The BEST DIY PCB Tutorial (ForOurGood PCB Method)

If you want to know the best way to make your own DIY PCB's, well you have stumbled on the right video...
This is a special episode, spun off from my original series, where I challenged myself to make the best possible DIY PCB’s using commonly available low cost CNC technologies. (Spoiler, I succeeded!)

[@ForOurGood]

This is a pinned comment where I will try to provide extra information in the replies of this comment...

[@ForOurGood]

====================== [Single Sided PCB Process] ====================== 1. UV Resin Coat PCB - If PCB surface has notable corrosion or scratches, use Scotch Brite to create wire brush finish first - Use metal polish (e.g. Pikal or Brasso) to polish to a high shine finish - Using tissue paper, clean thoroughly with IPA and dry with (maintain dust free from here) - Apply UV Resin and use small spatula to push into the entire surface first - Use a piece of Mono thread #150 silkscreen mesh to level surface - Remove silkscreen mesh, INSPECT and REDO as required (minimize bubbles or dust particles) - ALLOW TO SETTLE for 2 minutes (keep covered to avoid dust!) - Expose with UV lamp for 30 minutes (WNB UV-W100 Green UV resin) > Note: Store the silkscreen mesh in a dark & dust free place so it can be used again in Step#5 > Note: Silkscreen mesh can also be washed in IPA and reused, but the cost/benefit should be considered case by case > Note: If you are working with a large PCB, then consider putting the mesh in a frame or mesh to make handling easier 2. Affix PCB Locating Guides - Make 4 "L" shaped PCB corner brackets from PCB offcuts - Cover PCB corner brackets with masking tape, leaving PCB contact surface bare - Cover PCB working area with masking tape with generous margins - Clamp working PCB firmly in place, centered to the masked area - While applying pressure to corner bracket, add CA glue to outside edge - Apply CA glue accelerator to lock corner bracket in place - Repeat above 2 steps for all 4 corners - Leave for 5 minutes to allow CA glue to fully set - Remove PCB - Lightly clean CA glue accelerator residue from the PCB with IPA - Replace PCB and check for movement. (Redo if not a positive/firm fit) - Clamp working PCB firmly in place. > Note: A different locating method may be used, as long as it allows for accurate re-installation of the PCB during the job 3. Create Location Reference - Make locating marker from UV resin coated PCB offcut (min 10W x 40H) - Determine location of locating marker, close but allows PCB to be flipped. - Make sure edges and shape is square for both location marker, and PCB. Adjust with file as need be. - Apply masking tape to cover area where locating marker will be glued (min 25W x 50H) - Place locating marker in the center of the masking tape, oriented square to CNC - While applying pressure to locating marker, add CA glue to margin - Apply CA glue accelerator to lock locating marker in place - Leave for 5 minutes to allow CA glue to fully set - Lightly clean marker with IPA - Home CNC & Set XY to zero - Probe Z-Height of work PCB surface and set laser to correct focus height - Locate laser relative to lower left corner of locating marker X+3mm Y+3mm - TAKE NOTE OF OFFSET COORDINATES (will refer to as "LM_OFFSET" hereon) - Set X & Y coordinates to zero - Laser burn reference markings onto surface (use "Laser_Make_Reference.nc") - Move the laser to zero X&Y point for the work PCB - TAKE NOTE OF OFFSET COORDINATES (will refer to as "PCB_OFFSET" hereon) - Set X & Y coordinates to zero 4. Laser Clear Etch Area & Etching - Prepare Ferric Chloride by warming in boiled water bath - Load G-Code file "??-?_Cu.gbr_sub_edit_iso_combined_cnc.nc" - Run G-code to laser clear area to be etched - Remove PCB and lightly wash in IPA using fine hair toothbrush, then dry. - Place PCB in Ferric Chloride, continually agitate, occasional brushing. - At 5 minutes observer carefully until all visible copper has been removed - Wash in water, and check with microscope or zoom camera for residual copper - Repeat etching in 30>45 second intervals until etched enough (do not over etch!) - Clean with IPA and fine toothbrush, then dry with dabbing tissue paper 5. UV Resin Top Coat & Print PCB Silkscreen - Repeat Step#1 procedures to apply top coat (Use silkscreen cutoff for small area) - Wash thoroughly with IPA and dry with tissue paper (maintain dust free from here) - Apply UV Resin and use small spatula to push into surface (Fast UV curing resin!) - Use a piece of Mono thread #150 silkscreen mesh to level surface Note: Use as little UV resin as possible and create a thin even coat - Remove silkscreen mesh, INSPECT and REDO as required (minimize bubbles or dust particles) - ALLOW TO SETTLE for 2 minutes (keep covered to avoid dust!) - Use hotplate to heat PCB for 10-minutes @80degC then cool to 50degC Note: Finish should look flat and somewhat dull after this process - Mount PCB in CNC and clamp firmly in place. - Load G-Code file "??-?_Silkscreen.gbr_iso_combined_mt_paint_cnc.nc" - Run G-code to expose the PCB Silkscreen pattern - Remove PCB from CNC - Use IPA soaked tissue to dab off majority of unrequired resin - Scrub with IPA and fine toothbrush to reveal the PCB Silkscreen details Note: Important to scrub pretty hard to test adhesions, dont be scared to scrub Note: If the some or all of the PCB fall off, redo with higher power settings - Final clean with IPA and dry with tissue paper - Expose with UV lamp for 1 minute (Mechanic UVH900-WY White UV resin) 6. Laser Clear Pads, Laser Clear Outline & Laser Cut PCB @45% depth - Mount PCB in CNC and clamp firmly in place. - Load G-Code file "??-?_Mask.gbr_sub_iso_combined_mt_paint_cnc.nc" - Run G-code to laser clear the pads while using STRONG AIR ASSIST for cooling pads Note: If air-assist is not available, then use the Gerbal editor in Flatcam to split up the pad clearing operation into 2 or more files, with pads having smaller thermal capacity, using lower power levels. (avoids burning PCB substrate) - Load G-Code file "??-User_8.gbr_iso_combined_mt_paint_cnc.nc" - Run G-code to laser clear the PCB edge outline (NO air assist needed from here) - Load G-Code file "??-User_9.gbr_iso_combined_cnc.nc" - Run G-code to laser cut the PCB - Remove PCB and scrub the pads clean using IPA and fine toothbrush, then dry. 7. Install & Align Spindle, Setup PCB for Drilling - Remove laser and install spindle - Home CNC & Set XY to zero - Install 1mm drill in the spindle - Probe Z-Height of locating marker - Move CNC to LM_OFFSET X&Y coordinates and then move Z to G0Z0.3 - Adjust XY position by eye to be centered with location cross mark - Set X & Y coordinates to zero - Move CNC to coordinates G0X0Y4 - Set X & Y coordinates to zero a - Drill test holes in locating marker (use "Drill_Align_Test.nc") b - Magnify and inspect marks and determine X & Y adjustment (COMPENSATE INSIDE HOLE DISTANCE!) c - Move to X0Y4 adding the determined adjustment amount for X & Y (typically +/- 0.01>0.1) d - Set X & Y coordinates to zero Note: If the drilled hole has an ugly burr, consider changing to a new 1mm drill. - Repeat steps "a" thru "d" until good alignment achieved - Move CNC to coordinates to G0X0Y0 then G0Z0.3 - Move CNC by 1mm steps until centered with location cross mark - Set X & Y coordinates to zero - Move CNC to coordinates to G0Z15 to ensure the drill will be clear - Move CNC to PCB_OFFSET X&Y coordinates - Set X & Y coordinates to zero - Re-install PCB with 0.2mm waste sheet underneath, ready for drilling - Probe Z-Height of work PCB 8. Drill Alignment Holes - Load G-Code file "Alignment Drills_cnc.nc" - Run G-code to drill 1mm alignment holes in work PCB - Use a small drill to bevel the copper edges (REMOVE ANY COPPER BURR!) 9. Drill PCB Through Holes - Ensure all drills to be used have collet stoppers adjusted to same height - Load G-Code file "??-PTH.drl_cnc.nc" - Install first drill in sequence in the spindle - Probe Z-Height of work PCB (if not already done) - Move the spindle to the location of the first hole at a Z height 0.2mm - Inspect drill location, if noticable error exists,then adjust the origin to compensate - Run G-code to drill holes in work PCB, changing drill bit as per the drill sequence Note: Alignment holes are made into a copper surface, but PCB holes will drill into softer resin Blunt drills and/or Z axis play, can cause the holes drilled in copper to shift position This likely will result in the PCB holes being drilled slightly off center due to alignment error For consisent errors, consider adding an adjustment amount to the origin before drilling PCB holes For inconsistent errors, don't substract holes from CU & Mask layers in Flatcam prep (i.e. don't etch holes)

[@ForOurGood]

10. Install & Align Laser - Remove spindle and install laser - Home CNC & Set XY to zero - Probe Z-Height of locating marker and set laser to correct focus height - Move CNC to LM_OFFSET X&Y coordinates - Adjust XY position by eye to be centered with location cross mark - Set X & Y coordinates to zero - Move CNC to coordinates G0X0Y4 - Set X & Y coordinates to zero a - Burn test marks in locating marker (use "Laser_Align_Test.nc") b - Magnify and inspect marks and determine X & Y adjustment (COMPENSATE OUTSIDE CIRCLE DISTANCE!) c - Move to G0X0Y4 adding the determined adjustment amount for X & Y (typically +/- 0.01>0.1) d - Set X & Y coordinates to zero - Repeat steps "a" thru "d" until good alignment achieved - Move CNC to coordinates to G0X0Y0 - Move CNC by 1mm steps until centered with location cross mark - Set X & Y coordinates to zero - Move CNC to coordinates to G0Z10 to ensure the laser will be clear - Move CNC to PCB_OFFSET X&Y coordinates - Set X & Y coordinates to zero 11. Burn Alignment Marks, Flip & Align PCB - Remove PCB, remove waste sheet, and re-install PCB without waste sheet - Check alignment holes are all clean and clear - Probe Z-Height of MOUNTING surface (i.e. waste board) - Load G-Code file "Alignment Drills_conv_edit_iso_combined_cnc.nc" - Run G-code to burn alignment marks into the masking tape on the mounting surface - Remove PCB - Remove Locating Guides, being careful not to disturb the alignment marks - Flip PCB backside up, and using magnification, align alignment holes and marks - Clamp PCB firmly in place - Probe Z-Height of PCB surface 12. Back Side PCB Markings & Laser Cut - Load G-Code file "??-B_Silkscreen.gbr_iso_combined_mt_paint_cnc.nc" - Run G-code to mark the bare P - For symmetric PCB -> Load G-Code File "??-User_9.gbr_iso_combined_cnc.nc" - For asymmetric PCB -> Load G-Code File "??-User_9.gbr_1_iso_combined_cnc.nc" - Run G-code to laser cut the PCB - Masking tape, locating guides and reference can be removed 13. Laser Cut SMD Solder Paste Stencil (If SMD is used in design) - Mount bare copper clad PCB in CNC as working surface - Apply masking tape to the bare PCB over area larger than the stencil - Load G-Code File "??-F_Paste.gbr_edit_iso_combined_cnc.nc" - Set the XY origin of the CNC so the stencil will be burnt at the center of the masking tape - Cut out 2 pieces of white weak adhesive label sheet large enough for the stencil - Carefully apply the FIRST layer of label sheet to the center of the masking tape - Use Scotch Brite to dull surface of the FIRST layer of label sheet, clean lightly with IPA - Probe Z-Height of the FIRST layer of label sheet - Run G-code to cut the stencil design into the FIRST layer of label sheet - Now stick the SECOND piece of label sheet carefully on top of the FIRST layer - Use Scotch Brite to dull surface of the SECOND layer of label sheet, clean lightly with IPA - Run G-code to cut the stencil design into the SECOND layer of label sheet - Carefully remove COMBINED layers from masking tape, leaving cutout areas on the masking tape - REDUCE STENCIL ADHESION by pressing sticky side to your fingers a few times (need only minimal adhesion) 14. Prepare Pads for Soldering - Magnify and inspect, using a pin tip to remove any obvious resin remaining on pads - Use fine toothbrush and metal polish (Pikal/Brasso) to bring pads to a high shine > When polishing, focus on scrubbing inside the pad cavities cut into the resin > Rotate the PCB and polish a couple of times from different angles to ensure all areas covered > Take care with this step, it's CRITICAL to remove residue from pads to get good solder take up - Scrub the pads clean using IPA and fine toothbrush, then dry. - Magnify and inspect, confirming all pads are highly reflective. Repeat polishing as needed > Note: If the PCB was polished well in Step#1, then getting a good shine on the pads should be easy. 15. SMD Assembly Process (If SMD is used in design) - Affix the stencil to the PCB so it is accurately aligned with the pads, then press down firmly - Magnify and inspect stencil, ensuring all holes are clear, use pin tip to clear as needed - Apply solder paste into stencil cavities with spatula, press in hard, scrape of excess - Carefully remove Paste Stencil so the paste remains in the PCB undisturbed - Place all SMD components onto PCB - Place PCB on hotplate and PRE-HEAT to just below solder paste melting point, hold for 2min > for 138degC solder paste this is approximately 125degC > for 183degC solder paste this is approximately 145degC - Raise the hotplate to the REFLOW temperature for the solder paste, hold for 1min > for 138degC solder paste this is approximately between 150degC > 150degC > for 183degC solder paste this is approximately between 205degC > 225degC - Nudge any out of place components - Carefully remove from hotplate and allow to cool naturally (do not cool quickly) - Clean with IPA and fine toothbrush, then dry. - Inspect with magnification, remove any unwanted solder beads & further clean as need be 16. Snap Out PCB & Clean Edges - Repeatedly bend waste part of PCB around laser cut edges of PCB, and snap it out - Use fine file to clean up outside edge of PCB to align cleanly with laser burns - Clean with IPA and fine toothbrush, then dry. 17. Install & Solder Through Hole Parts (If thru-hole components used in design) - Install, solder and trim wires of any through hole components - Clean with IPA and fine toothbrush, then dry. ====================== [Double Sided PCB Process] ====================== a) Step 1: UV Resin coat BOTH sides of the PCB (One side at a time of course...) b) Steps 2 thru 10, Skipping Step-9: Alignment & Front Side Process (Skip drilling PCB Holes) c) Steps 11 & Step 2 Combined: Flip PCB, Align and add new Locating Guides for back side processing d) Steps 4 thru 9, Skipping Step-8: Back Side Process* & Drill PCB Holes (Skip drilling Locating Holes again) e) Steps 13 thru 17: Assemble PCB (Note: Manually solder on back side components as need be) * When etching the back side of the PCB, make sure to protect front side with tape or adhesive sheet

[@ForOurGood]

=================== [ KiCad Layers & Creating Files ] =================== @@@@@ Single Sided PCB @@@@@ F.Cu Clear resist UV resin F.Paste Create solder paste stencil + F.Silkscreen Expose UV Resin for PCB silkscreen + B.Silkscreen Burn into board the PCB silkscreen +* F.Mask Clear mask resin from PCB pads User8 Clear mask resin from PCB edge User9 PCB Shape, Cutout PCB ** PTH-drl PCB Through-Holes (Gerber & Excellon) + Optional * Mirrored in FlatCam ** If PCB not vertical line symmetric, make mirrored copy in FlatCam @@@@@ Double Sided PCB @@@@@ F.Cu Clear resist UV resin B.Cu Clear resist UV resin * F.Paste Create solder paste stencil + F.Silkscreen Expose UV Resin for PCB silkscreen + B.Silkscreen Expose UV Resin for PCB silkscreen +* F.Mask Clear mask resin from PCB pads B.Mask Clear mask resin from PCB pads * User8 Clear mask resin from PCB edge ** User9 PCB Shape, Cutout PCB ** PTH-drl PCB Through-Holes (Gerber & Excellon) ++ + Optional ++ Make mirrored copy in FlatCam * Mirrored in FlatCam ** If PCB not vertical line symmetric, make mirrored copy in FlatCam @@@@@ Creating Output Files in KiCad @@@@@ - Select Menu > File > Fabrication Outputs > Gerbers(.gbr)... - Plot format = "Gerber" - Select above layers - [Plot] - [Generate Drill Files...] - Drill File Format = Excellon - [Generate Drill Files] - Drill File Format = Gerber X2 - [Generate Drill Files]

[@ForOurGood]

============= [ FlatCAM 8.994 File Preparation Information ] ============= Note: Offset adjustment and PCB tiling is possible within FlatCAM, but is not covered here. *** Laser Remove Mask Resin for Etching Design *** - File > Open > Open Gerber... > Select "??-F_Cu.gbr" & "??-PTH-drl.gbr" > [Open] - Subtract Tool (ALT+W) (Use to remove pad holes) - Delete source = YES, Target = "??-F_Cu.gbr", Subtractor = "??-PTH-drl.gbr" - [Subtract Gerber] - Double-Click "??-F_Cu.gbr_sub" - [Gerber Editor] (Use for cleanup & etch compensation) - Inspect carefully for any defects or hidden unrequired track pieces - Select and delete any leftover track pieces (or fix in KiCad & Reload) - Select all remaining elements - Click [B] icon in toolbar - Buffer distance 0.022 - [B Buffer] - [Exit Editor] > [Yes] - [Isolating Routing] - Tool Table 0.075, Tool Dia 0.075, Passes 19, Overlap 33.3333% (Increase "Passes" to fit design) - [Generate Geometry] - Check Geometry has all tracks properly isolated. Modify KiCad design to resolve any issues. - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.075, Tool Dia 0.075, V-Tip Angle 1, Feedrate XY 650, Laser Power 5000 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Laser Expose Fast Mask Resin for PCB Front Side Silkscreen *** - File > Open > Open Gerber... > Select "??-F_Silkscreen.gbr" > [Open] - Double-Click "??-F_Silkscreen.gbr" - [Isolating Routing] (Use to generate outline use with paint tool) - Tool Table 0.0001, Tool Dia 0.0001, Passes 1, Overlap 0% - [Generate Geometry] - [Paint Tool] - Tool Table 0.1, Tool Dia 0.1, Overlap 0%, Offset 0.01, Method = "Standard", Connect = NO Contour = NO - [Generate Geometry] - Click [Project] Tab - Double-Click "??-F_Silkscreen.gbr_iso_combined_mt_paint" - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.1, V-Tip Dia 0.1, V-Tip Angle 1, Feedrate XY 200, Laser Power 650 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Laser Remove Mask Resin from Pads *** AIR ASSIST REQUIRED *** - File > Open > Open Gerber... > Select "??-F_Mask.gbr" & "??-PTH-drl.gbr" > [Open] - Subtract Tool (ALT+W) (Use to remove pad holes) - Delete source = YES, Target = "??-F_Mask.gbr", Subtractor = "??-PTH-drl.gbr" - [Subtract Gerber] - Double-Click "??-F_Mask.gbr_sub" - [Isolating Routing] (Use to generate outline use with paint tool) - Tool Table 0.0001, Tool Dia 0.0001, Passes 1, Overlap 0% - [Generate Geometry] - [Paint Tool] - Tool Table 0.075, Tool Dia 0.075, Overlap 33%, Offset 0.0128, Method = "Lines", Connect = NO Contour = Yes - [Generate Geometry] - Click [Project] Tab - Double-Click "??-F_Mask.gbr_sub_iso_combined_mt_paint" - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.075, V-Tip Dia 0.075, V-Tip Angle 1, Feedrate XY 650, Laser Power 4400 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Laser Remove Mask Resin from PCB EDGES *** - File > Open > Open Gerber... > Select "??-User_8.gbr" > [Open] - Double-Click "??-User_8.gbr" - [Isolating Routing] (Use to generate outline use with paint tool) - Tool Table 0.0001, Tool Dia 0.0001, Passes 1, Overlap 0% - [Generate Geometry] - [Paint Tool] - Tool Table 0.075, Tool Dia 0.075, Overlap 33%, Offset 0.038, Method = "Standard", Connect = Yes Contour = Yes - [Generate Geometry] - Click [Project] Tab - Double-Click "??-User_8.gbr_iso_combined_mt_paint" - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.075, V-Tip Dia 0.075, V-Tip Angle 1, Feedrate XY 650, Laser Power 2000 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Laser Cut Front Side of PCB *** NOTE: Symmetrical PCB's can use the same G-Code file for both the front and back - File > Open > Open Gerber... > Select "??-User_9.gbr" > [Open] - Double-Click "??-User_9.gbr" - [Isolating Routing] - Tool Table 0.1, Tool Dia 0.1, Passes 22, Overlap 99.9999% - [Generate Geometry] - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.1, V-Tip Dia 0.1, V-Tip Angle 1, Feedrate XY 650, Laser Power 10000 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Drill Alignment Holes *** - Right-Click "??-User_9.gbr" > Enable Plot - Tool > 2-Sided PCB (ALT+D) - Select "??-User_9.gbr" - [Calculate Bound Values] (if does not generate values, then disable then enable plot, and try again) - PCB Alignment: Drill Dia 1.000, Axis = Y - Click the desired location of the lower-left hole and press [+Add] (manually modify value if needed) - Click the desired location of the upper-left hole and press [+Add] (manually modify value if needed) - [Create Excellon Object] - Click [Project] Tab - Double-Click "Alignment Drills" - [Drilling Tool] - CutZ -1.9, TravelZ 1.5, FeedZ 300, Spindle 10000 - ToolChange = NO, EndMove Z 25, EndMoveXY "0,0", PreProcessor = "GRBL_11_no_M6" - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) *** Drill Pad Holes *** - File > Open > Open Excellon... > Select "??-PTH.drl" > [Open] - Double-Click "??-PTH.drl" - [Drilling Tool] - CutZ -1.9, TravelZ 1.5, FeedZ 300, Spindle 10000 - ToolChange = YES @25, EndMove Z 25, EndMoveXY "0,0", PreProcessor = "GRBL_11_no_M6" - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) *** Laser Burn Alignment Marks *** - Click "Alignment Drills" (don't enable plot!) - Edit > Conversion > Convert Any to Gerber - Double-Click "Alignment Drills_conv" - [Gerber Editor] (Use to modify mark size) - Change Aperture Size from 1.0 to 0.8 - [Exit Editor] > [Yes] - [Isolation Routing] - Tool Table 0.0001, Tool Dia 0.0001, Passes 1, Overlap 0% - [Generate Geometry] - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.075, Tool Dia 0.075, V-Tip Angle 1, Feedrate XY 650, Laser Power 3000 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (saftey step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Laser Burn for PCB Back Side Silkscreen *** - File > Open > Open Gerber... > Select "??-B_Silkscreen.gbr" > [Open] - Right-Click "??-User_9.gbr" > Enable Plot - Tool > 2-Sided PCB (ALT+D) - Select "??-User_9.gbr" - [Calculate Bound Values] (if does not generate values, then disable then enable plot, and try again) - Select "??-B_Silkscreen.gbr" - Mirror Operation: Axis = Y - [Mirror] - Click [Project] Tab - Double-Click "??-B_Silkscreen.gbr" - [Isolating Routing] (Use to generate outline use with paint tool) - Tool Table 0.0001, Tool Dia 0.0001, Passes 1, Overlap 0% - [Generate Geometry] - [Paint Tool] - Tool Table 0.075, Tool Dia 0.075, Overlap 20%, Offset 0.00, Method = "Standard", Connect = NO Contour = Yes - [Generate Geometry] - Click [Project] Tab - Double-Click "??-B_Silkscreen.gbr_iso_combined_mt_paint" - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.075, V-Tip Dia 0.075, V-Tip Angle 1, Feedrate XY 650, Laser Power 6000 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT

[@ForOurGood]

*** Laser Cut Back Side of PCB *** NOTE: Symmetrical PCB's can use the same G-Code file for both the front and back - File > Open > Open Gerber... > Select "??-User_9.gbr" > [Open] (Opens second instance) - Right-Click "??-User_9.gbr" > Enable Plot - Tool > 2-Sided PCB (ALT+D) - Select "??-User_9.gbr" - [Calculate Bound Values] (if does not generate values, then disable then enable plot, and try again) - Select "??-User_9.gbr_1" - Mirror Operation: Axis = Y - [Mirror] - Click [Project] Tab - Double-Click "??-User_9.gbr_1" - [Isolating Routing] - Tool Table 0.1, Tool Dia 0.1, Passes 22, Overlap 99.9999% - [Generate Geometry] - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.1, V-Tip Dia 0.1, V-Tip Angle 1, Feedrate XY 650, Laser Power 10000 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT *** Laser Cut Solder Paste Stencil *** - File > Open > Open Gerber... > Select "??-F_Paste.gbr" > [Open] - Double-Click "??-F_Paste.gbr" - [Gerber Editor] (Use for burn compensation) - Select all elements - Click [B] icon in toolbar - Buffer distance -0.04 - [B Buffer] - [Exit Editor] > [Yes] - Click [Project] Tab - Double-Click "??-F_Paste.gbr_edit" - [Isolating Routing] - Tool Table 0.0001, Tool Dia 0.0001, Passes 1, Overlap 0% - [Generate Geometry] - Preprocessor = "GRBL_Laser" (!!CHOOSE FIRST!!) - Tool Table 0.075, Tool Dia 0.075, V-Tip Angle 1, Feedrate XY 650, Laser Power 3900 - [Generate CNC Job Object] - [Save CNC Code] > [Save] - Click [Project] Tab > CTRL+A > Right-Click any item > Disable Plot (clears plot area > wait till complete) - CTRL+S or File>Save>Save Project... (safety step in case of crash) - MODIFY G-CODE AS PER INFO AT BOTTOM OF THIS DOCUMENT ##################################################################### ##### ##### ##### REPLACE TEXT FOR FLATCAM GENERATED G-CODE LASER FILES ##### ##### ##### ##################################################################### *** Replace all matches with nothing as follows *** ------------ G01 F60.00 G01 F650.00 (Check this F value and replace as need be) *** Replace remaining header with below *** ------------- G21 (Millimetere units mode) G90 (Absolute distance mode) G17 (Select normal XY Plane) G94 (Feedrate as units per minute mode) $32=1 (Enable laser mode - just in case!) M5 (Stop spindle/laser) G0 F1100 (Set the rapids speed to be fast) G0 Z7.0000 (Lift Z a bit) G0 X0.0000 Y0.0000 (Send X-Y to home) G0 Z0.0000 (Send Z to home) M0 (Wait user confirmation) G00 x?? y??

[@darthjeder]

A small ultrasonic cleaner (one for glasses and the like) might be able to stir the resin inside the syringe. It works in seconds for pigmented ink that would take "hours" of shaking by hand.

[@ForOurGood]

@@darthjeder I have one of those, I'll give it a try, thanks!

[@IkechukwuNnantaAnanaba]

I have not finished watching the video but let me start by appreciating the effort that you have put into this. Thank you

[@ForOurGood]

You are very welcome, thanks for the comment!

[@atahanerginer686]

Thank you very much for the in-depth video covering all the details. As a maker with experience in FDM printing and plotter machines, I’m interested in exploring DIY PCB making. I have a question regarding laser power. I have access to a 3018 CNC machine with 2.5W, 5.5W, and 15W diode lasers available. Which one would you recommend for this purpose? On a side note about alignment: there’s a system called ARMS (Auto Registration Mark System) used in plotter cutters with CCD cameras. It aligns roll media (e.g., 20cm x 50m) to a cutting knife. At regular intervals, it uses L- and T-shaped corner marks to align the cut vector with the media. My hobby-grade machine achieves about 0.1mm accuracy even on a 20x30cm grid. I believe a similar system could be implemented with a side camera attached to the tool head. Once again, thank you for sharing your knowledge and for inspiring makers like me to explore new areas. Your content is truly invaluable!

[@ForOurGood]

The LaserTree one I use is sold as 40W, but in reality, most people agree it is less than 10W. I don’t have any experience with other lasers, so I am not in a great position to make a recommendation, but maybe the 15W laser could be better at least when it comes to cutting the board, so I would at least try with that first. But if you find you have trouble dialling in the settings for the lower power settings due to lack of setting granularity, or alternatively the settings do not give consistent results, then I would switch to the 5.5W. I suspect the 2.5W one will also likely work fine, but cutting the PCB could be problematic. Camera based alignment has actually been on my mind for some time, I even bought a ESP32-CAM with that in mind. Obviously 0.1mm accuracy is just not going to cut it (is that a pun?), so I would likely use magnification to try to achieve 0.01mm or less. The mark up and alignment process shown in the video could then be fully automated, and I am pretty confident I could achieve the desired result. But I am also cognisant that adding too much hardware complexity or cost to my process will reduce accessibility, which is a core tripe of what I have been selling here. So, for now it is on the back burner, but still hopefully will happen at some point, even so, the manual aligning process works really really well, and once you get used to it, can be completed pretty quickly, and anyone can replicate it easily enough. You are very welcome for the video, and as you are clearly you are my target audience, I am always happy to hear it might have been of some use. Sadly, for the amount of effort that goes into the video, and considering I am showing a pretty good result compared to many other videos out there which get hundreds of thousands of views, my motivation ultimately takes a hit. And the primary reason I do KZbin videos is to get some motivation and purpose, so what can I say really, this is a problem I continue to fight to overcome. Thanks for taking the time to comment, and I wish you success with your PCB making experimentation.

[@atahanerginer686]

​@@ForOurGood I truly believe your short content has massive viral potential. Your videos are filled with hidden gems, including captivating and satisfying close-up shots from vendors, repair technicians, and content creators. Even without commentary, the visuals speak for themselves and create a uniquely engaging experience. The quality of your work far surpasses competitors, setting a new standard in this space. To maximize your reach, creating short content for newcomers and general audiences could be the key to going viral. Adding timestamps to long-format videos would further improve accessibility and keep viewers engaged, especially for those returning to replicate steps. Wishing you continued success and countless viral moments ahead!

[@atahanerginer686]

@@ForOurGood Looks like I made a couple of mistakes in my first comment. Sorry, English is not my first language. What I meant by 'I have access to a 3018 CNC machine with 2.5W, 5.5W, and 15W diode lasers available' is that I can get any of these three versions from the vendors here. Also, I misremembered the details about the plotter. It actually has an accuracy of 0.01mm when cutting and a 0.084mm return-to-position accuracy. Keep in mind, it’s a belt-driven machine with a +120cm belt (Siser Romeo). What impressed me the most is its skew correction using the ARMS system, aside from the impressive cut accuracy. Since it’s a roll-based system (up to 300m), it can cut continuously while maintaining 0.01mm accuracy, even accounting for media flex, angle misalignment, and uneven feed rates. The system achieves this by detecting L and T marks (3mm wide) and finding the center of the intersection using a camera. I’m not sure if there’s an open-source version of ARMS, but there is a camera-based alignment system for multiple tool heads called TAMV (Tool Aligning Machine Vision). It’s an automated system for aligning X-Y offsets and could potentially be used to automate tool changes between laser and router heads. As for bed leveling, I’ve tried nearly all the available technologies, and there’s a new standout called Eddy (Beacon). It’s incredibly fast and works like a LIDAR. Although it’s compute-intensive (I’m using a Raspberry Pi 4 with 4GB RAM and Klipper), the results are incredible, with over 300k samples on a mesh. I’m not sure if this level of precision is necessary, but it can easily compensate for bent or warped PCBs down to micron-level accuracy. My runner-up would be the Pinda probe (induction-based) since it’s also a no-contact leveling system. Lastly, I highly recommend checking out the KZbin video titled DIY PCB: Professional Results with CNC + Laser by Alvaro’s Workshop. He showcases two great ideas: combining the silkscreen and solder mask application in one step and selectively removing layers using varied laser power. He also demonstrates a fast method for electroplating. Before watching his video, I had planned to follow Levi Jahssen’s method with two tool heads for electroplating, but Alvaro’s approach seems more efficient. I hope these insights are helpful and spark some ideas for optimizing your workflow or future projects. Your work is already impressive, and with these additional techniques, I believe you could push the boundaries even further. Wishing you continued success and innovation!

[@ForOurGood]

@@atahanerginer686 If the 5.5W is the true optical power of the laser, and you are budget conscious, then the 5.5W one might be the better choice. If you are not confident that it is really 5.5W, or money is simply not a concern, then go with the 15W one first, and if does not work well, then buy the 5.5W one too. The less than 0.01mm alignment accuracy makes more sense for professional equipment, so honestly, I am not surprised. But I am also very confident that this could be replicated at the hobby level too (based on my experiences so far). Hopefully I will be getting around to trying this myself soon enough. Super high accuracy in Z-probe (or surface scanning) is simply not required with the laser process I am using, small Z-axis deviations really do not seem to have much effect on the beam shape that is working on the PCB. If you can achieve accuracy in the range of around 0.1mm it will fine for sure, anything worse than 0.5mm you “may” start to see issues. I have actually seen Alvaro’s videos at some point of developing my process (they are very good), as well as Wegstr videos and many others who undoubtably provided me various inspiration for the process I have come up with. Some of my processes are undoubtably borrowing from a mix of these videos, some of them are just coincidental inspiration, but what from where I could not really say because it is all a mental fog for me. In EP7 of my original PCB series, I actually do use a similar “burn & fill” process for the silkscreen, similar to what Alvaro demonstrates, and the results I got were amazing as can be seen by my surprise in the video.. I even saw another more famous KZbin maker “inventing” this process for himself in a very recent video. Unfortunately, none of us realised at the time of making these videos (some still don't realise to this day), that this process is actually very dangerous in some situations, and can destroy your PCB. If you want to find out more, then please see EP8 of my original PCB series. It is because of this experience I stopped using this process, and searched for a new solution, and came up with what I demonstrated in this video (which works really well and is safe!) As for the “hole plating problem”, honestly, I have not seen anyone come up with a solution for this that I really like, but I also keep thinking about this problem myself too. As I state in the video, radically increasing the resolution of my PCB’s has basically made this a low priority problem since over 90% of the PCB’s that I need to make can now be done on a single sided PCB, and using a single sided PCB also comes with some huge cost, time and application benefits. Even so, hopefully one day I will crack this nut too. BTW: I just saw a very old video called “PCB Plating without Chemicals”, that process actually looks fairly simply to replicate.. but I don’t know how the result would compare to real electroplating, or if the material used is widely available. If you do end up making some PCB’s, I hope you take the time to upload a couple of videos and show us your results. Good luck!

[@TankTurk]

I've been searching how to make a PCB using a diode laser for a couple of years now and you solved it with this video. I just did a test of each step in your video and it all worked out perfect. My next project is your UV light box. I have a very old, probably about 30 years old tube UV light box. I might just take the tubes out and fit in one like yours but I think I might go with the 100W version, it's only about £4 (GPB) more so might be worth it. Thank you very much, I am one happy old man. All the best to you, have a great Christmas and a great new year. PS: I did subscribe to your channel a while ago but didn't keep up so missed this episode when you posted it. 👍👍👍👍👍

[@ForOurGood]

@@TankTurk Well I am so happy to hear that this video was finally useful to someone in a comprehensive manner.. I had half given up on it. Thanks for taking the time to let me know, and hope you have a wonderful festive season too! BTW: in the end, how did you (re?)discover this video?

[@TankTurk]

@@ForOurGood, I was searching how to make a PCB with a Diode laser. After watching a few, your video popped up. I was a bit reluctant to watch it as it's almost 3 hours long but I am really glad I did. I still need to practice and dial in my laser but last nights tests turned out good so I have high hopes of producing some good looking PCBs, at least a lot better than what I have been producing till now. Thanks again

[@blabby102]

I'm so glad I found your channel. I'm an Australian living in Tokyo and I have a 3030 desktop CNC and a cheap and weak laser module. I'm currently trying to make PCBs with solder mask at home. I bought a silkscreen frame at Daiso and are about to try it out. Even though I mill the traces with the CNC instead of etching, your videos are so helpful. I would love to hear any more information you have regarding makers in Japan. I've found it's hard to get cheap materials like acrylic or metals for the CNC. Currently I get most of my materials and parts from Aliexpress.

[@ForOurGood]

@@blabby102 20 years ago I used to mainly get stuff from Akihabara, but now it's only useful for heatsink, cable ties and cable. Actually, there are a one or two new hobby component stores there now, but it's basically repackaged stuff from China, and since Ali-Express is cheaper, and delivery very fast, that's where I get most of my stuff from now too. RS Components, Mouser and Farnell are all good for getting IC's that are the real deal, but it's never all that cheap, but for some types of chips it's the only way to go (buy Western brand IC's from Ali-Express at your peril... 99% are fake and either barely work or don't work) For raw stock materials, originally I was getting stuff mainly from SuperViva Home, but actually what they have for machining stock is either not so great, or over priced. Tokyo Hands also used to have a good line up for stock, but they're even more expensive. I have also used some online Japanese stores to get aluminium stock in the past, places like Misumi, prices are not bad, but the shipping can be costly, so it's only practical if you need a heap of stuff. For smaller stock or plate materials, Amazon Japan is always a good place to check. Generally I always check Amazon for a wide range of stuff, sometimes it is surprising what they sell, and some good pricing too. However, I got lucky a year or so ago, and stumbled on a physical store that exclusively deals with selling stock materials, thier prices is very reasonable, and they can get almost anything, and as I can pick it up at the store there is no shipping charges, even for small orders. The stores name is ㈱西山商店, typically they only deal as a wholesaler selling to other companies, but even, so they have been very helpful with selling me what I need, and currently is my go to place for raw stock materials. Glad you found the video useful, good luck with your PCB making, and thanks for taking the time to leave a comment!

[@AndyJenkins999]

Yes outstanding as usual from F.O.G. Been looking forward to this since the teaser shots of the board, looks great. The silk screen is incredibly clean. I was thinking there may be something else to use for your stencil material but this clearly works for you so all good and seems to accept a good amount of paste with the two layer thickness. Thank you for the exceptionally detailed and time consuming to make video a great addition for makers. Ps I think you were generous most hobbyists I know go for the bodge wires on the first batch of board mistakes (after much swearing) never mind ordering second set haha

[@ForOurGood]

Ha Ha, yep, I reckon your probably right, most people will just live with their mistakes after the first order, it’s gotta be pretty disappointing😐. Personally, I could not stomach the stress of getting it wrong (or likely getting it wrong), and then having to live with it, which is part of the reason I have never taken the step of ordering a board from China myself 😊. Actually, even in a professional environment, I could rarely justify going more than one revision before needing to put a board into the required operation (my requirements were never mass production however, max 100 or so boards). The great thing about surface mount, is that with the help of the hotplate, it is super easy and fast to reuse components, so that definitely takes the sting out of the idea of having to redo a DIY PCB too. Using my new process, I honestly believe I would be inclined to do the necessary iterations to at least get the circuit design and track layout of the PCB correct. Thanks as always for the comment and support!

[@mitchellhw2006]

great video. I watched the entire thing in two sittings. On the solder stencil; I've had recent luck using .15mm plastic sheet that I spray painted matte black before the laser cut. Makes a larger, washable stencil that I can tape down when batching PCBs. Adding a few "mouse-bites" around the edges of oddly shaped PCBs makes them so much easier to remove from the parent material. Great video. I cant imagine how many hours you have in the production.

[@ForOurGood]

Yeah, sorry for the long video, but it seems tutorials do much better if they are not split up. Even so, I am glad you found it interesting. I did actually try using plastic sheet, but I could not get a clean cut... more of a melty twisty result... I didn't use matte black paint, but I did try a matt black paint pen if I remember correctly. The feature size also is a challenge, with a paper label I could barely eek out the 0.2mm width required for the QFN20 package, so I sort of came to the conclusion that trying to do that with plastic sheet would simply be futile. What sort of resolution are you getting? Do you know what type fo plastic the sheet is made of? Currently, I am usually only doing one or two PCB's with a given stencil, so it is not really a big deal if it is not reusable (of course that would be nice if it was reusable). What is more of a concern, is getting the correct thickness (currently good), nice clean edges (currently not great), smooth surface (currently not really), required resolution (currently passable) and relatively sturdy (currently sturdy enough for a couple of uses, but not really sturdy as such) In know the locating of the stencil, is doable without an adhesive backing, have seen this in another persons video, so I am not super worried about that aspect either way. Thanks for taking the time to comment and provide some info too!

[@ForOurGood]

Just noticed your "mouse-bites" point, not sure why I missed it first read, but I will try to keep that in mind, thanks! As for amount of hours in production.. yep, don't want to think about it much, especially because I already know the video will not really get that many views. But they say, about an hour per minute, is what it takes just to properly edit a KZbin video.... but that excludes everything else

[@mitchellhw2006]

@@ForOurGood The first plastic stencil I made was from the diffuser sheet from a cracked LCD monitor. Since then I have used old school transparency sheets made for laser printers. I found two lower power cuts are better than one high power. Also I cut the outline only and not the field area.

[@ForOurGood]

@@mitchellhw2006 Got it, thanks! I will see if I can find something similar and give it a go. I actually used to use those laser printer transparencies back in the 90's when I was making PCB's back then, I guess it makes sense that they are somewhat heat tolerant.

[@reginaldmitchell-w1o]

Awesome method. Nice results. What would you think of the first step being to drill a few alignment holes in the scrap part of the PCB. Then have alignment pins in your CNC waste board. Then when you place and remove the board, it always goes back to the same place. As well, keep the pins the same for every board that you make. That is.. the alignment pins are forever in the same known/fixed place. As they are in a know position, and hence the board will always be in a known position, you no longer have to do any alignments for each board, or each step when placing the board (laser, drilling, etc). You would do one calibration when initially placing the pins in their permanent positions (to essentially tell the CNC where the pins are located), but none after that.

[@ForOurGood]

What I am trying to demonstrate here is different way to achieve high precision PCB locating, that is easy to replicate, even with all different sizes of blank PCB stock. I have actually been using alignment pins right up until recently (see my earlier videos), including going to the effort of machining my own precision sized pins to ensure an exact fit, but it just does not work good enough for this level of high-resolution PCB, and that is why I came up with these new methods. The reality is, that it is extremely difficult, actually nearly impossible, to reliably get the repeat fitment accuracy of less than +/- 0.01mm, using the locating pin method. Thanks for the comment!

[@Multimen76]

Nice Job and fantastic research well done . I try make ploter using IR led because hawe regular raund dot but problem whot i hawe is mechanical movement in axis . I use GRBL board and small stepers belts GT2 and teoreticly resolution is 10 micron but when i try clean area betwen traces all lose aligment like stepers lose steeps but is not this because if i use onli contur its ok its like board dont respect position from gcode and use tolerance to make more round numbers ich movement incres this tloerance error and if my dot move every 0.03mm its more error in every mm but if i move 200mm axis move perfect at 200mm. Whot leadscrew you hawe in your router ? Probably i leaw my idea use belts and i use leadscrews with 2mm pitch or 4mm pitch . Whot you think

[@ForOurGood]

Thanks for your comment and query. I don't think you will get the required precision using belts, I am using basic leadscrews purchased from Ali-Express with 2mm pitch and 4mm lead, they work well with good precision up to speeds of 500mm/min. Please see other videos on my channel for more information.

[@trhosking]

Put a piece of scrap PCB under the other end of the clamp so that the clamp is parallel to the machine bed when you tighten the clamp screws. That will provide a better clamping area on the board, i.e. flat against the surface of the board and not just along the edge. I know lasers don't impart any force on the workpiece, but it's just good practice.

[@ForOurGood]

@trhosking the clamps are long, thick and cut with a slot to ensure the clamping force is relatively flat (see episode 2 for more info). Possibly the slots could be a little deeper, but overall it does the job good enough for this application. Your point however is very much valid, the way the clamping force is applied is important! Thanks for the comment.

[@billybob3811]

In KiCAD you can control the export and plot origin by placing a "Drill/Place File Origin" mark wherever you want the (0,0) origin to be on your board. Simply select Place->Drill/Place File Origin from the Place menu and then left click where you want the origin to be. Then when you export/plot make sure the appropriate tick box is checked to force the output to be referenced to that (0,0) point.

[@ForOurGood]

I did mess around with that for a while, but failed to get it to work. It's more than likely that I checked the wrong box or something. Next time I make a PCB, I will be sure to give it another try. Thanks!

[@billybob3811]

@@ForOurGood First of all, and I should of said this before, but thank you for all the effort you put into your method and for documenting the process as well as your journey to get there. The end result is very impressive and is nearly indistinguishable from purchased boards. I personally made my first pcb over 40 years ago using pre-sensitize coper clad boards and photocopies of artwork on overhead transparency papers while exposing them to sunlight. Since then, I've tried lots of different methods, including toner transfer, resist pen plotting, and isolation milling. All to varying degrees of success, and at this point I've all but given in to just using fast turnaround Chinese vendors. Your method has rekindled my desire to produce my own pcbs for first article prototyping and same day iteration testing, as such I would be most interested in learning the optical power of your laser module. I believe this is just the excuse that I needed to justify adding a laser to my CNC machine. With regards to the origin issue, the tic box you are looking for is in the upper right quadrant of the plot dialog and is labeled "Use drill/place file origin".

[@ForOurGood]

Wow, 40 years ago, I did my first probably in 1991, though I have probably had a 15 or so year hiatus before my latest attempts were started, either way, you have got me beat! I am glad to hear that the video might actually stimulate someone to get back into the hobby, so thanks for the comment! The laser I am using is a “LaserTree Gold 450nm 40W-Pro 12V 20mm Fixed Focus Laser”, apparently it puts out less than 10 watts of optical power in reality. I bought mine a couple of years back on Ali-Express, so it is likely this exact model is not sold now, but I am sure you will find something similar, though I do recommend you buy one with an air assist nozzle included, even if you don’t plan to use it immediately.

[@Cruvs7]

Always watching every video, great work you should have a ton of subscribes!

[@ForOurGood]

Thanks for watching & the nice comment! 😉 You are clearly one of the lucky few to have somehow connected my channel, because I don’t get much love from KZbin, that is for sure... Clearly I am doing something wrong.... 😐 And just for a laugh, KZbin just yesterday recommended me to watch a 3-year old video on how to make PCBs, this is a 10 minute video which shows how to make PCB’s just like I was making them back in 1993.. it has 1.4million views... so who ever said life is fair!

[@robindebreuil]

Very nice, I’ll have to try the silkscreen, looks great. I have an old scratch milling video with a similar title, hopefully it doesn’t break KZbin :). Yours looks way better of course, though the scratch milling is one step (only milling). Need to think about the silkscreening for that process, great tips. Love the dedication to the process you’ve shown, and I have an inkling of what all that involves!

[@robindebreuil]

Btw, you should submit this to hackaday if you haven’t, always pcb makers lurking there :)

[@ForOurGood]

The title is unlikely to help much, at best my videos top out at about 20K after a few years, worst cast a couple of hundred views. Basically, KZbin does not promote my videos in any sensible way, I laugh when I see my videos continuously recommended on things like Polish Opera videos, then KZbin Studio gives advice that the poor click rate is because my thumbnails suck... yeah right.. The conspiracy theorist in me makes me want to say this is because I disabled ads for all my videos, but KZbin says its policy is that this makes no difference, so it likely is because my videos are not so great after all. Anyway, impressions for this video have already dropped to zero after 5 days. (Rant over). BTW, I originally got interested again in making PCB’s when I saw a Wegstr video, and was fully expecting to be milling PCB’s, and in my second episode made a sincere effort to do just that. But milling PCB’s on my 3018 just was just not doing it for me, so I quickly decided to try a different path. Again, likely due to some bias, I really like this process, and at least comes to etching an amazing result, it’s just so damn reliable and stupid proof. Once you take the mechanical interaction out between the CNC and the PCB, it totally transforms what a cheap CNC is capable of delivering. As a contrast, a Wegstr PCB mill is selling for $3.7K shipped, for me that does not seem like the typical DIY’er spend, maybe it is ok for some people. Let me know how you go with the silkscreen, and thanks for the comment!

[@ForOurGood]

@@robindebreuil Thanks for the advice, but I am not so good at self-promotion, I just throw it out there and see what happens. I don’t even bother doing the obligatory beg for likes or subscriptions in my videos anymore.

[@robindebreuil]

@@ForOurGood Ha yeah, right there with you, promotion is unpleasant work. The Wegstr machine seems amazing, but yeah, different thing. I’m working towards a small dedicated machine for simple PCBs. There are so many ways to remove copper, it makes for an interesting obsession. Very inspired by your dedication to the craft here, thanks for sharing it in such a detailed manner.

[@EnverTorlakovic]

Great effort @ForOurGood, thank you.

[@hilfsbereit1437]

Great video. What power does the laser have? Which items ( paste ) do you use (green and white)? is green Solder mask? Sorry for my bad englisch ;-))) Thank you

[@ForOurGood]

The laser I am using is a LaserTree Gold 450nm 40W-Pro 12V 20mm Fixed Focus Laser, apparently the optical power is in reality less than 10 watts, but still that works fine for me. I do show all the items I use in this video, and also many of my earlier videos too, please check them out for the all details. Thanks for leaving a comment!

[@PeetHobby]

Very interesting process. I've been making my own PCBs for the past 20 years using UV reactive paint and an exposure mask made with clear overhead projector sheets. I typically use bare copper print, but if I need some protection, I use clear PCB varnish or even nail polish. Maybe it's time for an upgrade after 20 years. 😋

[@ForOurGood]

Give it a try! I guarantee you will be surprised by the results. Thanks for the comment!

[@silicium_lab]

Excellent video and even better results! Congrats

[@ForOurGood]

I hope you and your fellow engineering students can now solve your pad clearing problems, good luck and thanks for the comment!

[@arthureschner9319]

Fantastic work! The boards came out really good. Thanks for taking the time to share this with us. What power is your laser? Is there any reason why you didn't cut out the pcb when you engrave the components on the back of the board?

[@ForOurGood]

The laser I am using is a “LaserTree Gold 450nm 40W-Pro 12V 20mm Fixed Focus Laser”, apparently it puts out less than 10 watts of optical power in reality. I highly recommend getting one with an air assist nozzle. Sorry, I am not sure I understand your question regarding the PCB cutout. The laser is not powerful enough to cut through the central glass fibers of the board, which is why I need to laser cut on both sides of the board. I show cutting the front of the board at timestamp (1:49:40), and the rear of the board at timestamp (2:08:25). The final step is to snap it out, breaking the central glass fibers, as shown in timestamp (2:35:10) Thanks for the comment!

[@arthureschner9319]

@@ForOurGood Flip, sorry I somehow missed you cutting the outline on back of the board. Thats what I was asking about. Thanks for the info. I have a creality 10W (with air assist) upgrade kit fitted to my ender 3 v2 neo. So hopefully, following your steps I should be able to acheive the same. I thought my laser might be too weak to do this. Thanks again.

[@ForOurGood]

@@arthureschner9319 no worries, the video is pretty long, so not surprising if people miss bits here and there. Anyway, sounds like your laser will do the job, so good luck with that 👍

[@Segphalt]

What are we considering a decent power laser? I have a 3w 405nm laser that I can use to engrave things I suspect that is enough?

[@ForOurGood]

@@Segphalt if that's a real 3W optical power laser, it should be enough to remove the resin. Maybe a little weak to cut the board, but you could do that a different way.

[@snrsmh]

Hello, thank you for the video, it would be nice to have a low-budget version of this series.

[@ForOurGood]

The process is actually relatively cheap, considering the end results that can be achieved. If you are smart with the purchases, I expect you could get started for around about $300USD. You also need to keep in mind that the CNC is a multi-functional device, and you can use it for things beyond just making simple PCB’s. Thanks for the comment!

[@yaroslavbaranovskiy3170]

Great job, thanks a lot for doing such long and detailed video with no ad-crap! Just a side note: From my experience, blue laser doesn't do really well on white. The darker the better! Even lighter vs darker masking tape makes a difference. And if I simply paint it with black sharpie - end result is way better. You were complaining on laser sometimes not being able to start burning glossy white solder mask film.. That should be exactly the case. I think simply switching to darker colors would save a lot of frustration + give you much larger margin for possible error. I.e. black for solder mask, something darker for silkscreen and dark red or black film for stencil. But yeah, you'll have to re-do all your speed/power settings :)

[@ForOurGood]

I am glad you liked the video, it seems that you are one of the few people who noticed I turned the ads off.. weird. I did a lot of testing in an earlier video in my series, check that out if you missed it, and thankfully found this green resin burns extremely reliably. But you are correct however, burning anything very shiny, glossy or translucent will not burn with this laser. In this case, using a dark coating or sanding the surface is the only way to get a reliable burn. I have not had trouble with burning normal white paper however, so I don't think the colour is the main problem, more so the reflectivity or transparency appears to be the culprit. I have not tested burning white UV resin yet, but I do plan to try to do that in the future. Thanks so much for taking the time to comment and leaving some suggestions, people sharing their experiences have helped me a lot with this project.

[@Cybernetic_Systems]

I’m only 17 mins into the video (great explanations), and I just wanted to discuss the board origin issue. Have you trying moving it in KiCad to the button left corner of the work area? I’ve never had any issues in FlatCam since doing this.

[@ForOurGood]

I tried for a while, and in the end gave up. Maybe I was doing it wrong or something. Another viewer also left a comment about how to do this, so I will give it another go next time!

[@ForOurGood]

Sorry, ignore what I said before. I can't remember if I specifically tried moving to the bottom left, I sort of think so. I did move it around a bit and tried a few things. I will try bottom left next time.

[@Cybernetic_Systems]

@@ForOurGood I'm not the sharpest tool in the shed, so you'll figure it out, I'm sure. That said, I haven't tried it in KiCad v8 yet, but in v7 it was straightforward.

[@Cybernetic_Systems]

@@ForOurGood Bottom left gives you a positive X and Y Axis, which makes things simpler for sure. :)

[@TinNguyen-uu2tr]

Video finally arrived. Hope to see you try with galvo laser one day. CNC laser has very low acceleration which make burning very uneven. I got exhausted with this process when burning all of my small pads so I quit and try with the mask instead. Probably high acceleration laser like galvo or at least coreXY klipper laser might have enough acceleration to encounter that

[@ForOurGood]

I can pretty conclusively say that acceleration does not play a notable role when using this process. I was in fact worried about that given the CNC3018 with its cheap lead screw accelerates pretty slowly. Undoubtably there is minor affect, but it is not noticeable from my experience. The key issues about clearing the resin from the pads without burning them off the board, comes down to a very simple calculation. That is how much “heat over time” you inject into the pad, and how much “heat over time” the pad can dissipate. If you add far more heat than the pad can dissipate in a given timeframe, then that will result in enough heat build up to eventually burn the substrate (And on top of all that, of course you’ll always need enough laser power to clear the resin). The best solution I have found is to use active cooling, such as air-assist, by doing that you can stop the pad from building up too much heat, meaning you can get away with a much higher power level that allows you to clear all pads with a single power setting. If you do not have air-assist, then the only other option is to break up the pads into different groups/jobs, applying an appropriate power level, depending on how well you expect them to be able to dissipate/cope with heat. Naturally this requires experience and good judgement to get this right, that is why I recommend air-assist as demonstrated in this video. Thanks for your comment and I hope you enjoy the video.

[@TinNguyen-uu2tr]

​@@ForOurGoodif so then that's very different from my result. I got full blast of air assist on all the time but can still see very clearly that the corners will get burned much more than s straight line. There are some settings to help even out the laser power but then it will reduce the power by a lot and still somehow doesn't get that consistency. There is one more thing is that if I don't align the laser well enough and some how burn into the resin area beneath then 100% guarantee that the pads will peel off. Those got me so frustrated with the laser so I developed my own process using printed light mask to cure the solder mask. I've been having consistent results since then and much simpler process and less tedious. Hopefully I can share it some days but I'm quite lazy with making videos and busy with out projects also. But I'm very glad that you succeed with your process. I hope more people will get into this so someone will try to fix FlatCAM. 😅

[@ForOurGood]

@@TinNguyen-uu2tr If the acceleration is that slow, then there must be a problem with your settings of your CNC (mine is 500mm/sec/sec for all axis) Also, using good FlatCAM settings is pretty critical, and that's why I discussed in detail in this video. Finally, maintaining perfect alignment is also critical, which is why I came up with the new alignment method which is working very good (alignment pins just don't provide enough precision). Finally, you can't have the laser set too hot, there is a limit to how much air assist can help. In any case, I am glad you came up with a process that is working for you, I am very much looking forward to your video 👍😁

[@retu191057]

Is it possible with fiber laser? I can clear mask okay. But what about sillkmask?

[@ForOurGood]

@@retu191057 Actually, I don't know for sure (since I don't own one), but If the laser wavelength is close enough to what the UV resin requires to cure, and you can throttle the laser power low enough, then I don't see any reason as to why it wouldn't work. Give it a try and let us all know! Thanks for the question.

[@zaprodk]

Amazing tutorial, I'm glad you are highlighting the importance of the correct version of FlatCAM etc. as it wouldn't be obvious for someone new that they are very different. Also what's going on with the wiggly track snaking under R8? :D

[@ForOurGood]

Yep, the devil's in the details as they say. Some people will think video goes into too much detail, but I know from experience, that it is hard to put this puzzle together when your starting from zero. BTW: The track is squiggly to tune it to an exact length because.. nah.. just kidding, actually the resistor is at a weird angle and I wanted to leave room for the ground plane to sneak through.. sadly I am still a nube when it comes to KiCad, and could not figure out how to freely set the angle of the track to match with the resistor.. Do you know how?? Thanks for taking the time to comment.

[@zaprodk]

Why don't you install the laser and spindle at the same time so the offset between them is constant, then it's a lot less alignment.

[@ForOurGood]

Originally when I first started this series I did give it serious thought, but did not follow through in the end. Without a doubt, it would save some time assuming the relationship would reliably remain precisely locked. But, likely just adjusting the height of the spindle in the mount will knock this out of whack (I use the CNC for various things, not just making PCB's). Physically having them both mounted at the same time will also likely affect the physics of the CNC3018, and that may affect the precision, and naturally the usable range will be reduced, and then we also have a risk of precision errors in the lead screw affecting the result (as we would be working in different areas of the lead screw for laser vs spindle) Phew.. actually there were more concerns than I remembered, some of them likely are not a concern at all, but only testing would confirm that. Anyway, I really like the idea of making the reference marker for a given job, so I would be making that in any case. This provides a way to precisely recover the position should your controller lose its marbles or something else that causes you to lose position. It also provides a way to check with confidence that your alignment is correct, before you start doing the actual work.. so peace of mind and all that. Actually the alignment process is not the difficult part, really it is physically fiddling around, swapping between laser and spindle, that would be the primary benefit of having them co-mounted. Sorry for the waffling! 😁

[@Cybernetic_Systems]

I've been mulling over your process and how I could apply various aspects to my own half-baked one (it's similar to your old process), and I was wondering if you'd experimented with using your "Silkscreen process" on the main PCB mask? In theory, it would make cleaning the pads much quicker and easier with much less elbow grease needed - the only downside I can think of is the laser curing process would take a lot longer. Btw, Unexpected maker (another Aussie) has a really nice "reflow master" kit that makes creating your own reflow oven really simple. I made one some time ago - it has so many additional uses with the "bake" mode, plus the ability to run proper solder reflow temp curves.

[@ForOurGood]

After seeing the great results of the silkscreen, I also had various ideas about how I could apply this for different purposes, maybe even trying to set the positive etch pattern with it. Setting the resin around the pads should also be plausible. Naturally, I would need to switch to the 10sec Mechanic resin for everything, and of course this is going to mean much much longer jobs (although it may be possible for it to be faster if the resin is not a couple of years old, like mine is). Then there is the question as to if the result would be cleaner using an expose method vs the burn method, and how reliable/consistent that might be given the expose method is a little more finicky, compared to burn method that just works and gives a consistent result every single time. For a silkscreen, +/- 10% thickness variation does not make a big difference, just as long as you can get it to stick, but a for a etch resist or pad clearing operation, this could be a real problem when working with small stuff. I also have other ideas about ways to clear the pads by laying something down in advance that could easily be removed later, along with any resin on top, but this has not progress beyond the thinking stage at this time, I generally think this might be a non-starter. For now, I am super happy with the results I am getting with the current process, and I probably make it look harder to do than it actually is, this is mainly because I had a lot of bad experiences before finally getting to the point where it actually works well. I still have a few improvements to my process in mind, but mostly minor tweaks to materials, tools and methods. Even so, I think at some point I will experiment with the resin expose method more, at least to test my assumptions, and see if this might actually be something practical/possible/worthwhile to do more with. As for the oven, I did actually think about doing something like that for this video, but in the end, I deemed it not a pressing concern, that old Panasonic clothes iron is doing what I need it to do for now. Maybe one day!

[@Cybernetic_Systems]

@@ForOurGood Cool, I figured you'd likely at least thought about it. Btw, what power is your laser? I've got a dinky little 5w laser that is essentially useless for anything other than marking things or exposing resin, and I really need to find a better one.

[@ForOurGood]

The laser I am using is a “LaserTree Gold 450nm 40W-Pro 12V 20mm Fixed Focus Laser”, apparently it puts out less than 10 watts of optical power in reality. I highly recommend getting one with an air assist nozzle.

[@Cybernetic_Systems]

@@ForOurGood Cheers, I'll have a look for a similar unit.

[@boumsiedyoucef1064]

Very good result, thanks

[@ForOurGood]

Glad you liked it! Thanks for the comment!

[@toddfisher8248]

Thank you! The big thing I want to figure out next is plating vias with copper

[@ForOurGood]

Vias vias vias.. sigh.. every one wants an easy solution for this.. not at all feeling confident with this part of the problem, but I am still thinking about it. You never know 😉 Thanks for the comment!

[@ShopperPlug]

Excellent PCB tutorial

[@ForOurGood]

Thanks so much 🙏

[@laszloszell8753]

I always wanted to try make my own PCB,my problem that I never did learn about it,but I always know what I want and how I want at basics level. I mean some examples: I want to make a led panel I know I want to use 12v dc source and 3v leds. Or I wanted to make last time a mop robot. I can do the coding and I know all the accessories I need with ardunio, but obviously is to much wire,so I thought make a beautiful pcb board in Fushion 360 with connectors and everything. My problem thst I don't know when to use and which type ,diode,resistors,capacitors, transistors etcetera. I was trying to looking for a pcb design program like at led panel example I give the input volt 12v and the amount of led I want to use and the program help me out what I have to use between them to don't get overcurrent etcetera.

[@ForOurGood]

There are plenty of resources out there if you want to learn electronic. You don’t need to learn absolutely everything to be able to start making your own PCB’s. Start small, learn the very basics, and do some testing to see if what you learnt makes sense with the results you get. Each small piece of knowledge is a building block you can use to improve your skills as you go, and in this way, it is very similar to learning writing code. I started writing code at the age of 12, but I did not start doing anything with electronics until age 17, but learning code helps you with structural thinking, and that helped allot when learning electronics. Good luck!

[@civismesecret]

I'm just starting electronics, this look perfect and promising.

[@Cybernetic_Systems]

lol, it’s taken me a quite while to get through this video, so much excellent information and great ideas. Regarding the ability to super easily to the pads, have you tried tinning them with liquid tin? I use MG Chemicals Liquid Tin - it’s not cheap, but lasts a really long time. I tin my board before I do anything else. Btw I get it from eBay here in Australia.

[@ForOurGood]

Just in case you are not taking the piss, and simply missed the section, I do talk about this at 2:24:30 Glad you liked the video! 😁 I am feeling pretty buggered after making that one.

[@Cybernetic_Systems]

@ForOurGood I managed to comment just before I got to you mentioning liquid tin, so that’s on me for jumping the gun. 🤦🏼‍♂️ no piss was harmed in the making of my comment. 😜

[@Cybernetic_Systems]

@ForOurGood I’m not surprised you are exhausted after that movie length video, I felt tired after making my latest 15 minute video. 🤦🏼‍♂️

[@Cybernetic_Systems]

@@ForOurGood I just got to the end, your silk screen imo looks better than the manufactured PCB silk screen. Iirc the cheaper boards use inkjet printing.

[@Error404_Handle]

how well are the solder joints? cause resin looks to be quite thick...

[@ForOurGood]

There are some pretty close up shots at the end, so you can judge for yourself. From what I am seeing, they look and work very well. I have also been using similar boards, however made with my older process, for a number of years now without a problem. Thanks for the question!

[@Error404_Handle]

@@ForOurGood good to hear. the result looks quite good, it just looks like there's a very high layer of resin. But if it works out, it works out :D I just use black spray paint to cover the board and use the same 3018 with the laser it came with. But i only use it for emergency boards... so i'm not very picky :D

[@ForOurGood]

@@Error404_Handle I think the magnification method I use makes it seem thicker than it really is. Laying down the resin with the silkscreen results in a super thin layer in reality, and even with 2 layers it is still pretty thin.

[@Error404_Handle]

@@ForOurGood that and maybe the sharp angle that's left after the laser burns the resin off. Definitely could be the reason why it looks so thick. But great job either way.

[@witawat]

you can share the chemicals you use? Where can I buy them?

[@ForOurGood]

@@witawat I think I cover this is this video, as well as earlier videos. The only "chemicals" I use is generic 99% IPA as well as generic Ferric Chloride. This stuff is typically produced and branded locally in each individual country, so the local Japanese brand I am currently using is likely not useful information for the bulk of viewers (I don't even remember the brand I am currently using). Typically just buy this stuff from Amazon or other similar sites. I am not sure if that answers your question, so if you need more information then could you be a little more specific? Thanks for the question!

[@witawat]

UV Color and 1:20:50 What TURBO ?

[@ForOurGood]

@witawat Green "WNB" UV resin and White "Mechanic" both purchased from AliExpress. The Turbo is a simple CA glue (super glue) accelerator, it is a local Japanese brand, but you should be able to find something similar locally.

[@TheRainHarvester]

It seems like a silk screen might only be bragging rights: if you're creating your own boards, you have enough to know what every component is!! Great process though! Thank you!

[@ForOurGood]

Glad you liked it, and thanks for taking the time to leave the comment.

[@toddfisher8248]

You can speed up the curing of the UV resin with hot air. I've been finding that my resin can cure in less then a minute with hot air...

[@ForOurGood]

Seems I am a little late to the game with regards to pre-baking UV resin before exposure, but it does not exactly seem to be common practice either. Actually, just yesterday I noticed on an Ali-Express store with the Mechanic UV resin, a set on instructions indicating both pre & post baking as a part of the process (the instructions are really not that clear however, and they talk about baking times of 50~60mins for each, which seems excessive). Unfortunately, I see no such similar instruction for the green WNB UV resin I am using. I also didn’t see a specific improvement in curing time with the Mechanic UV resin after baking, even though the viewer who commented indicated instant curing was achievable. The clear benefit I did get was a quality improvement, which in that case is the most critical thing anyway. It is possible that either my testing was not thorough enough, or my UV resin has just aged poorly (it is about 2 years old, and Mechanic UV resin is stated as a 12-month shelf life) Even so, I was already thinking to give pre-baking a try with my green WNB UV resin, so I will likely do that next time, and see how it affects the curing time (30mins is way too long). Don’t think I will be hitting with a heat gun however, that would likely disturb the finish, so I will probably try the hotplate, basically using a similar method to what I am already doing with the Mechanic UV resin. Would be nice to get the time to less than 15mins in total, otherwise it is probably not worth the extra messing around. Thanks for the comment!

[@trhosking]

You have been watching This Old Tony though.

[@teodulosoriano513]

wow creating via on the pcb