machine gone rogue! (motors at least)

[mcmasterp]

I have been using my awesome mpcnc lately to build a bartop arcade for my father. It’s probably the most work I have asked of my machine as I have since building it only cut a few signs and lasered, but not much time lately.

I was having issues with chatter or run out or deflection or whatever its called while working acrylic and found that the legs of the machine where not very rigid leading to a lot of movement. The belts had also loosened over time or maybe the zipties. And finally some bolts on rollers and middle parts that effect z tubes. So I tightened the belts a bit but not too much, gave the bolts a small tighten but in no way tight, added gussets between the table legs in the front which made it rock solid. I took care not to alter the positions of the legs as to keep the geometry of the machine the same. I then tried to cut the same piece with 100 times better results, I still have to go slow to avoid chatter which is tough in arcylic where you need to go fast to prevent melting. I use a 300w spindle with speed control so I am able to turn the speed down to help but too low of a spindle speed has its own chatter issues. In the end piece looks great except I notice my sacrificial bed which was leveled long ago must not be because one side of the pocketing is 1.7mm deep, the other only 1mm!

SO, I get ready to run bed leveling gcode I made, I’m jogging the machine around as I have a 100 times before and one motor stops for a split second on the y axis! I check things out, all seems normal. I jog back and forth and it does it more, randomly, and the other side of the same axis as well! then a swear they sometimes go in opposite directions!

Mind you these are long ago proven wired correctly with all connections very secure and if possible soldered and shrinkwraped. I don’t use the quick connect screw down terminals on the rollers and the connections to my driver boards is secure and more over unchanged in all this time, no strain etc.

I should also say that during my build I already owned a usb controller board, 2.5amp driver boards and control software from buildyourcnc.com, they are very nice and purpose built so you would assume better than ramps setups.

I checked everything I can think of. I loosened all the bolts back just to be sure, I turned up the power on the driver board which seemed to help but not completely and the board and motor got too hot. I can’t see anything that would cause the motor to act like this, the belts still not too tight, wiring good.

Too say I’m sad would have to be an understatement. I originally bought a mydiycnc.com kit which was a total pos. I spent hours and $2k trying to get it to work (also why I had the premium electronics) before I found the mpcnc project. I was in love. the parts printed issue free which was a sign from god since my reprap is a pos too lol. I built a 4’x3’ machine on a sturdy table and all the issues that made ruined my hobby where no more! aside from the expected flex ( which I now wonder if gussets wouldn’t have solved) and limitations of a diy machine I was in heaven. My custom spindle mount has led lighting and line laser targeting, a dust collector, laser mount spot and wire plug. it’s perfect. I’m hoping someone has an idea or some insight I’m yet to think of

Thanks in advance for any help,
Paul

[vicious1 (Ryan)]

Well I can help with the chatter. Raise the acrylic up, all the way. Top out the spindle and shim up the bed until it moves the work piece all the way up. Put some 2x4’s or something underneath it. Then if you are still having issues try a single flute bit, then you literally only need to move half as fast as a 2 flute. If you still have an issue try a single down flute…my fav. If you keep a vacuum on it or compressed air it will keep it cooler if you really need it. You should also get rid of the vacuum straight piece and just mount the hose. That straight rigid piece pulls on the cutter, a lot, like a breaker bar. There are tons of little things that can be done. CNC machinists get payed very well for a reason.

As for your movements, it is hard to say. I would guess a bad connection, but you say they are good. Next would be the drivers but I have zero experience with external drivers, I would say check for signal noise. Maybe end stop issues or a power wire crossing your driver input. You might need to isolate your wires more, or use shielded, or maybe they got cut on the conduit edge somewhere?

[vicious1 (Ryan)]

I say this a lot. If you can half the length that your z axis hangs down you double your rigidity, same goes for the length of the legs. A 3" machine is twice as rigid as a 6" and a 1.5" is twice that.

[mcmasterp]

yes my z axis drop is about 4" with tool in. I wish the spindle mount was higher on the pipes but I guess it would hit something? i dont know I 'm in front of the machine at the moment. I was planning on printing the new middle and z anyway soon.

The movement issue is the killer one obviously. no problem found, intermittent issue, it’s a nightmare. I appreciate your quick response and suggestions. I the end it is all operator error some how. I just have to figure it out. You make house calls? I’m pretty sure the local cnc repair techs would either laugh me off the phone or back out the door after arriving and seeing my unholy “not from a manufacture!?!?!” machine.

[vicious1 (Ryan)]

Well I know it would suck but you can either switch axis drivers to see if the problem stays with the axis or with the driver. Then Either replace the driver or the wires. Quickest way to test, plus you don’t want to screw up any more jobs. Bummer, I had this happen with a printer of mine. one of the wires broke inside of the insulation and caused random issues. I found it but clamping on a continuity tester and wiggling each wire till I found the little booger. Then ended up relacing them all anyway…

[jweers (John)]

Let’s see pictures of all the electronics and connections. I don’t think noise with a reversal of direction, I think cross talk or momentary shorted connection between wires (thus the pictures I’d like to see).

Make sure all the pins on the boards are clean and tight. Connections snug. Check for over temp on the boards. Look where wires are close to metal and see if the have scarring on the insulation.

[nerdyrcdriver]

Agreed, it almost has to be a connection issue or faulty stepper drivers. There is also a chance that another program on your computer could be sending random stuff over the line connected to your machine.

Vicious had the right idea to switch axis and driver combinations. X and Y would probably be the easiest to see the problem. If you plug the x axis into the y driver and the y axis into the x driver you will pretty quickly figure out where the issue lies.

He also had the right idea with checking continuity through each wire while moving it (machine powered off and motors unplugged). I have wasted many hours troubleshooting problems by assuming that wires with no visible damage are fine. If you don’t already have one, a multimeter with a continuity testing function is worth every penny.

[mcmasterp]

I’ll post some pics of the electronics.

I did check all the connections but I have all my steppers soldered and shrink wrapped so I’ll have to unsolder I guess to check each one.

Here’s a strange development that may be a clue. My max speeds set in the usb cnc software is 4800 but I usually jog at 3600. On a desperate attempt I set the max speed down to 2800. And now the problem has yet to reappear. That’s a horrible solution of course and I’m running a simple bed leveling gcode now to see if it looses steps.

I should say my driver boards have stick on heat sinks and get pretty hot. As do my steppers. I of course have consider that the motors could just be cooked or drivers or both and that the timing was completely unrelated

[mcmasterp]

More pics I have on phone right now

[konp (Nate)]

The problem sounds like maybe one of your stepper drivers is overheating and going into protection mode - they stop momentarily to allow the heat to dissipate. This is a common problem on a lot of 3D printers, usually alleviated by adding a small fan somewhere near the drivers.

So try that, just a little 30mm 5v or 12v fan blowing air across the stepper drivers. See if it makes a difference at all.

[barry99705 (Barry)]

Is that a garbage can lid they’re attached to??? Swap that out with something that’s not a static generator as soon as possible. Doesn’t take much to kill the electronics.

[mcmasterp]

@nate yes I have a small desktop fan pointed at them because they do get hot. I may add tiny fans directly to each board

@barry. Haha it almost is a trash can lid. It’s the lid from storage box from Lowes. It’s made from a rubbery kind of plastic that is great for electronics mounting among other reasons because it is static free. I usually cut them up into sheets before use but I never got to it

[nerdyrcdriver]

Sounds like you might want to turn down the current limiting on your drivers if you can. I think most people here run their machines with the steppers getting warm to the touch at most. My steppers are barely above ambient temp when my machine is running.

[mcmasterp]

@nerdyrcdriver I had to turn them up actually to help with the issue. I loosened some of the bolts on rollers and middle to make it move easier as well. All things I should not have to do. I think drivers and motors replaced is an option to look at. They could be permanently damaged at this point. I’m wondering if nema23 steppers and bigger drivers would help. I don’t think there are new style motor mounts for nema23s yet.

At that point I wonder if I should upgrade the belts to chains. Sprockets and idler bearings for #25 bicycle chain is cheap. If it further solidified the machine by eliminating belt backlash that wouldn’t be bad!

[vicious1 (Ryan)]

23’s will rip your machine apart. the 76 oz 17’s we use are only being used at half power. There really should be no reason to upgrade. Look for some other issues. chain has massive backlash, belt is the most efficient way to transfer motion, and is less moving mass. Maybe the drivers you had are not made for this kind of stepper. Double check the specs of the drivers and the steppers.

[mcmasterp]

Yes I got the drivers and steppers as a package. Each driver Powers two steppers but the 2.5amp drivers are supposed to be capable of l that type of setup and as you said we shouldn’t be pushing things hard

[vicious1 (Ryan)]

Are there other settings for those drivers like decay rate and dwell time that maybe need to be set?

[vicious1 (Ryan)]

So those drivers are the same as we use, drv8825. You need to add cooling at anything above .7V heatsink and a fan. If those drivers were built corrctly you will be adding the heatsink to the bottom of the board and will need a fan on it. The way you have them mounted might be causing them to have thermal shutdown from being mounted hot side close to the board.

[mcmasterp]

I don’t remember seeing decay or dwell in the software which is plantcnc usb software that works with buildyoucnc.com hardware. I believe the boards are clones of plantcnc boards in some way. I’ll have to check more on that.

It sounds like you found the spec sheet for these drivers on buildyoucnc.com? I do have heatsinks on the chips that get very warm. Are you saying I need ones on the underside on the pcb as well?

I also have a small desktop fan that points just at the three driver boards but I could add a tiny fan to each board. Guess I’d gave to put one on both sides IF you agreed above that heatsinks on underside are needed.

[vicious1 (Ryan)]

Just have a look at your board, If there is a large copper pad under the biggest chip that is a heat pipe. It conducts heat better than the black chip itself. These are the same drivers we use on the ramps, just with screw terminals instead of pins. you have 62 oz motors so .7V should be plenty.