Union Stars not same size

[jhone22]

I'm hoping someone can help me. I carved this file using a 90 deg vbit. It ran fine. When I tried to do a second run. The picture below happened. I am using the new star size on my layers. I have included the crv file. if you need my tool path as well I can include that too.
Thanks for any help.

[highpockets]

File look fine.
Did you surface the material, looks like there might be some highs and lows in the surface of the material.

[jhone22]

I planed it to 5/8" there is a little unevenness, but I didn't feel like it was too much. Is it better to surface it than use a planer?

[scottp55]

Personal opinion,
but there's not much better for uniformity than surfacing with your machine for consistency.
People say thickness sanders do the job if spoilboard is Flat, but no personal experience.
The smaller you go on fonts and delicate carvings...the more important it is.
Sanding to final grit before carving will let you know Exactly what it will look like.
My 2 cents:)
scott

[highpockets]

I prefer to clamp the material to the spoilboard then surface it, this will ensure the material surface is parallel to the XY of the machine. Plane the material can release some stresses in the material cause it to warp/twist before getting it clamped to the table. If the material requires planing first it's still a good idea to run a surfacing pass once clamped to the spoilboard.
JMTCW

Update: What Scott said

[jhone22]

Ok, I appreciate the input, I'm new to cnc and I am learning a lot as I go. This forum is very helpful. I'll definitely pick up a surfacing bit. Are there any other things I might check or do you think its an uneven surface problem?

[jhone22]

So I just went out and took a closer look at the edge of the board and sure enough, It was actually more uneven than I thought. I'm impressed you could tell that from a picture. I'm going to fix that and re-run the carve. Thanks again for your help!!!

[Rcnewcomb]

I'm impressed you could tell that from a picture.

Once you have been bitten by that type of Gotcha! you tend to remember it.

[Mark Bolton]

The bonus is, if you can register the piece where it was, and the finished thickeness isnt critical, you can just deck off that one and re-run your stars and save it if its worth saving.

[Savannahdan]

The controller program for the first cnc I had would take 5 readings across the workpiece which helped in carving on an uneven board. It worked really nice and had the side issue with it that you tend to not make sure a board is flat to the bit. The cnc and service from the company that made it were not good, so much that I changed cnc machines. I recently purchased a planing bit that uses carbide inserts and I very much enjoy that part of my routine.

[scottp55]

Yep:)
Like Mark just said, many pieces can be "saved" IF you notice Before removed!
Once taken off spoilboard, all bets are Off!

IF using an RO sander....using a "Hard" pad (or staying away from edges with a "Medium")...
then you can sand to finish grit before carving, so "what you see...is what you get".

When purchasing a surfacing bit, I recommend you do a bunch of trial cuts in your chosen material
to find the "Sweet Spot" for speeds/feeds for minimal sanding.
Often depth of cut, sharpness of bit, and orientation of cut, can cut sanding down to 220G.
( depth of last pass(or running twice) with a sharp bit can save me lots of sanding time)...
MAY be a Myth, but I keep all my surfacing bits(and most of my Finishing bits)....
separated by MDF and Hardwoods.
(Recommended by Shopbot people when I started, and Seems to work for Quality cuts in Hardwoods.
Seems like the glue in MDF dulls bits faster than normal...and that results in lower than normal quality
in my Hardwood cuts).
MDF people seem to Accentuate the importance of a "Sharp" bit for"Curls"...
I use MDF so rarely, I keep bits separated so "Fine Details" don't require sanding on my "Tiny" to "Smalls".

It works for me....but some people claim enhanced audio by coloring CDs Green

Best quality when surfacing for me and my machine seems to be >.02" in Hardwoods...
(Cherry + in Janka Hardness)...Running surfacing twice seems to reduce sanding.

IF you notice a "Pattern" that does not disappear with your finger or Very light sanding...
then check "Trammel" or "Surfacing marks" in threads.
Your spindle/router may be Not perpendicular to surface of material.
Worth your while to fix!

IF you want to "sign" or dedicate a piece on the backside....
Highly recommend you surface Both sides so they are parallel for small fonts such as
Signature/Trademark/material or finish(for future generations...(bit me a few times on restorations) )

PITA when CNC means "Pressing a button" and winding up with a perfect piece
scott

OH! Rollers on most Planers(Thickness Cutters), press the wood so firmly down..
That it May be flat, BUT most likely has the same warp as the original board when it comes out!
https://www.youtube.com/watch?v=SbIzd1I ... e=emb_logo
(LOVE this vid because it shows some of my "branches" )

FLAT spoilboard is imperative!!

P.P.S. Like Dan says....sometimes watching your surfacing can Enhance a carving!
Surfacing "Found Wood" is one of my most enjoyable Tasks:)
Welcome to "The World"

[scottp55]

Dan, can I say that the "Virtual Z" did Not always work as advertised???
Kinda hard to believe unless so many samples of "Z" as to make it almost a 3D sample?
THAT takes time, depending on parameters!
Rough is good Sometimes....NOT always!!!

Shopbot does Not have that function unless 3D Probe is used.
PROBABLY WRONG, as no Shark experience, but no not trust a 'ONE of' to a function.

Only going by a few examples in last few years.
It may be better at "sampling now' i'm ignorant of that.

[Mark Bolton]

A true "virtual Z" would seem pointless unless your customer/job is forcing you to drape a toolpath on a fixed existing surface that can not be modified in any way. At which point the toolpathing will be distorted anyway as projected in a 2D environment. It really doesnt solve anything other than to make a toolpath which was intended to be on a flat/monolithic plane machine distorted. How out of flat the part is would effect how well the distortion would work but with a field of star and so many points it would be a fail without a ridiculously high point cloud for probing.

Might be cool for one of these carved wavy flags to have the stars try to follow the waves but at that point your pretty much in 5 axis world.

The cost per minute of probing (whats baseline cost for CNC time period? a buck a minute?) and further debugging would make a job with a 15-20 minute run time ($20) cost several hundred dollars (couple hour run/design time?). Then add in materials and shop time.

[Tex_Lawrence]

I thought "Virtual Z" came out of cutting traces in printed circuit boards since there is a very small margin for change in Z.