Hi, I want to build a 4-axis CNC hot wire foam cutter to cut RC airplane wing cores. Using the linear actuators from OpenBuilds seems to be the fastest and most convenient way to do this. I imported the OpenBuillds step files from GrabCAD and created a quick assembly in Fusion 360 to see how this could work. I will be honest with you, this design is highly inspired by the 2-gantry-cnc-hotwire-machine build. I really like this design, but I think using C-Beams is a bit overkill for the job and also increase significantly the overall cost. So this is where I need help: 1- Are theV-Slot NEMA 17 Linear Actuator Bundle (Lead Screw) the right choice. If not, why? Maybe it would be better to use the V-Slot NEMA 17 Linear Actuator Bundle (Belt Driven) instead (all axes)? Or maybe belts on the x-axis and lead screws on the y-axis? 2- What hardware is needed to connect 2 of the linear actuators together. This should be simple to resolve, but I have no experience with the OpenBuild ecosystem. I'm trying to avoid having to place multiple orders and to minimize delays and having a bunch of parts with no use in the end. Thanks in advance, Marco
No reason belts themselves shouldn't work in either direction. You might take a look at Hotwire cnc Surfboards | OpenBuilds as an example of a belted system. The only issue with belts is that in most of the designs, the gantry plates ride on the narrow edge of the extrusion which is not near as rigid as plates having a wider grip on the rails. You might need to put a spreader bar across the top to keep the system stable. As for connections, depends on the gantry plate. In most cases you can run bolts through the back side of the gantry plate into t-nuts in the channels in the back side of the rail.
This is true, due to the nature of the mini V gantry plate actuation system, though with a small amount of effort, a simple angle bracket can fix that without changing the orientation of the motors: Portable Diode Laser Cutter Would that work for a 4-axis hot wire, given that the "z-axes" need to move independently?
You are correct on that one. (Brain fart on my part.) I was just looking at the Cali surf board set up and thinking it must have a spreader at the top as there is way too much leverage on that system. That's kind of an apples to oranges solution. It would be easier to rotate the motors on a V-Slot NEMA 17 Linear Actuator Bundle (Belt Driven) - OpenBuilds Part Store based system 90 degrees to use the flat side and upgrade the extrusion to a 20x60 and use the full size plate. But this is not a stock solution and would require a far more detailed order. The true solution though lies in how large the OP wants the system to be. If its just for aircraft wings, the axes wouldn't need to be more than 500mm in the horizontal and 250 in the vertical and that would be inherently stable on its own. If it's bigger than that vertically, a second horizontal rail with a free riding gantry plate would solve the problem.
Exactly! I think OpenBuilds should offer a NEMA 17 bett driven linear actuator where the gantry rides on the large side of the extrusion. That being said, since I can't really make custom angled brackets, I'll have to stick to the lead screw actuator for the x-axis. On the other hand, I don't see why the belt driven actuator would not work for the y-axis. Plus it would save weight and they're cheaper. Aren't the screw heads will prevent having the y-axis extrusion being flat against the gantry? Or maybe the screw heads fit inside the channels of the extrusions? Thanks guys for your advices! I need to fiddle around a bit more with the concept, but I think this is working.
That's the dimensions I had in mind(x:500, y:250). I used the 500mm actuators in my 3D model for the y-axis because that's what was available on GrabCad. I should have mentioned this, but you totally interpreted the RC wing application right!
They don't offer it readily on the site but if you ask the shop, they're more than happy to put together custom kits. Here is a basic idea of what it would look like. If that's the route you'd like to take, just reference this thread. It's all a minor case of substitutions for them. Just give them a few days to provide a response as orders waiting to be filled generally have priority. This should give you a better idea. M5x8mm screw through the back side of the horizontal axis gantry plate threaded into a t-nut set in the vertical axis channel. And finally, a general overview of a combined system using basic configurations. It should be noted that the range of motion in this configuration will be approximately 14" horizontal and 4" vertical and the vertical range may be a bit limiting for what you seek to achieve. You may want to go with 500mm vertical actuators. Overall the price difference is negligible. If you want to keep the build low and only want the verticals to be say, 350mm, just put a comment on the order and the shop will be happy to cut the extrusions for you. Sketchup files attached for reference.
Thanks for all the time you put into helping me with my project @Rick 2.0!!! It's much appreciated. I really like your first drawing with the 20x60 extrusion where the gantry run along the wide side. I might just go for with this. That's exactly what I had in mind for mating the extrusion on the gantry. But I still have concerns about the wheel screws. According to the CAD files I have, you cannot fit the screw heads into the channels. In case I don't want to custom, using the lead screw actuators for x and belt driven actuators for y would also work well.
Are you referring to an interference issue between the screws for the center wheels and the vertical extrusion? If so, that's fairly easy to address. Just put a Slot Washer - 15x5x2mm - OpenBuilds Part Store on the screws you using to mount the plate to the channel between the plate and the channel as spacers. That holds the extrusion clear of the screw head. You'll need to up the screws to 10mm in length to accommodate. The screws for the wheels should be tightened and adjusted on the plate long before mounting the two assemblies together anyway. It'll work, just realize the two move at substantially different speeds. Belts move five times faster than screws per rotation of the motor. As far as asking for custom quotes, it's nothing to worry about. The people at the shop are happy to help. The goal of OpenBuilds is help people create the best, most useful systems they possibly can. If you're in a hurry and don't want to wait for a quote, just add a couple 500mm lengths of V-Slot 20x60 Linear Rail Aluminum Extrusion Profile (openbuildspartstore.com) and a couple V-Slot Gantry Plate - Universal - OpenBuilds Part Store to your cart and make the switch yourself. It will total out to less than upcharge for the screw actuators. You can use the leftover 20x40s for the support framing.
Only the lead screw actuators have screw heads in the middle to hold the anti-backlash nut that would prevent the extrusion to sit flush on the gantry. Sorry about the confusion, since this is not present on the belt driven setups. I didn't think about that. I mean speed is not an issue with a hot wire foam cutter. The speed of belts is not required at all - cutting too fast results in drag in the wire and throws off your profiles. I was thinking more in terms of accuracy. That means that a lead screw has about 5 times more accuracy that belts. The real question is how much accuracy do I need? Everything is a trade off... speed, accuracy, weight and cost. Can't have it all!
That's a non-issue then. The lead screw nut doesn't have to be attached to the center holes. There are 2 other sets of holes down the center of the plate that could be used to attach the lead screw nut.
I think this is how it's going to be: 2 x V-Slot NEMA 17 Linear Actuator Bundle (Belt Driven) - 500mm 2 x V-Slot NEMA 17 Linear Actuator Bundle (Belt Driven) - 250mm 2 x V-Slot Gantry Plate - Universal 8 x V-Slot 20x40 Linear Rail - 250mm Any red flag? What's the best way to mate 2 extrusion pieces together? T-nut + holes T-nut + Angle corners T-nut + L brackets Since I'm using 20x40 extrusions, I guess I need 2 corners or brackets per "corner"?
It seems you're learning fast regarding the OpenBuilds system. Fairly simple isn't it. Visually I see a need for 12 Cast Corner Bracket - OpenBuilds Part Store. 2 at the top of each vertical (on the sides of the vertical, screwed to the lower channel on the back of the horizontal) and 4 at the bottom (2 front, 2 back). You might get away with 2 at the bottom of each, but you might not. Hard to say. I generally shoot for the sure thing first try. The bottom of the legs could also be secured with a pair of T Joining Plate - OpenBuilds Part Store.
Alright, so I've placed my order to my Canadian OpenBuilds distributor, @Makerparts. While I'm waiting for the parts to arrive, I was wondering if anyone has recommendations on how to configure the blackbox for a 4-axis foam cutter? Oh, I also have the Interface, but I don't have to use it right away if it's easier via the OpenBuilds Control software. For example, I will need to use 4 limit switches, one per axis. The blackbox only has 3 inputs for limit switches. Can I reconfigure the probe input to accept a 4th limit switch? I'm sure I'll figure it out, but any tips are welcomed!
Our standard 4-axis firmware does that already Wizards and Tools menu in CONTROL > Firmware Flashing tool If our default setup doesn't quite suit you, you can customize your firmware completely by modifying pinmaps and configs and then recompiling: github.com/grblHAL/ESP32
@Peter Van Der Walt, connected everything and noticed that the 4-axis firmware is only available for the new BackBox X32 in OpenBuilds CONTROL v1.0.362. Is there a reason why? I don't think a 32-bit processor is required to handle a forth axis. Is there a way to compile the firmware for the BlackBox 4X and share it?
The BlackBox 4X does not support 4th axes no, as you were mentioning ordering parts and didn't mention 4X or X32 - I assumed it would be the newer X32 (available since Nov last year), sorry Grbl (8bit) is 3-axes only. See Home
Oh... I see. The limitation is the firmware, not the hardware. I was planing to "borrow" the blackbox 4X from my CNC mill. I guess that won't work. I'll try to repurpose an old 8-bit 3D printer board I have. The timing seems to be good - Marlin was recently forked to control coreXYUV machines (like a 4-xis CNC foam cutter)! GitHub - rcKeith/Marlin: CNC 4 axes Foam Cutter firmware forked from Marlin2ForPipetBot I guess if I loose patience or decide to go plug-and-play in the end, I'll get the new BlackBox X32!
The ATmeag328P 8 bit micro does not have the space either. Its both hardware and firmware. Your 3D printer board probably runs a ATmega2560 or larger though, so could work.
Just quick question then... say I get the BlackBox X32 for my 4-axis foam cutter: 1- Can I use the interface with a 4-axis machine? The graphical interface currently only shows 3 axes. 2- Can you swap the Interface to one BalckBox to another (assuming it works with the BlackBox X32 using the 4-axis firmware, or even just with another one running 30axis firmware (one day I'll get a LEAD1010))?
Not directly no. Interface predates the X32s existence, so doesn't cater to all the newest features. It will still stream 4axis jobs, but no jogging etc Yes its just a host. Machine settings are stored in the controller not the host
I got all my parts and put the machine together. I got the BlackBox X32 in the end. Now I'm trying to setup GRBL using the CONTROL software I have some issues for the 4th axis and I was wondering if someone could help me out. So far, the main issues I've encountered are: I don't know how to activate the limit switch connected to the probe port for the A-axis. I don't know how to reverse the stepper direction for the A-axis. Units for the A-axis are in "steps/deg" instead of steps/mm and deg/sec2 instead of mm/sec2. How do I home A-axis? The X32 also has WIFI and Bluetooth capabilities thanks to its built-in Espressif ESP32. Is there any ways this can be exploited (say wireless connection to controller)? Maybe to have a more flexible way to configure GRBL settings? Thanks, Marco
Standard configuration: "A" would be a rotary, rotations are in degrees Shouldn't need anything other than plugging it in Home All should Home A as well CONTROL includes connectivity over wifi, set up Wifi using USB first (Grbl Settings tab, scroll down to Wifi SSID and PSK ($74/$75) and Save. Press the reset button on the front of the BlackBox so it boots up fresh with the new network details. After its properly on the network, use the Scan button in CONTROL to find and connect to it. Currently the UI doesn't give quick access to do it, so you can either calculate the bitmask manually, or (easier) swop the wiring (If its wired 1-2-3-4 reverse it so it becomes 4-3-2-1)
Something is wrong with the limit switch connected to the probe port and I can't figure it out. Maybe the Xtension limit switch is defective? It the LED turn on like the other ones when it gets triggered. It does get detected properly in CONTROL though. I get "ALARM: 1 - Hard limit triggered. Machine position is likely lost due to sudden and immediate halt. Re-homing is highly recommended. [ undefined ]" message. It just reappears if I try to clear it. If I hit cancel and home the machine, the Z, X and Y get homed, but not the A. Is it a bad limit switch, a bad X32 controller or just the wrong settings? I attached my settings in case an expert eye can see something wrong/the source of my problem. Code: [16:21:50] [ connect ] PORT INFO: Port is now open: COM3 - Attempting to detect Firmware [16:21:50] [ connect ] Checking for firmware on COM3 [16:21:50] [ connect ] Detecting Firmware: Method 1 (Autoreset) [16:21:50] [ ] ok [16:21:51] [ connect ] Detecting Firmware: Method 2 (Ctrl+X) [16:21:51] [ ] GrblHAL 1.1f ['$' or '$HELP' for help] [16:21:51] [ ] [MSG:'$H'|'$X' to unlock] [16:21:53] [ $$ ] $0=10.0 ;Step pulse time, microseconds [16:21:53] [ $$ ] $1=255 ;Step idle delay, milliseconds [16:21:53] [ $$ ] $2=0 ;Step pulse invert, mask [16:21:53] [ $$ ] $3=4 ;Step direction invert, mask [16:21:53] [ $$ ] $4=0 ;Invert step enable pin, boolean [16:21:53] [ $$ ] $5=7 ;Invert limit pins, boolean/mask [16:21:53] [ $$ ] $6=1 ;Invert probe pin, boolean [16:21:53] [ $$ ] $7=0 ;Disable spindle with 0 speed, boolean [16:21:53] [ $$ ] $10=511 ;Status report options, mask [16:21:53] [ $$ ] $11=0.010 ;Junction deviation, millimeters [16:21:53] [ $$ ] $12=0.002 ;Arc tolerance, millimeters [16:21:53] [ $$ ] $13=0 ;Report in inches, boolean [16:21:53] [ $$ ] $14=0 ;Limit pins invert, mask [16:21:53] [ $$ ] $15=0 ;Coolant pins invert, mask [16:21:53] [ $$ ] $16=0 ;Spindle pins invert, mask [16:21:53] [ $$ ] $17=0 ;Control pins pullup disable, mask [16:21:53] [ $$ ] $18=0 ;Limit pins pullup disable, mask [16:21:53] [ $$ ] $19=0 ;Probe pin pullup disable, boolean [16:21:53] [ $$ ] $20=1 ;Soft limits enable, boolean [16:21:53] [ $$ ] $21=1 ;Hard limits enable, boolean [16:21:53] [ $$ ] $22=1 ;Homing cycle enable, boolean (Grbl) / mask (GrblHAL) [16:21:53] [ $$ ] $23=7 ;Homing direction invert, mask [16:21:53] [ $$ ] $24=100.0 ;Homing locate feed rate, mm/min [16:21:53] [ $$ ] $25=1000.0 ;Homing search seek rate, mm/min [16:21:53] [ $$ ] $26=250 ;Homing switch debounce delay, milliseconds [16:21:53] [ $$ ] $27=5.000 ;Homing switch pull-off distance, millimeters [16:21:53] [ $$ ] $28=0.100 ;G73 retract distance, in mm [16:21:53] [ $$ ] $29=0.0 ;Step pulse delay (ms) [16:21:53] [ $$ ] $30=1000.000 ;Maximum spindle speed, RPM [16:21:53] [ $$ ] $31=0.000 ;Minimum spindle speed, RPM [16:21:53] [ $$ ] $32=0 ;Laser-mode enable, boolean [16:21:53] [ $$ ] $33=5000.0 ;Spindle PWM frequency [16:21:53] [ $$ ] $34=0.0 ;Spindle off Value [16:21:53] [ $$ ] $35=0.0 ;Spindle min value [16:21:53] [ $$ ] $36=100.0 ;Spindle max value [16:21:53] [ $$ ] $37=0 ;Stepper deenergize mask [16:21:53] [ $$ ] $39=1 ;Enable printable realtime command characters, boolean [16:21:53] [ $$ ] $40=0 ;Apply soft limits for jog commands, boolean [16:21:53] [ $$ ] $43=1 ;Homing passes [16:21:53] [ $$ ] $44=4 ;Homing cycle 1 [16:21:53] [ $$ ] $45=3 ;Homing cycle 2 [16:21:53] [ $$ ] $46=0 ;Homing cycle 3 [16:21:53] [ $$ ] $47=0 ;Homing cycle 4 [16:21:53] [ $$ ] $62=0 ;Sleep Enable [16:21:53] [ $$ ] $63=2 ;Feed Hold Actions [16:21:53] [ $$ ] $64=0 ;Force Init Alarm [16:21:53] [ $$ ] $65=0 ;Require homing sequence to be executed at startup [16:21:53] [ $$ ] $70=7 ;Network Services [16:21:53] [ $$ ] $73=1 ;Wifi Mode [16:21:53] [ $$ ] $74=NA ;Wifi network SSID [16:21:53] [ $$ ] $75=NA ;Wifi network PSK [16:21:53] [ $$ ] $100=26.667 ;X-axis steps per millimeter [16:21:53] [ $$ ] $101=26.667 ;Y-axis steps per millimeter [16:21:53] [ $$ ] $102=26.667 ;Z-axis steps per millimeter [16:21:53] [ $$ ] $103=26.667 ;A-axis steps per degree [16:21:53] [ $$ ] $110=2500.000 ;X-axis maximum rate, mm/min [16:21:53] [ $$ ] $111=2500.000 ;Y-axis maximum rate, mm/min [16:21:53] [ $$ ] $112=2500.000 ;Z-axis maximum rate, mm/min [16:21:53] [ $$ ] $113=2500.000 ;A-axis maximum rate, deg/min [16:21:53] [ $$ ] $120=150.000 ;X-axis acceleration, mm/sec^2 [16:21:53] [ $$ ] $121=150.000 ;Y-axis acceleration, mm/sec^2 [16:21:53] [ $$ ] $122=150.000 ;Z-axis acceleration, mm/sec^2 [16:21:53] [ $$ ] $123=150.000 ;A-axis acceleration, deg/sec^2 [16:21:53] [ $$ ] $130=295.000 ;X-axis maximum travel, millimeters [16:21:53] [ $$ ] $131=80.000 ;Y-axis maximum travel, millimeters [16:21:53] [ $$ ] $132=80.000 ;Z-axis maximum travel, millimeters [16:21:53] [ $$ ] $133=295.000 ;A-axis maximum travel, degrees [16:21:53] [ $$ ] $300=Grbl ;Hostname [16:21:53] [ $$ ] $302=192.168.5.1 ;IP Address [16:21:53] [ $$ ] $303=192.168.5.1 ;Gateway [16:21:53] [ $$ ] $304=255.255.255.0 ;Netmask [16:21:53] [ $$ ] $305=23 ;Telnet Port [16:21:53] [ $$ ] $306=80 ;HTTP Port [16:21:53] [ $$ ] $307=81 ;Websocket Port [16:21:53] [ $$ ] $341=0 ;Tool Change Mode [16:21:53] [ $$ ] $342=30.0 ;Tool Change probing distance [16:21:53] [ $$ ] $343=25.0 ;Tool Change Locate Feed rate [16:21:53] [ $$ ] $344=200.0 ;Tool Change Search Seek rate [16:21:53] [ $$ ] $345=100.0 ;Tool Change Probe Pull Off rate [16:21:53] [ $$ ] $370=0 ;Invert I/O Port Inputs (mask) [16:21:53] [ $$ ] $376=0 ;unknown [16:21:53] [ $$ ] $384=0 ;Disable G92 Persistence [16:21:53] [ $$ ] ok [16:21:53] [ $I ] [VER:1.1f.20220325:custom] [16:21:53] [ $I ] [OPT:VNSL,35,1024,4,0] [16:21:53] [ $I ] [NEWOPT:ENUMS,RT+,HOME,TC,SED,WIFI,SD] [16:21:53] [ $I ] [FIRMWARE:grblHAL] [16:21:53] [ $I ] [NVS STORAGE:*FLASH] [16:21:53] [ $I ] [DRIVER:ESP32] [16:21:53] [ $I ] [DRIVER VERSION:220327] [16:21:53] [ $I ] [DRIVER OPTIONS:v4.3.2-555-gb4ef1b80a2-dirty] [16:21:53] [ $I ] [BOARD:BlackBox X32] [16:21:53] [ $I ] [AUX IO:1,0,0,0] [16:21:53] [ $I ] [WIFI MAC:NA] [16:21:53] [ $I ] [IP:NA] [16:21:53] [ $I ] [PLUGIN:SDCARD v1.05] [16:21:53] [ $I ] [PLUGIN:ESP32 WebUI v0.03] [16:21:53] [ $I ] ok [16:21:53] [ $G ] [GC:G0 G54 G17 G21 G90 G94 G49 G98 G50 M5 M9 T0 F0 S0.] [16:21:53] [ $G ] ok [16:21:54] [ connect ] Firmware Detected: grbl version 1.1f on COM3
Try updating to the latest firmware: - might be caused by WiFi \ Hard limit error · Issue #61 · grblHAL/ESP32 1) Go to http://svn.io-engineering.com:8080/?driver=ESP32&board=BlackBox X32 2) General Tab: Set number of Axes to 4 3) Plugins Tab: Our default firmware has SD card Enabled. 4) Networking tab: Our default firmware enables WebUI, Telnet, Websocket and HTTP server 5) Advanced tab: We don't change any defaults there Generate and download the firmware Extract the zip, and flash firmware.bin (ignore the other files) using CONTROL > Wizards and Tools > Firmware Flashing tool > BlackBox X32 > machine style: Custom firmware.bin May also need to tell the machine to also home A by setting $46=8 ($44-49 controls homing sequences, see Additional or extended settings ). You can customize the order things must home in there, defaults + $46=8 should do Z first, then XY together, then A
The latest software removed the alarm and $46=8 enables the a-axis homing. I had to invert the homing direction by using $23=15 (all my limit switches are at the closest position). I have one problem left to solve: the A limit switch (plugged into the probe port) gets triggered (I can see it in the troubleshooting tab), but it doesn't stop the stepper motor. It's like GRBL doesn't know that the probe limit switch is there to prevent the a-axis to crash. I went through the grblhal documentation and I cannot find any command that would allow me to modify that behaviour. Since the A-axis is typically a rotational axis, it doesn't need a limit switch. On the other hand, @Peter Van Der Walt wrote earlier that it should work as is... so is it software or hardware (remember, the LED on that switch don't turn on (even when connected to other ports)?
Will have to check for you when I get to the office later. Some other bitmask that needs to be enabled for A axis I think
If that might help, triggering the limit switch plugged into the probe port does not trigger an alarm like the other ones.
I think I found the issue. We only defined probe as limit for Autosquare configs Gotta update ESP32/BlackBoxX32_map.h at 57e54a19dd4613415703aff5b51cad371120bd7f · grblHAL/ESP32 to also define it for 4th axes configs. Logged issue with grblHAL Enable Hard Limits for A-axis with BlackBox X32 · Issue #67 · grblHAL/ESP32 Good catch, thanks for report!
