I have an MPCNC from several years ago that gets used occasionally. While playing around with cutting speeds and depths, it recently started to make a high-pitched harmonic noise (from memory around 800-1400 Hz) while cutting MDF and produced cuts that were visibly wrong. Lines that were supposed to be straight had obvious wavy imperfections, for example. It also buzzed at medium to high speeds and became somewhat more difficult to move by hand when unpowered.
Trying to figure out the issue, I took apart the entire machine to see if any components were visibly broken. A few bearings were difficult to spin and the central spacer that does not have the nut for Z-axis engagement with the allthread was shattered, so I modeled and re-printed that.
After re-assembling, but not connecting the steppers to the controller, the X and Y axes were still difficult to move by hand, and the buzzing became much louder. The noise and resistance disappears completely when the two steppers are disconnected from each other. The linked video demonstrates this.
What steps should I take to troubleshoot this behavior?
Assuming in your video the steppers are engaged, you should not be able to move your router without putting in a lot more effort than you appear to be in this video. The sound is not the typical lost step buzzing either. Telling us about your build (including a photo of your control board) will help us direct your troubleshooting. I suspect either a wiring issue, or the current is set too low on the steppers, but there are other possibilities.
If your controller is off, then the motors are acting like a generator and that is normal. The rule of thumb is to not do that fast enough that the screen lights up. The logic is that the voltage would have to be much larger than that to do damage.
If the controller is on, then you shouldn’t be hot unolugging motor, is you can avoid it. But I have the same questions as robert.
The steppers are not plugged into the controller in the previous post, so they are definitely not engaged. The same sound occurs when the motors are plugged in and I send a move command with a feedrate of at least ~1800 mm/min.
The video example is set at 2000. I’ll note that the speed appears to be a bit off, as it takes about 12s to travel 350mm → ~1750.
The controller is set up as shown below. I believe it’s a Ramps V1.4, but I’m not certain. Firmware as reported by Repetier on startup is Marlin 510S 2.0.7.2 (Sep 5 2021). The steppers are wired in parallel. I’ve confirmed they’re wired the correct orientation relative to each other. I’ve also confirmed the cables are all unbroken and have continuity.
The recommended configuration is to wire the steppers in series. You will find a series wiring diagram on this page. If you leave them in a parallel configuration, I believe you have to up the current on the stepper drivers to compensate. I’m sure the next time Jeff checks this topic, he can give you precise values.
Another option is to bring all the stepper wires back to your Ramps 1.4 board, and give each stepper its own driver. You would need two more drivers (it looks like you are using DRV8825 drivers), and you will need to update the firmware to one of the dual ones. You don’t need endstops to run dual-endstop firmware.
For the DRV8825 drivers, the current is set manually by setting the vref voltage. Did you set the voltage?
I’ll note that the speed appears to be a bit off, as it takes about 12s to travel 350mm → ~1750.
When you get other things sorted, you want to check this again, but with acceleration and deceleration, and the rough “about,” I would not be concerned. What you really want to check is does the router move the precise distance you specified in your move command.
So my best guess based on your information is that with the parallel wiring, your steppers are not individually receiving enough current to generate the torque you need. Note that steppers do “sing.” In fact there are programs that play music using stepper motors, so the noise is not concerning. What is concerning here is that you do not have enough torque to hold the line when cutting with the router.
Edit: I may be telling you something you already know, but here is a video of singing steppers so you have an idea of the range of sounds they can be make.
The top picture in the page you linked appears to be in parallel? That’s the setup that came with the kit that I got. I can rewire the whole thing if I have to, but it’s some hassle and I’d prefer not to unless you’re confident that it’s contributing to the problem.
I haven’t touched any configuration settings related to the drivers, they’ve been left exactly as they were when they arrived in the mail. The machine used make what I would consider to be “normal” stepper motor noises (e.g. sinusoidally varying pitch when tracing a circle), now it’s significantly harsher and louder. It sounds like the whole frame is vibrating, although there’s no change in noise when I hold the conduit as it moves. That noise is concerning me more than the poor quality cuts, as I was pushing the machine pretty hard when that happened and I can always turn the cutting feedrate and depth down.
To give a bit more context, this video has 2 moves, first at F1000 and then at F2000. The first move is what I would consider normal and not concerning. There’s a marked difference once it switches to the faster feedrate. You can hear a distinct buzzing quality, not just the higher pitch and slightly higher volume I’d expect from running a stepper motor faster.
The top picture in the page you linked appears to be in parallel?
No, the top picture is the two steppers connected to the serial wiring harness. Further down the page is this picture detailing the wiring of a serial wiring harness. In the past, Ryan/V1 sold a serial wiring harness, but he phased it out when he stopped selling the Rambo-mini (the only board that required serial wiring), so you would have to construct your own version.
That’s the setup that came with the kit that I got.
If you didn’t purchase the “kit” from Ryan/V1, then the harness is almost certainly parallel.
I haven’t touched any configuration settings related to the drivers, they’ve been left exactly as they were when they arrived in the mail.
I’ve purchased DRV8825 drivers, and the original vref voltage settings are random. This is likely the root of your issue. You must set the vref voltage for each driver. There are multiple tutorial pages and YouTube videos on how to set the vref voltage for various drivers. Look at the purchase page to verify the model of your drivers.
I can rewire the whole thing if I have to, but it’s some hassle and I’d prefer not to unless you’re confident that it’s contributing to the problem.
If it were me, I’d rewire the machine and give each stepper its own driver, but I think you can get away with parallel wiring if you set the current of the drivers appropriately. It was mentioned recently that the original version of the MPCNC sometimes had its steppers in parallel. @jeffeb3 or @vicious1 will pass back across your question and give you recommendations. If you provide the model and specs for your stepper motors, it is possible that someone can be more specific about recommending the current (vref) setting. But no matter what, you must set your vref.
this video has 2 moves, first at F1000 and then at F2000.
Very strange noise. It is possible this is the steppers losing steps at a high feedrate, but this is not the noise when my machine loses steps. Note that if you did not have the current set high enough in the stepper drivers you might lose steps with faster movements. I’m not sure what happens to the steppers if you set the current too high.
My activity on this forum (and my experience on with the MCPCN) don’t date back to 2017, so Ryan or Jeff will need to give you feedback. That sure looks like a parallel wiring harness. As for the drivers, it is possible that Ryan was setting the vref before shipping his Ramps boards out. He will have to comment. Once Jeff or Ryan gives you their recommended current settings for a parallel setup, I would still recommend checking the vref voltage. As long as you are careful, it is simple to check the vref.
Since you bought your kit from V1, I’m far less certain this is a vref/current issue, but I cannot suggest what else it might be nor what troubleshooting steps you could take next.
Oh that is from waaaayyyyy back. You even have the faux therm resistor!
I did set the drivers back them, it is a good idea to check them though.
You are wired in parallel, this is exactly what happens when two steppers are connected. One feedbacks to the other and tries to stop or drive it in reverse. This is not an issue.
We need to see this to determine what the problem is. All the rest you describe doesn’t seem like any issue.
You are using an 8bit ramps, 32nd stepping, and 12V. Your controller is limited. Off the top of my head I do think 30mm/s is the limit, so 1800/60=30mm/s You should keep your moves under that speed. The Z axis is all thread so that needs to stay under 8.4mm/s.
Fairly certain this is not an issue your processor is limited and will stop moving any faster and if you try to go too far over that limit it will randomly stop sending commands and you will have errors in your cuts. I think you are just trying to move it too fast. At that point it was built to run at 8-15mm/s XY and 2-8mm/s Z.
So it sounds like there’s no actual problems, I’m just pushing it beyond its design constraints and it’s (obviously) complaining about that. I’ll try some test cuts and if there are any issues I’ll get back to you.
Well that is what I am thinking but I could be wrong. Try slowing down and see how that goes. If you want to go faster, we can start you on the upgrade path. For the amount you use it though might just want to stick with it for a while longer. Maybe I will get another version out by then…
