GrblGru: Free CAM and 3D-Simulation for mills and lathes

[GrblGru]

@pettyofficernick
I am not sure how you used GrblGru.
The DXF file won't load because it consists of 3 lines that partially overlap.
I suspect that you are still leaving too much material when roughing. This puts too much load on the machine during finishing and the motors lose steps, which then causes a crash.
If you provide me with the project file here, I can say more.

@melvinardan
I can reassure you in this case. The size is not an obstacle.

[OuldMan]

Hello gentlemen.
Allow me to provide a translation into Ukrainian.
I made a translation for a better understanding of how the program works in lathe mode.
Unfortunately, I did not succeed in setting up the lathe, but there is a translation.
I have not checked the translation for all types of machines, I do not have such a variety of machines, nor experience with all.
But I responsibly treated the translation and, I hope, it is correct enough for different types of machines.
Thank you.

[OuldMan]

As for the lathe mode. I wanted to repeat a project of turning threading.
Like in the video on Youtube.
I connected the stepper motor as the main motor (spindle drive).
What didn't work:
The Z-axis feed is very slow and has nothing to do with the Y-axis feed i.e. spindle speed.
For example, when changing spinde speed as Y-axis feedrate ($101=...) from 50 to 300 steps/mm (over 50 steps), the travel time per 1 mm for Z axis was approximately 88 seconds in each case.
This is for Z-axis feedrate 888 steps/mm
If set a feedrate for Z-axis 222 steps/mm and with the same range of change in spinde speed the time to move by 1 mm for Z axis, was the same 88 seconds.
I tried on GrblGru versions 3.45, 4.8, 5.0, 5.1 and lathe 1.0
With equal unsuccess.
Unfortunately, I don't have version 3.31, as in the video.
Where and what to configure I did not understand. Here such a problem arose.
If you have any ideas where and what to turn, please share.
Thank you.
Sorry for my English.

[GrblGru]

@OuldMan

first of all I would like to thank you for the translation.I know how time consuming it is, so I really appreciate your work !
At the moment I am splitting the program into 2 parts.One part will be only for milling and robots, and the other only for lathes.
I will include your translation in both programs.

To the problem with the speeds:
I unfortunately don't have the settings from the video anymore either, but I remember that I also had problems with the feed rate at that time.
The reason for this is that the original GRBL version could only control linear axes and not rotary axes. That's why the feed values don't match the real speed at all.
I simply increased the values until the real speed of the axis seemed correct to me.

When I have finished the splitting conversion, I will add a machine model with a real rotary axis to the program. But this will take some time.

Thanks again for your great help, and please let me know if you have any wish for the program.

[OuldMan]

Thanks for the kind words.
Thanks for the quick response.

I included other parameters in the game of settings and something started to work out.
I continue to experiment.
I inform you later.
Regards.
Alex.

[Calibre41]

Hi,

A few months ago I began purchasing the parts I thought I needed to allow me to 3d scan circuit boards including their components. I found an old youtube video of GrblGru using the G38.2 and thought "YES, this is just what I need!!!"

I have finally today got my 3018 CNC router setup with 3D touch probe, unfortunately it does not work as I'd hoped, and I wondered if you could help me? It's a Z height issue, I only have >40mm to play with, and the part I'm scanning is probably circa 25mm from low to high with vertical components, therefore, I have set the retraction to 25mm, therefore after touching a high component, it then tries to retract 25mm from that height. This is past the physical limit switch and causes the whole thing to stop. Rather than retract an offset from the measured point, can I retract to the same "known safe" Z value? - my current safe Z=-2 because my probe is a bit low, and my part is a bit high on a board and metal standoffs for ease of mounting and rigidity for scanning.

With any luck, I've attached a photo of a random low cost PCB I have setup just to test the process






Please also note, I am a beginner, I can see my measured points are way over on the right hand side off the visualisation of the machine, I haven't worked it out yet, when I CNC, I have a X/Y bottom left 0/0, and Z at the top. I wanted GrblGru software to home to my limit but it wasn't working - I haven't googled it yet, and it doesn't matter for my point cloud really, so long as I position and set "workpiece zero" so the scan happens in the correct place.

[Calibre41]

I'm not sure my pictures worked.... take 2

[GrblGru]

The current algorithm assumes that the target has only slow rising/falling shapes.
The input value "Pull back" is the path length that the probe takes off relative to its last measurement.
In this way, objects can be measured relatively quickly that have larger measurement differences.

But this does not fit to your measuring task, which contains steep rising edges.
I see here only the possibility to always lift up to an adjustable height, which would be very time consuming.
At the moment I am revising my internal program structure, which will also take some time.
But after that I could set up a new mode for you.

May I still ask what the scanning of the PCBs is needed for ?

[Calibre41]

objects can be measured relatively quickly

Yes, as soon as my scan failed, I realised why it was this way. What I did is I set the tool 1mm above the highest point, and then measured down to the spoil board, and used this value as the retract value.

I'm basically creating a CMM machine, https://en.wikipedia.org/wiki/Coordi...suring_machine

The purpose of the machine is to digitise objects, accurately so close fitting components can be designed and machined to fit. I have tried several optical 3D scanners but none of them achieve any reasonable accuracy for this type of application regardless of what they advertise. Typically I make custom coolers and cases for my computers for fun, I design them on CAD or just make them by hand from profile or manual milling machine from solid flat bar perhaps. These might be air or liquid coolers. The height from one component to the next has to be accounted for accurately, ideally <0.1mm as even small gaps lead to very large increases in thermal resistivity if one cooling block is directly cool 2x components with high thermal concentration, some of the chips are typically 15mm square or less, but can output anything from 5w to 200w of heat, it's an incredible amount of heat, in a very very small area, hence the custom coolers!

Another application I have for it, was scanning a metal brackets that incorporated unusual angles, it looks simple, but it has proved very difficult to replicate, if I had an accurate 3D model, I could create a sheet metal part in CAD and then a flat pattern and machine a simple die block to press a flat pattern into shape. This is for a classic bike I have, I know of only one original example of this metal bracket that I can possibly get my hands on, and I certainly can't squash it flat just to measure the flat pattern - Although that would the simple solution, I don't think my friend would be very happy!

So in short, I'm making highly accurate 3d models of parts, many times more accurate than an optical scanner could achieve, with high confidence in measured points.

[antidesco]

finally, I'm finish the upgrade to my lathe, I'm happy

[Cator]

Hello everyone, and thanks to the creator and developer of Grbl Gru for his amazing program. I have this problem, when I click on STL instead of appearing the 3D modification options this menu appears in gray. What do I have to do to make it work?
Thanks in advance

[jacksin]

Hello, thanks for this great program, it's helping us to make a lot of parts on my lathe since five years ago. As I tried to switch from Grbl1.1f to GrblHal cause of the benefit of backlash compensate setting, everything seem to be running well but the Y axis is not moving, therefore, we cannot use the Y axis for threading . Here I post the setting copy of the GrblHal . And, one more problem on GrblHal wifi setting, as I config the $73, $74 and $75, all the setting data of above setting will be gone after the program restarted. This data disappear after restarting problem was only happened on GrblGru version 5.0 , 1 and 1.1 but not happened on former version 4.7, please advise how to solve the problems on the latest version 1.1, thanks.

[jacksin]

Hello, the problem on Y axis stop moving is solving by set the $4 value equal to 4. And now the Y axis is running as Grbl version do. But the problem on program restarting which caused the $73, $74 and $75 valve disappeared was still arousing.

[GrblGru]

@antidesco
Thank you, it's always nice to get a positive feedback.

@Cator
in the 3D view you can select an STL object with CTRL left mouse button. The STL is then displayed in a box.
After that you can open the 3D context menu with CTRL right mouse button.
To deselect the STL you have to click with CTRL left mouse button somewhere outside.

@jacksin
Thanks for the GrblHAL settings. I'm currently working on restructuring the milling program and so unfortunately haven't gotten around to trying out the backlash compensation yet.But I have a lot of interest in it myself.
I am glad to see that the Y-axis works again. Unfortunately I can't say anything about the parameters $73, $74 and $75.
I do not describe them. However, in the initialization I perform a soft reset ($RST), but in my opinion it is necessary.
I assume that this resets the parameters. Please try if you can clear the parameters in terminal mode with %RST.
If so, we will find a solution.

[Cator]

Thanks for the reply, I managed to do a lot of things.
Now I have another question.
The dxf that I produce from my Cad are usually read well by every CAM but in GrblGru I have a problem with the splines, because they are not recognized. Is there anything I can do?

[GrblGru]

Thanks for the feedback. I had not considered the spline function until now.
I am in the process of separating the program into two parts. One for milling, and one for lathes.
So at the moment I can only offer you my current alpha version, which should also read splines.
https://www.grblgru.com/Alpha
It would be nice if you would test it out.

[Cator]

Thanks, I've already downloaded and put the program to work. Splines are handled if dxfs are saved from AutoCAD®2000 onwards.
If export is used AutoCAD® 14 files or earlier are not read. I say this for the use of those like me who work with Alibre Design who can export the different editions of the Dxfs.
Do I here other reports regarding the alpha test or do you have a dedicated mail?

[GrblGru]

Feel free to report here if something doesn't work or you have suggestions for improvement.
But I also just sent you a PM with my email address.

[Cator]

Hi, unfortunately I had private messages disabled and I didn't receive your email, could you resend it again?

[GrblGru]

Sorry, but it is still not possible. Please check your settings again.