My simple design is here: http://www.thingiverse.com/thing:2018609 Short summary: A bracket replaces the IR probe circuit board. One additional wire needs to be run in the wiring harness ( for the servo control line ). Firmware changes to support the BLTOUCH are included in Marlin RC7 and RC8 source code ( editing the source file to un-comment the appropriate lines and then recompiling is required). IF you use RC7, there is a bug that requires a work-around (might also be in RC8...). The work-around involves making sure that the bed lowers fast enough, during Auto Bed Leveling, to avoid tripping the fault detection firmware in the BLTOUCH. I can find the thread in the Marlin discussion if someone wants more details.
Question? Why are the brown and black wires redundant? Also if you have your working RC7/RC8 code can I see it? Wait I think I see it!! So if you plug the Red, Black and Yellow into Z_MIN +,-,S pins respectively you do not require the other wires, also you can enable: //#define Z_MIN_PROBE_USES_Z_MIN_ENDSTOP_PIN in the firmware. Am I getting there?
You are really close! Both the brown and black wires coming from the BLTOUCH are connected to Ground on the BLTOUCH, so only connecting one of them to the Rumba board is required. This means that the three wires that connect to the existing IR probe connector, coming from the BLTOUCH, are the orange ( to red +), black OR brown (to black -) and white (to white S) wires. That leaves the yellow servo control wire, which I connected to pin 5 on the Rumba board (this is the default servo 0 pin for the firmware). Since there is not a wire already in the set of existing wires in the wiring harness that can be used, a new wire must be added to the wiring harness. That location on the Rumba board is covered with the EXP3 Dupont connector. As shipped by BigBox, there is not a wire in the Dupont connector location that corresponds to the pin 5 location, so I crimped one of the female Dupont pins provided in the BLTOUCH kit onto a long wire which I then inserted into the existing EXP3 Dupont connector ( pin is located at second row, 3rd from the right ). The other end of the long wire needs to be connected to the yellow servo control wire coming from the BLTOUCH, so I crimped a male Dupont pin to the other end of that long wire. I have Configuration .h files for both RC7 and RC8. I have not debugged RC8 yet. I will zip the files and post them later today or tomorrow.
I have both RC7 and RC8 working - I am going to stick with RC8 for now... Here are the customized files; I only made small changes on top of work other people have already done... In RC8, I have Auto Bed Leveling set up for a 6 x 4 set of points with bicubic interpolation. Also, check the steps per mm - it matches my setup. UPDATE: Make sure you edit "DEFAULT_AXIS_STEPS_PER_UNIT" to match the drivers you are using and the jumper positions on the Rumba board.
@W1EBR.Gene - What's your machine config? I'm trying to find RC7 or 8 firmware for a single hybrid Titan. I'm expecting a Prometheus dual extrusion system soon and also switching to the BLTouch probe.
My machine is a Dual Titan Hybrid, so the firmware might work for you as is. I have noticed that the auto bed leveling might not be getting applied in the bugfix release of RC8 that I compiled, so it makes sense to recompile it with the latest bugfix release and then watch the z axis as something is printed.
Hmmm... I tried the rcbugfix rc8 firmware with your configuration and pins modifications. Machine homes in x and y but then heads of the build plate in the upper right corner and rams into the frame. Not sure what I missed. Anyone know where I should look in the firmware... Like where Z homing location is specified.
I'm sorry, but I don't understand your description of the problem you are having ("heads of the build plate in the upper right corner and rams into the frame"). A picture would help. Also, are you sure that the BLtouch is connected to the Rumba the same way as I did? I wonder if the z limit switch is being tripped.
After homing x and y, the end effector travels all the way from x=0 y=0 to a position that's greater than x=300 and y is roughly at 100-150. It rams into the RH side of the top panel at that point and I have to switch power to stop it. There's a little more to the story as well. I haven't implemented the BLtouch yet, so I commented that portion out in your config. And I'm just running one hot end in the right hand side of the hybrid x carriage, so I set the config to one extruder and turned off the second temp sensor. I'm thinking I have something wrong in the bed leveling or z homing. Does your auto bed leveling start at x=270 y=185? Maybe my probe offset is wrong and the following variable is sending the x carriage off the bed in X: Code: #define RIGHT_PROBE_BED_POSITION 270 My configuration.h is attached.
I think I might know the problem. I changed #define DEFAULT_AXIS_STEPS_PER_UNIT { 320, 360, 400, 409.2 } because I am not using the original stepper motor drivers. If you are using them, then the steps per mm (of the x axis, I think) is off by 2x and what you described would happen (because there is no x,y,or z MAX endstops). Check the steps per mm for each stepper motor and make sure they match the drivers you are using and the jumper settings on the Rumba board that you have configured.
Yeah, that would definitely cause an issue! I changed my values and all is well now. The z adjustments during travel appear to be working, but I haven't done a test print yet. It's going to be SO NICE to have auto mesh bed leveling
I just learned that "M420 S1" needs to be added to the Scripts section so that bed leveling is actually turned on!
This must be the Marlin issue that G code addresses: https://github.com/MarlinFirmware/Marlin/issues/5869 Thanks for the heads up.
Well, as it turns out, the bilinear auto bed leveling was NOT working properly, but now the unified bed leveling is being merged into the RCBugfix branch, so I have hope for that. Right now I'm running the latest RCBugfix with manual mesh bed leveling. I spent a good amount of time messing around with UBL, but I think I'll let the dust settle on that for a few weeks.
Interesting... was the ABL moving the z-axis steppers at all for you? It does move them for me (but I haven't verified that what it was doing was correct)