Raster Disaster (Or - How to avoid Conventional direction)

This forum is for users to post tips and tricks they have found useful while working with VCarve Pro
User avatar
Adrian
Vectric Archimage
Posts: 14544
Joined: Thu Nov 23, 2006 2:19 pm
Model of CNC Machine: ShopBot PRS Alpha 96x48
Location: Surrey, UK

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by Adrian »

ZipperHead55 wrote:
Thu Feb 16, 2023 2:42 am
Rcnewcomb wrote:
Wed Feb 15, 2023 9:56 pm
Take a rotary raster gcode and turn it into continuous raster, essentially reversing every other raster line.
There is no need to do this if you use the Milo Scott method.
The OP isn't using Aspire, so he can't use Gadgets.
Gadget work in VCarve Pro and Cut2D Pro as well not just in Aspire. The Desktop versions don't support gadgets.

User avatar
SteveNelson46
Vectric Wizard
Posts: 2282
Joined: Wed Jan 04, 2012 2:43 pm
Model of CNC Machine: Camaster Stinger 1
Location: Tucson, Az.

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by SteveNelson46 »

Also, the Milo Scott method of rotary turning isn't a gadget. It's a technique used to carve 3D with a spiral instead of a 3D toolpath.
Steve

User avatar
adze_cnc
Vectric Wizard
Posts: 4325
Joined: Sat Jul 27, 2013 10:08 pm
Model of CNC Machine: AXYZ 4008
Location: Vancouver, BC, Canada

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by adze_cnc »

I guess the confusion lies in the "1001 Vectric Continuous Turning Round 27 Mar 2021.pdf" file. Section 6 (6.1 especially) mentions the use of the "Wrapped Spiral Layout" gadget.

User avatar
TReischl
Vectric Wizard
Posts: 4584
Joined: Thu Jan 18, 2007 6:04 pm
Model of CNC Machine: 8020 48X36X7 RP 2022 UCCNC Screenset
Location: Leland NC

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by TReischl »

Just some terminology clarifications:

The term "post processing" is not unique to Vectric products. This term has been used since the earliest days of NC/CNC dating back to at least 1974 when I first started with NC machines.

The term "pre processing" was/is used to describe modifications to a g-code file after post processing but before loading it into the machine control. I know, I know, it makes no logical sense whatsoever to call "pre". But that is what it has been called for as far back as I can remember. For a frame of reference, way back in the day I worked for MDSI, Numridex and Encode. So I am not the one making these terms up.

Frankly, I always thought it should be called post-post processing. . . . .

Edit, a little more info on these terms: It was called post processing because the computer software prepares what is commonly known as a "CL" (cutter location) file that is not machine specific. Then that file is "post" processed into machine specific g code. So, internally, there are two "processing" passes to create a gcode file. These days a file is generally not used, just memory. But back in the day of 8Kb and 16Kb RAM computers, files had to be used. Yup, the first programming system I used had a Data General 8Kb RAM computer, and 8 inch floppy disks. Forget about a hard drive, at the time, a 10Mb hard drive was north of $10K. How times have changed!
"If you see a good fight, get in it." Dr. Vernon Johns

eitantal777
Posts: 15
Joined: Sat Dec 31, 2022 11:25 pm
Model of CNC Machine: CNCEST 6090

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by eitantal777 »

Hello friends! Hope you're having a lovely weekend

I have results for you. Though I didn't yet write the script that I promised, I did find a way to create a climb-only raster toolpath, using existing tools. I simply do (an over-extended) rectangular offset toolpath, then I clip the parts that I don't want, using this tool: https://www.scorchworks.com/Gcoderippe ... pper.html . The end result is very similar to my original plan, which was to take a raster, reverse every other line, and start from the middle outwards.

I did a test run, with deliberately bad circumstances, to emphasize the effects. (bad circumstances: 0.25mm TBN, 40% step-over rate. Other settings: 18k rpm, 1000 mm/min)

I made the following runs:
(1) Cross-grain raster
(2) With-grain raster
(3) Climb Offset
(4) Climb-only raster (using With-grain orientation). I'm aware there are errors in the GCode. The point is to assess the fuzzies
(5) 2-pass raster: With-grain raster, followed by cross-grain raster
(6) Climb-only raster (using Cross-grain orientation).

Offset looks the best IMO, but it does have this obnoxious "X". Surprisingly, the 2-pass looks worse than I thought.

(1) Cross-grain raster
Image

(2) With-grain raster
Image

(3) Climb Offset
Image

(4) Climb-only raster (using With-grain orientation).
Image

(5) 2-pass raster
Image

(6) Climb-only raster (using Cross-grain orientation)
Image

User avatar
adze_cnc
Vectric Wizard
Posts: 4325
Joined: Sat Jul 27, 2013 10:08 pm
Model of CNC Machine: AXYZ 4008
Location: Vancouver, BC, Canada

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by adze_cnc »

eitantal777 wrote:
Sun Feb 19, 2023 2:49 am
I did a test run, with deliberately bad circumstances, to emphasize the effects.
The acronym GIGO comes to mind.

In those TReischl links I posted earlier he goes emphasizes using deliberately good circumstances.

eitantal777
Posts: 15
Joined: Sat Dec 31, 2022 11:25 pm
Model of CNC Machine: CNCEST 6090

Re: Raster Disaster (Or - How to avoid Conventional direction)

Post by eitantal777 »

adze_cnc wrote:
Sun Feb 19, 2023 7:56 am
The acronym GIGO comes to mind.
But it's not garbage-in-garbage out, tho. You can see how going in the climb direction, even with the grain, results in a reasonable work with almost no fuzzies, while going raster with grain results in garbage. It would stand to reason that under better conditions, the same effect still applies, except on a smaller magnitude: You'll get Good work vs. Excellent work. (Or, alternatively, if you use climb-only-raster, you can pick larger stepover % for the same quality of work)
Not only that, common wisdom says that the maximum stepover% is 50. So according to this, my conditions weren't techinically garbage in the first place.
adze_cnc wrote:
Sun Feb 19, 2023 7:56 am
In those TReischl links I posted earlier he goes emphasizes using deliberately good circumstances.
Fair enough. I will repeat the same experiment twice: Using reasonable conditions (20%), and using favorable conditions (10%). It is possible that the effect will disappear entirely, and you get Excellent work vs. Excellent work, in which case climb-only becomes a must, only when using 40% step-over. climb-only-raster becomes a must if you don't want the X mark you'd get from offset.

Post Reply