Hi all, My MPCNC’s about a year old now, and it’s worked flawlessly. All electronics were purchased new and from Ryan, save for the controller, which is a rambo 1.3L I had from an old 3d printer. I’ve encountered some funky control behaviour recently that I cannot figure out.
The Problem(s):
The problem appeared after I left the machine powered up in the idle state for a few days. The machine stops part way through a job. It does not return to zero position, the LCD does not show the ‘Turn off spindle’ directive as if the job finished, it’s just in the idle state like the board was reset. The selection knob button on the full graphic display also becomes very sensitive when this happens, often times registering a button press when left alone, sometimes as often as multiple times per second (the rotary encoder still works fine, it’s just the button). I believe there to be a communication issue as well, as the machine seems to be mixing up lines of gcode, moving to the right spot at the wrong point of the toolpath. I have tried using multiple sd cards, and the problem persists. I have tried running the exact same file at least 5 times, and every time it stops at a different location, and mixes up different parts of the gcode.
At first I thought the Full Graphic controller was bust, so I ordered one off of amazon only to find the problem persisted. I’ve tried multiple ribbon cables, and that has not helped either. I’ve examined the traces on the rambo LCD adaptor as well as the rambo board and see no obvious burn outs or damaged components, I’ve checked the LCD adaptor pins and can’t find any shorts.
Does this sound like a rambo problem? I have the tools and skills to swap components on the board if need be, but I can’t seem to find any damage. It is an old board(bought in 2014) and I’ve done some other repairs to it in it’s lifetime, but I don’t feel too excited about dishing out $250CDN+ for a new board so please let me know if there’s anything else I can check/ do to figure this out. The board has worked great the past year, and this problem appeared quite suddenly.
Thanks for the help, I plan on posting all of the projects I’ve accomplished soon, I just have to get the machine back up and running first to get back on schedule with my planned projects
Thanks for the reply, code is attached. It’s an adaptive clearing operation that was generated with the fusion360 post processor, feeds and speeds are within the realm of what I usually run at, and I’ve ran the code at 50% feed rate and the problem still persists.
I’m not sure if the feed rate is to blame, the 41mm/s is used for a rapid move at the beginning and end, all the cutting is done with 12.7mm/s or less. I’ve been using fusion 360 with the same settings since I built the machine and it’s had a stellar track record, I’ve checked my past successful posts and confirmed that they used arcs as well with no issues. If I can find time this weekend I’ll set up a post with Estlcam and see if the problem remains.
Thanks for the help Ryan, the amount of community support you give is top notch! I should have updates soon, whether or not I find the issue or just call it beat and see if a new controller solves it.
It sounds to me like a wiring gremlin. Something odd, like a pin losing a connection after thermal expansion or something. Leaving it idle for a few days shouldn’t affect it. The lcd knob is especially odd, which makes me think intermittent connection. I would try running a job from a computer as well, hoping it will print some error message when it fails.
That’s what I was thinking as well. It has gotten much colder here in the recent weeks, but I’ve run it reliably in similar temps with no issue. I’ll try it from the computer this weekend and see if it gets me an error message
So I think I’ve narrowed down the problem. It seems very odd though.
I found a good deal on a new rambo 1.3L so I bought it and flashed the identical firmware that’s on the other one on to it. That did not really change anything. The real problem centers around the encoder button behaviour.
After ~30 minutes of running, and progressively getting worse after that, any agitation to the EXP1 ribbon cable will cause the machine to register an encoder button hit. It starts out with things like bending the cable slightly, shaking it a bit, but after a 40 minute job even bringing my finger within ~1mm of the cable (no contact with the cable itself) will do it. This happens independent of what control board, graphic board, or even ribbon cable I use.
I took just the control board and graphic unit inside and ran the gcode job, and the issue only occurred once or twice when I was shaking the cable, that was after ~1 hour of running it.
So to me it looks like 4 possible things that could cause it: the power supply, the LCD to rambo adaptor board, the ambient temperature, or the geographical location of the board. The power supply seems ok, and I can’t find any shorts on the adaptor board, could this be some kind of weird EMF issue?
So I’ve talked to the electronics engineers at my work, as well as consulted with several other forums, and they seem to agree that this behaviour is due to the lack of an appropriate pullup resistor within the arduino mega/rambo design, although nobody was really able to address the circumstances surrounding it, i.e why it isn’t a more common problem/ why it’s only begun to happen now. I think I’m gonna try modding one of the ribbon cables I’ve accumulated to jump a couple 5K resistors between the vcc and encoder lines to see if that gets rid of the floating input behaviour. If that doesn’t work I’ll try new firmware. I’m also taking steps to insulate/heat up my garage to see if that helps. Will report back once I’ve accomplished that,
I haven’t seen this before, and the rambo is very well built, so I am not sure I agree with the pullup diagnosis. The microcontroller can be configured to have internal pullup resistors, so that may be why they are missing from the board. The LCD may also supposed to have them, which I am more suspect of.
That video is wild though. That definitely shouldn’t happen, and if it’s happening this easily with your setup, I would think it would happen at least sometimes with mine, or any of the other thousands out there.
Are you sure there’s nothing funny with the wall wiring ground? You’re not using a 2 prong AC cable or something, are you?
Another thing that can happen is the board power supply on one house circuit and the laptop/usb on another and it having issues.
But my guess would have been cable but it looks like you have tried swapping them. And you even tried swapping LCDs. Are you using the 1.3 lcd bridge or did you wire it yourself by hand?
I agree Jeff, that’s exactly what I was thinking, it’d be a way more common problem if it was the rambo design. I think it may be the AC feeding it, have to check my power supply and the extension cord and powerbar it plugs into. I’m using all 3 prong plugs, but maybe something happened.
What’s the GPS coordiantes of your machine and which way does it point? Feng Shui dictates that MPCNC machines must point East to work properly.
I do remember having issues with the LCD on my Ramps board if the I used too long of cables going to the LCD. I went back to shorter cables and the problem went away. I don’t remember what length I was trying as I threw the cables away after switching back to the shorter cables.
Good idea. The 23.5" cables I have are at the limit of stable. Some have used longer (we used to have a diy bundle), but even a few inches past 23.5 seems to usually cause issues.
