General discussion

[martin2250]

as the title states. please only use this thread for questions and discussion and open new feature requests for actual issues with OpenCNCPilot.

Martin

[fantommunky]

Hi. I'm really new to cnc and this may be a retarded question, But..... How do I jog the machine in OpenCNCPilot??? I have come from USG and can see no way to jog in OpenCNCPilot. Thanks in advance for having to deal with a simpleton :)

[canbaytok]

Expand the "Manual" Tab Press on "Enable Keyboard Jogging" Press the textbox with the text "Jogging Keyboard Focus" so that you can see the cursor Use your arrow and PageUp/PageDown keys to jog the machine

[canbaytok]

Hi there @martin2250

I had to reinstall Windows yesterday and after backing up all the necessary files, reinstalling and starting OCP I noticed that all the settings were defaulted. I would love to know where the settings are saved, so I can back them up next time!

[martin2250]

Hi @sirsenor,

the settings are stored under %appdata%/martin2250/OpenCNCPilot (not 100% sure, could also be in %appdata%/../Local somewhere). I'm not sure if it's possible to just copy them from one PC to the next though.

[canbaytok]

Alright I found them. I'll try the copying process with a virtual machine and see if it works!

Thanks a lot!

[classbproject]

There is an XML file in appdata, not sure what it is called but I can look it up. Copying that file has worked for me (with the same version of the app).

[deHarro]

The settings of OCP are stored uner "C:\Users\x_username_x\AppData\Local\martin2250\OpenCNCPilot.exe_Url_XXX" where x_username_x expands to the user where OCP was installed under and XXX expands to an arbitrary(?) GUID-like monster of characters and numbers. Beneath that path, there are one or more additional directories which are named according to the version number of the used (once started) version of OCP. In each of these directories is one file "user.config", which contains all settings for OCP which are accessed at least once by the user. Settings which are not changed yet, are not listed.

As to my experience you can copy and reuse the whole file if you use the same version of OCP (as sirsenor stated above). If you want to rescue your precious userdefined macros from one version of OPC to a different version, you have to copy just the part containing the macro texts, obviously surrounded by all accompanying XML stuff to get a valid XML entry. I'm not sure whether you have to take into account the position within the file or at least against preceeding or following entries.

Harald

[martin2250]

I would like to add that this is not my wrongdoing, I used the default settings manager in C# :joy: The easiest solution would be to back up user.config, run OpenCNCPilot once on the new PC to generate the config folder and copy user.config there. Theoretically, you should be able to copy folders from older versions and reuse them with new versions, OpenCNCPilot will try to load the old version, but only if the new version does not exist yet.

[beckmx]

hello @martin2250, I have been using opencnc pilot I think for a year now, very happy with it, I would like to suggest a feature that comes handy. For instance, when you are about to mill something, you make a probe map and start doing it, however sometimes I notice that because the thickness of the material the final depth was not enough, I would suggest that in the application you can select a section that you would like to repeat. So you draw a square in the screen to select what section to mill again.

Thanx for your work!

[martin2250]

Hi @beckmx,

thanks for the feedback! That's something I too wanted for some time now, but not as a part of OpenCNCPilot (I want to keep OpenCNCPilot lightweight, so the only edit commands available are ones that can't be replaced by external programs). Right now I'm too short on time to take on such a project, but maybe in the future I'll release a second program for editing gcode to go along with OpenCNCPilot.

Martin

[bigshug]

Hi Martin. Really handy looking bit of software! Is there any way to export the data as a model? I'd rather use your software than BCNC although it exports the data as STL.

Thanks.

[martin2250]

Hi @bigshug. Moving to a better file format for height maps is on my to-do list, but I don't plan on using the STL format. That said, it should be easy to write a small python (or whatever floats your boat) script that converts .hmap files to stl.

Martin

[bigshug]

That's great news! Will keep an eye on it. :o)

[eternal-optimist]

Hi Martin I use a 3mm aluminium strip as a touch plate for setting Z zero as well as bed-leveling. I do a lot of engraving on acrylic/acetal sheet which means I have to have a metal plate for this. Question is, is there any way I can use the TLO section of the program to somehow store this 3mm Z offset which I can later use to reset my 3mm 'higher' (false) Z zero back to a true Zero for the Z Axis before running my gcode? If so, exactly how would I go about it?

Thanks for your fantastic program, Richard

[martin2250]

Hi Richard,

no, the TLO is built to reference machine coordinates, for using a fixed tool length sensor. For your application, macros should work just fine. There should already be a "Probe and set Zero" macro. If you replace the "G92 Z0" with "G92 Z3", it should do what you want.

Best, Martin

[eternal-optimist]

Many thanks Martin, never thought about it being that easy...duh!

Richard

[eternal-optimist]

Hi again Martin, Have now got on top of setting my Z Zero correctly thanks to your previous advice - but now having trouble getting the heightmap to look correct on the drawing. Mine seems to put the heightmap below the visualisation of the item being worked: In your videos, the heightmap appears to 'float' over the top of the workpiece - whereas mine seems to sit below the workpiece. Not only that, even though I have set the Z Zero correctly, and I have checked it is at Zero using a feeler gauge, on starting to run the file, the bit is running at approximately 0.5mm above the workpiece - missing it completely?

A bit confused now but - I will include my heightmap and project file (deliberately misaligned on the bed to test) so that you can re-produce what I am seeing at my end. Hope you can help me out here (again!)

Heightmap and Project File.zip

Many Thanks, Richard

[martin2250]

Are you using the metal plate for probing? That will make OpenCNCPilot think that the surface is higher than it actually is. There is no way to account for this offset automatically, so you'll have to zero your Z-axis on top of the metal plate (3mm above the acrylic), run the probing cycle and then lower your Z origin by 3mm. (G0 Z0 & G92 Z3).

I missed the "as well as bed-leveling" part in your last comment.

Martin

[eternal-optimist]

Martin, Yes I am using the metal plate for probing (1.21mm thick - new lighter plate now), but I thought that I had taken account of this with my Probe Macro (following your earlier advice) as you will see:-

Still not sure why the project file movements on my heightmaps show below the workpiece - unlike your video which shows the the project file movements all parallel to, and above, the accentuated curve in the heightmap. Any ideas on why/how this could be happening?

Kindest regards, Richard

[martin2250]

That macro will only change the work origin. It has no effect on the probing routine. In my video, I'm using a much shallower engraving depth (prob. 0.04mm), so there is a higher chance that the height map is below the original tool path at some place. After applying the HMap, the toolpath should be below the height map everywhere. Follow the advice in my last comment and it should work properly. The height map will appear too high initially, but when it's finished and you change your origin appropriately, your toolpath will match the surface exactly.

Best, Martin

[eternal-optimist]

Martin, I'm getting there!

I've figured out now where I was going wrong, I was not allowing for the thickness of the probe plate when I 'Apply the Heightmap' and I have now been able to get it to cut at the same depth when cutting on a sloping plane - great progress (for me anyway)!

However, I have now discovered a new problem. Whenever the cutting head moves to a new position during the running of the file to start a new pocket in the job, which is further 'up' the slope in my case, the cutter is sent to Z zero before lifting back up to the correct height to continue to cut the pocket at the correct (HMap amended) height. This is leaving me with numerous half-drilled holes everywhere on the job, not what I want!

I don't think this is anything to do with the way your program alters the file when applying the heightmap, rather something I am doing in the software I use to create the file. I am using VCarve from Vectric and I am using the standard grbl post processor.

Unfortunately I'm not too good at reading the gcode so I don't really know quite how to get around this.

If you look at my Hmap amended file, you can see the moves causing the problem on line numbers 1014,1032,2656,3097,3493 as well as further throughout the file.

Once I get to the bottom of this little problem, I think I shall be 'Home and Dry' as they say.

Please take a look at my file would you and try to give me any pointers you can?

hmapamendedcoaster0.4mm.txt

[eternal-optimist]

Martin,

OK now - solved it. There's a Clearance/Plunge setting in VCarve that had (somehow?) been set incorrectly - mostly the plunge setting I think which I have now set to zero - and it seems to have mostly done the trick, just very small indentations now where I had 3mm holes before. Think I can hopefully sort it out on my own from here.

Once again, many many thanks for your support - without which I would have remained stumbling in the dark for weeks!

Regards Richard.

[martin2250]

Hi Richard,

that's great news, glad I could help. Feel free, of course, to share some results on the project gallery thread.

Martin

[eternal-optimist]

Hi Martin, Following upgrades to my machine, (increased X and Y axis sizes), I am now getting into macros a bit deeper. I happened to see a great gadget for doing the XYZ zeroing by Charley Thomas - called a Triquetra. I jumped straight in and started looking to make one of these (now done) and get it working on my new, much larger baseboard. Have now been at it for a couple of days but can't seem to make the thing work correctly - it always fails when attempting the final probe to determine the height of the 8mm touch plate. It does the X and Y fine and then moves over the top of the touch plate to measure the Z offset - but then ignores the 3 lines (G38.2 Z-25.4 F25:G92 Z8.1:G91 G0 Z3.175) which probe down to set the Z height, and carries on to send the router head 25mm in both the X and Y positive direction to clear the touch plate. This is my code:- G92 X0 G92 Y0 G92 Z0 G21 G38.2 X-12.7F25 G92 X41.575 G91 G0 X12.7 G91 G0 Y25.4 G91 G0 X-25.4 G38.2 Y-25.4 F25 G92 Y41.575 G91 G0 Y03.175 G91 G0 Z11.85 G91 G0 Y -15.875 G38.2 Z-25.4 F25 G92 Z8.1 G91 G0 Z3.175 G91 G0 Y25.4 X25.4 M30

I have tried the same macro in UGS and it all works fine, problem is I just don't like the program - yours is far superior in my view.

I'm wondering if anyone else had experienced anything similar, and if so, how it was resolved.

Any light you maybe able to shed on this would be appreciated. Have split the Macro for now as a workaround - X & Y on macro1, and Z is now a separate Macro2.

Later Note: Just to say I have just tried the above Macro on the latest version of the Openbuilds Software (re-flashed only minutes ago) and again, like UGS, it works perfectly. This is leading me to think that the problem I have outlined is definitely , in someway, linked with something within the openCNCpilot software - which is waaaaay outside my understanding to play with. Could be API calls, dll problems or whatever, those things are all Greek to me. Two other highly respected programs perform the Macro faultlessly - pity that I dont like either of them...........

[silicon-systems]

Hi Martin, A great auto leveling GUI, may I ask if you will be adding floating machine control panel so as to make navigating 'Z' probe over X & Y plan a breeze?

Cheers, Hanspeter.

[silicon-systems]

Hi eternal-optimist check out your gcode line 5 "G38.2 X-12.7F25" shouldn't you put a space between X parameter and Feed-rate setting?

Cheers, Hanspeter.

[martin2250]

Hi @eternal-optimist

sorry for the late response. Did you figure out a solution yet? My guess would be either the preceding zero in G91 G0 Y03.175 or the extra space in G91 G0 Y -15.875.

@silicon-systems the missing space should be fine. Whitespace between letters is allowed. OpenCNCPilot actually removes all spaces when sending a file to the machine. I thought this would also be the case for macros, but after checking it isn't. As the line G91 G0 Y -15.875 directly precedes the skipped lines, I'd bet on that.

Martin

[martin2250]

Hi Hanspeter,

can you elaborate on your suggestion a bit? Do you mean an extra window or a window 'inside' the main OpenCNCPilot window? Also what do you mean by "navigating 'Z' probe over X & Y plan"?

Best, Martin

[silicon-systems]

Hi Martin,

Sorry for lack of better explanation, see attached image file. This Jog control panel could be fixed or float above the OpenCNCPilot UI.

Best regards, Hanspeter.

[martin2250]

Hi Hanspeter,

I personally find the idea of controlling a machine that has potentially bone-crushing strength with virtual buttons on a screen a bit scary (imagine looking over to your machine and accidentally pressing the wrong button), so this is a feature that I won't be adding. Controlling the machine with actual keys on a keyboard is much more convenient anyways and doesn't require looking at the screen.

Best, Martin

[eternal-optimist]

Hi Martin and Hanspeter Just to let you both know that you were both correct, it was extra spaces which were fouling things up... Maybe others will find this useful as well as me, I hope so. Here is the amended code which now performs faultlessly. G92 X0 G92 Y0 G92 Z0 G21 G38.2 X-12.7F25 G92 X41.575 G91 G0 X12.7 G91 G0 Y25.4 G91 G0 X-25.4 G38.2 Y-25.4 F25 G92 Y41.575 G91 G0 Y3.175 G91 G0 Z11.85 G91 G0 Y-15.875 G38.2 Z-25.4 F25 G92 Z8.1 G91 G0 Z3.175 G91 G0 Y25.4 X25.4 M30

Many thanks to you both for sorting me out (again!).

[liam204]

Hi Martin. First off thank you very much for your work on this excellent program!

I have run into a problem where my engraving toolpath is larger than the probed area (I am unable to probe the entire surface due to clamps in the corners). The engraving works exactly as intended until it leaves the probed area, at which point the tool raises too high to make contact with the surface. The images show a clear cutoff point where the engraving leaves the work piece as it has left the height map area. Is this by design? Is there any way to extrapolate the height map out past the probed area. Do you have any other suggestions to help with finishing the engraving? Unfortunately moving the clamps is not an option.

Thanks in advance,

Liam

[AuBeTeX]

Hi!

Sorry my poor english.

I a new user, and sorry if i questioning wrong. I use a 3018 cnc with 1.1f board. I try to milling PCB. unfortunately while milling, my tool is damaged, and i need to change it. How to positioning again the tool along to z axis to use current height map? A new height map is not recommended a half done PCB because some probe points is milled, and this points damage the tool again. have somewhere an exact documentation to solve this problem?

Other questions: Where is the STOP button in the program to abort milling? I only see a pause button. How to use z probe once to set z height. I don't understand the manual probing box.

Thanks

AX

[beckmx]

@AuBeTeX firstly apparently you are doing it all wrong, if you break a tool is because your depth cut is excessive or because your tool is a poor quality one, ideally you will have to remap again because when changing the tool the tip might be moved from the previous center, but if you still don’t care about that because you use units bigger than 0.01 then you are safe for using the button called -probe and set zero- or the command is G38, it does exactly what you want, it tries to probe again the height to reset the z-axis.

Now going back to the broken tool, if you are using those tips from aliexpress those break very easily, in a couple of seconds of milling the tip disappears, you need to buy milling tips with 2 or 4 flutes for instance from precisebits.com I can tell you those do work, in any case for both a professional tip or a Chinese tip the max depth of cut is .09mm if you set a bigger one you will always break it, in order to achieve that depth or a higher one, you need multi depth from flat cam, with steps of 0.035

[martin2250]

Hi @AuBeTeX @beckmx a few thoughts:

• you don't have to remap after you break a tool, even if you do care about 0.01mm. Just re-zero Z and continue where you left off.
• cheap milling bits are not useless. I still use the same 4 or 5 cheap V-bits that I used with my old machine 5 years ago. I've used cutting depths of up to 0.14mm without a problem.
• the stop button is your espcape key. this will send a soft reset and stop the machine immediately. You can also use the 'hold' button in the status bar if you don't want to reset GRBL.

Cheers!

[silicon-systems]

Hi @AuBeTeX @beckmx a few thoughts:

* you don't have to remap after you break a tool, even if you do care about 0.01mm. Just re-zero Z and continue where you left off.

* cheap milling bits are not useless. I still use the same 4 or 5 cheap V-bits that I used with my old machine 5 years ago. I've used cutting depths of up to 0.14mm without a problem.

* the stop button is your espcape key. this will send a soft reset and stop the machine immediately. You can also use the 'hold' button in the status bar if you don't want to reset GRBL.

Cheers!

Dear Martin you are dead right. Have reused my 'V' milling bits so many times, these I bought from Amazon.co.uk and the supplier from China, I had to return a couple, as bits supplied was not 0.1mm.

I also use Flatcam as a pre-processor before milling out PCB.

Kind regards, Hanspeter.

[AuBeTeX]

Hi!

Yes, i'm use cheap chinese bits with a cheap chinese cnc. I'm only hobbist. But something not clear for me. Maybe the board is a crap maybe the machine. So, i have a flatcam nc file i loaded in the OCP generate a height map then start the milling. after that i push the tool on x axis cca 5cm resetting the woorking coordinates, create an another height map, then start the milling again.

BUT between the two milling is lot of differences not the path, but the depth. i check the tool, that's not damaged. and this happened me a bit lot of times. Until now i not figured it out why.

Sorry about my english spelling.

A

[deHarro]

Hi Aubetex,

just a thought:

1. Did you "apply hight map" to your CNC file after probing the map and before milling?
2. And did you reload the CNC file before applying the hight map the second time? That both is essential. If you omit 2. the corrections will be applied two times and the carving will definitely be too deep.

Additionally I made the experience that I better zero out Z axis before probing. Theoretically this must not be done (as far as I understood Martin), but I got decent and reliable results with taht procedure.

Harald

[AuBeTeX]

Hi Harald!

Yes i'm apply that. In my country the clean PCB-s like oceans. Thats Wavy. If i use one probe (with a little exaggeration) the PCB point Y0 is untouched Y100 is milled in to the desk. (i can show some :))

I'm use the Candle before OCP, and really missed the "remove height map" button or checkbox or something from OCP. So i figured out i need to reload to clear heightmap data from nc files.

So both done. but some milled PCB is beutiful some is ugly. Maybe i create a video about this.

I thinking... If the program is known the path. maybe test the height map before milling the path (of course if the probe available under milling.)

[deHarro]

Hi Aubetex,

I reread your posting and realized, you do not carve the same area two times but you carve two seperate PCBs, one after the other. Ok, I misssed that in my first answer, sorry.

Following your description, you seemingly do everything correct.

Perhaps you can try what I mentioned in my first post:

• Load your NC file
• Create a new height map, using "Size from G-code"
• Zero out Z axis before starting the height map probing.
• Start probing.
• Apply the height map
• Carve the first PCB.
• Move X to the second starting point.
• Zero out Z again at this position.
• Start probing.
• Reload your NC file.
• Apply the height map
• Carve the second PCB.

If this procedure doesn't produce two correct layouts, then your mill is the culprit.

One additional hint: Tape down your PCB board using double sided mirror tape. Use the tape all over the to be milled area. Why? I had similar problems and it showed, that I used too small patches of tape, so the board could give way. Different amounts af giving way when probing and carving!

Another problem you may stumble into is the surface of your raw material. I used to use old pcbs, I formerly bought for exposing by ultra violet light and etching with chemicals. The photosensitive layer prevented the stylus to touch without force, so the resulting hight map was some amount deeper as it should be and as it comes out when removing the photosensitive layer in advance.

HTH Harald

[deHarro]

Hi Guys!

As you perhaps know, I use Eagle to draw my designs and afterwards run John Johnsons "pcb-gcode.ulp" to generate the g-code for OCP.

I think I already mentioned my concerns in pcb-gcode's way to mill always clockwise or counter clockwise - one can choose in the setups, but that's it. John calls this "always climb" or not, which is "normal". In normal mode the drill bit cuts against the routing direction which results in well done edges at the thread. The downside is, all successive rounds are cutting in the same direction and so implicitely changing to climbing mode, resulting in not so well done edges on the outer sides of the trenches.

I once asked John whether it is possible to change the g-code generation to make the first round in normal mode and then switching to "climb mode" for the edges that are cut from second to last round.

I think he didn't get the idea since he told me, there is "no way around the way how Eagle generates polygons" and that pcb-gcode is nailed down on the current way it works. Ok, I was disappointed and retarded back then.

But these days I again felt bugged and decided to dive into Johns code and put it right on my own...

After some hard work I thought I got the mechanics of Johns program and added one single line. Guess what... now the milling is always (on both sides of the trenches) in "normal" or, if one wants that, in "always climb" mode.

I still have no feedback from John, but simulating designs with OCP seems to prove my modifiction as ok. I'm on making a pcb just tomorrow and will have a look at the result. I'll keep you posted.

Harald

[AuBeTeX]

Hi Harald!

I'm not use double sided tape. I'm always create fixing holes on my PCB, i'm create that first. Then i'm use that holes. On my worpad has fixing an 1 cm hardwood plate. I'm fix the PCB on that with a masking tape. Then i'm drill the fixing holes. After that i fix the PCB with screws. Then i'm create the height map. I think this is a good method. unfortunately some times looks like the height map is not correct, or my machine can't understand. I'm not figured out why. The height map applied to nc file, and i'm see the changes on Z axis while milling, but the result is varied. I'm thinking about why. The last few days OCP has lost the connect to machine. The time counter continue counting, the coords zeroed. The machine is halted, and wait the commands. I'm restarted the program and push the connect button, send commands, but only the green gauge is fill up, the machine is not moving. i need to start other program (sourcerabbit) to reset the machine, and clear lock, alarm, and whatever problem is. THEN OCP can connect to machine. Oh i'm wrong, OCP can connect to the machine and sended commands but nothing happened. If i'm use sourcerabbit to connect, resetting , after that OCP can command the machine and the machine is work. Maybe my less-than-one-month controller is dying?

[deHarro]

Hi Aubetex,

fix the PCB with screws. Then i'm create the height map. I think this is a good method.

I think that's ok.

If you are in doubt about the height map, you can visualize your design before you apply the map and afterwards, to get an idea of the correctness. In the first approach you can see the router pathes all in one level, overlaid with the colored height map. If your board is warped a lot, possibly some of the pathes are above, others are underneath the colored map. After applying the map the pathes should all "vanish" below the colored map and all have the same (negative) distance from the map. To see that clearly, just tilt the whole viewport and zoom in to show the design from the side. It's a bit difficult to get a good view, since the viewport is somewhat jumpy in great magnifications, but feasible.

Harald

[AuBeTeX]

Hi! I found, maybe, a problem with the program. I'm use on a Win10 64bit with a Chinese cheap CNC. Sometimes (cca 10~15%) when i'm connect to cnc, the program freeze. A lot. The task manager can't kill that. Only the reboot helps. If this is a problem, what need to resolve this? Logs? Files? Or simple drop my shitly cnc to trashcan?

A

[AuBeTeX]

Hi!

I'm search in the program a function . I don't know does exists. I'm created a PCB with SMD, and the PCB is a bit complex. According the program the work length 13 hours. In this 13 hours a lot things can happen. Voltage outage, tool breaking, cat attack, kid helping, termonuclear war, zombie apocalipse, and so on. If this things happen eg after 5 hours, the best of my knowledge i'm need to restart milling. If the program somehow record the progress with time code and visual confirmation, i'm reset the voltage outage, prevented the cat attack, kid's helping, termonuclear war, and zombie apocalypse, the broken tool is replaced, then the help with the fast forward, and the fast rewind buttons i'm searched the last good position. After that i overstarted the work from that position. In this case i'm don't need to 5 hours wait to nothing, because that part of the work already is done. In this 5 hours the CNC milling the air. I'm older 5 hours, and in this useless time the zombies incites the cat to push button to nuclear war. And the kid always helpful.

In this wersion have something similar?

Thanks

A

[silicon-systems]

Dear All Newbies,

Very important to configure your GRBL values setting, if not you might get very strange results.

Each CNC tool will require a range of values for successful milling. In my instance the CNC is CNC3020Z-DQ comes with parallel port and Mach3 software.

Stepper drivers are DRV8825 with forced air cooling, but yet to install new TMC2209.

DRV8825 setup $0=25 uS $1=255 mS $2=0 mask $3=1 mask $4=0 boolean $5=0 boolean $6=0 boolean $10=2 mask $11=0.01 mm $12=0.002 mm $13=0 boolean $20=0 boolean $21=0 boolean $22=0 boolean $23=0 boolean $24=25 mm/min $25=60 mm/min $26=250 mm/min $27=1 mm $30=12000 RPM $31=0 RPM $32=0 boolean $100=400 step/mm $101=400 step/mm $102=400 step/mm $110=150 mm/min $111=150 mm/min $112=70 mm/min $120=100 mm/sec^2 $121=100 mm/sec^2 $122=60 mm/sec^2 $130=250 mm $131=300 mm $132=45 mm

Hope above GRBL setting could be your starting point, you may tweak values for faster milling speed.

Milling bit tip size is 0.1mm or 100um you're safe with 20° - 30° angles. For 10° - 15° slow down your milling speed else will break them. Remember to use double sided tape to hold PCB in place.

First do set your Z axis, use this Macro: Call it "Safe Probe 2mm"

1. [G38.2 Z-10 F120]
2. [G92Z0]
3. [$J=G91 F150 Z2]

You do this by clicking on "Add Macro" "New Macro" shows up, right click mouse & select "Edit" Next give your Macro a New Name. Cut and paste command in square brackets up there into big square box under "Edit Macro Item" Do I have to say click OK?

Each time you change your milling "V" or drilling bit click on Macro "Safe Probe 2mm".

Next Load your PCB G-code by clicking on "File" then "Open". Next step click on "Probing" and select "Create New". Select "Size From GCode", note "GridSize" 6 - 12 is just right, but more warped PCB is will require smaller "GridSize" value, similarly a more flatten PCB will require bigger "GridSize" value. Did I say click OK to accept "GridSize" value? Next click on "Run" to acquire PCB HeightMap. You may choose to save this Map for drilling or PCB cutout but I had never found use for it as replacing your milling bit for drill bit, you will have to acquire new probe and set zero your bit.

Next step is to APPLY your HeightMap to your PCB GCode by Clicking on "Edit" and select "HeightMap" when done, click on "File" and select "Start" to mill your PCB, that's you have been hoping for is it not?

Kind regards, Hanspeter.

[AuBeTeX]

Hi!

Many thanks the instructions. At last i think i'm found the perfect mode to milling. But now i'm facing an another problem. The milling is nice, and perfect depth. My error is the use of the screws. If i'm not fixing the pcb with screws, only few positioner rod, and use the two sided tape, the work is nice.

The facing problem... My tool is "hitting" the 0.1mm 10° v shape milling tool in x0y0 position i send down z-0.1 the expected result is a 0.1mm depth, 0.1mm diameter hole. But, the result is a 1mm diamerer circle.

Have any idea to correct this, or i need to order a new spindle with tool holder?

Many Thanks! A

[martin2250]

Hi @AuBeTeX

I don't quite understand your problem. Does the tool get deflected by the cutting forces to make such a large hole?

Martin

[AuBeTeX]

Hi!

My tol is look like not centered. Maybe my spindle's axis curved, or my chuck is the problem. Dunno. My english is poor, but, i try to explain.

10° 0.1mm V tool New, "now get out the box" new tool

g0 x0y0 m3s5000 g01z-0.1 g0z2

Thanks

AuBeTe

[silicon-systems]

Hi @ AuBeTeX

I believe you are asking two questions.

One you expect 0.1mm milling path width but you get 1mm path width. Note! not all Chinese milling V bits tips are what is printed on the box. Have had several experiences from eBay or Amazon where V bits ordered was not 0.1mm. It's your responsibility to check each and every bit size, I'm afraid we will not be able to do that for you.

Check your spindle runoff, you can search the meaning of 'runoff' on google. If your spindle isn't centered you should notice some wee vibrations, just reduce spindle speed and check any off centering of your spindle.

In any case the doughnut at top left of your picture does not suggest a bad runoff. You seem to be running a kind of G-code in the background as there is some movement in both X and Y axis.

You may post the specs. of your CNC milling machine or a picture.

Kindly do a translation from your native language into english on google and paste it here for our benefit.

Cheers, Hanspeter.