three sheets of 1/2" birch plywood, with the first two layers glued and screwed, on an 8" grid then the third sheet of plywood glued and screwed, using brass screws, also on 8" centers (offset 1") then two layers of 1/4" tempered hardboard, glued, and then 8 sheets of plywood on top for about a day and a half whilst the glue dried. The table was then surfaced and I removed .040 to bring it true on the first go round, then I ran a surfacing routine a few days later and it was still flat, straight out... but I took off .010 just on GP.

Our SB has a home built, all welded steel table that weighs nearly 1,000# and I cannot discern ANY flexing of the spoilboard at all. Even if there were some flex, how would carving a 1/16" groove in what amounts to solidified termite barf throw such a wrench into our plans?

I know I may have rambled on a bit here, but we have had more trouble with Z issues than anything else, and to remain polite, it is quite frustrating.

Any ideas?

John & Rose
Lakeport, CA
3932

[richards]

Can you post the last dozen lines in each file and the first dozen lines in each file? I'm thinking that somehow the reference height on the deeper cut is different than the reference height on the shallower cut. Are you using the FP command to call the files from a master file?

I used to get that type of problem all the time when I cut plaques. My vacuum system wasn't consistantly pulling the plaque down to the same height from part to part; however, since you're screwing the boards to the spoil board, you shouldn't have to worry about height variation.

[knedert]

We used the FP command to call the files from a master file. The only difference was to use the "Offset in 2D" command after moving to the second X location. The boards were right next to each other, screwed to the spoilboard...

[btk]

Rose,

Just make sure that you were not selecting "Offset in 3D" (as opposed to 2D). Offset in 3D will make the current location 0,0,0 and would effect your Z.

Brian

[knedert]

Ha! Brian.... You made me look!
I rebooted the 'Bot computer just to see...
But... no sadly that was not the answer. The log for the file clearly said: offset 2D

Any other ideas guys?

[btk]

While you are at it, also check that you are in Absolute Mode (as opposed to Relative Mode).

This is something else that do funny things to your Z values.

Brian

[btk]

Rose,
I can not tell by the photo, but is the outer boarder of the design V-Carved (chamfered) or is that done with a flat bottom bit (and a tool change).
How are you doing the cutout? (are you doing a toolchange) and a Z-Zero Routine in-between.

Brian

[beacon14]

not to state the obvious, but are you sure both workpieces were firmly seated to the table, no sawdust or chips beneath - did you hold a straight edge across both faces to check that they really were in the same plane?

Was each entire frame routed to a uniform depth?

Is it repeatable? What if you put a single piece of MDF across that area and ran it again?

[johnm]

Folks -

Thanks for all of your remarks - Rose will chime in here momentarily...

To recap, we used the same file, with a verified 2d offset, and we were running in relative mode - we had moved zero X about 45 inches down the table, with zero Z at the top of the material.

We used 3 bits, a 1/4EM, a 1/8 ballnose and a 90 d chamfer bit. We had the material screwed down in the center waste area as well as each corner of the 16x16 blank. We registered each piece with dowels, so that we could mill both sides and maintain registration. I didn't put a straightedge across them while they were on the table, but I did check and they were flat relative to each other (they were butted up against each other)

The "inside" of each piece was milled down with the 1/4EM, and a glue channel with the 1/8"BN bit, then they were flipped and the milling continued with the 1/8 BN - the inside radius next to the inside chamfer, then carved and then cut out. We used bridges on the outside to hold the frame and the inside was cut clear.

We will mount some more MDF and run some test cuts to see if we are still having gremlins.

Hope this fills in the blanks and that you will all have some more ideas for us. Again, Rose will post here shortly with anything that I've overlooked.

John

[btk]

John,

You say in your latest post that you are running in Relative mode. I am assuming that you are Zeroing the bit inbetween moves.
For instance, if you zero before first cut (bit starts at z=+0.5") and then the file runs and when finished it goes to z=+0.375)
Then you JX 45 inches down the board.
If you are Relative mode and restart the file (witouth re-zeroing the bit), you will go deeper in to the wood by 1/8" then the previous cut.

If you were in Absoulute Mode (which is recommended), then machine would not care what Z you were at after first cut and would just plunge the same amount.

Brian

[btk]

I meant to say in the above example that you would end up going 1/16 deeper if you Clearance plane was 0.375 (in relative mode).

Also, i amd not sure what effect if any running in relative mode has on the Offset 2D command.

Brian

[knedert]

Hi Folks,

Sorry it took me so long to "chime in"... was busy vector bustin'.

First, let me make a correction. We are running in absolute mode.

We set our X,Y zero at the lower left corner of the left panel of material, and Z-zeroed there for all of the bit changes that John mentioned.
Since we were using the 2D offset, we did not Z-zero on the second panel.

FYI: The picture John posted should be flipped, as the deeper cut was actually the second offset 2D cut.

Again, sorry about the confusion and correction delay.

As soon as I make a few more asthetic changes to the file, we'll try again to see if we get the same results.

We truly appreciate all of the help we get here on the forum.

[paco]

Hey Jon (Rose)!

1)Is the part held securly/firmly?

2)Have you air cut the file with every thing running as (router/spindle, dust collector, vacuum) to try to duplicate the problem? You can check for that with a piece of say 1" material before running the file then, once ran, back to home position, check again with the gauge (1" thick piece of material)...

3)Have check your tool for free play; router/spindle collet/chuck/bearing, router/spindle mounting, Z axis pinion and rollers, Y axis, X axis... a dial indicator is useful for this.

4)Are you holding your part with vacuum; I recently discovered that my MDF support#spoilboard assembly is swelling with the vacuum running for long period... up to 0.03" upward movement read from above and up to 0.01" read from bellow...

5)Is the material consistent?!

6)Do you remove a lot of material around before the V process?

7)Do you run the SAME file on the two sides?!

8)Do you change tool in between the two side V Carving process?

Some thing is moving... the software is easy to sort out...

[gerald_d]

...did you hit the top of the z-axis, or a dirt chip in the rail, between parts/files and lose a step or two or more?