RE: Core One crash/restart
Yeah, actually earlier in the thread I think someone even said support indicated this wasn’t recommended.
If I had to guess they’re having a difficult time reproducing the issue to the point of finding a fix. It seems like our fix works, but it may not be the “proper” fix. Or at least, supportable one. Seems like the stepper harness should just add a ground lead that is connected to frame then buddyboard enclosure.
RE: Core One crash/restart
The frame that the motors attach to is grounded via a dedicated "faston" connection. There is no provision for a ground lead to the motors and it would be a complicated modification to existing printers to add a separate wire. The motors are designed to ground through the chassis, therefore, the simple solution to ensure that the motors are grounded is to ensure a low impedance interface between the screw and the frame either by removing powder coating or adding a "star" washer or some other piercing washer that would penetrate to the metal under the head of one or more motor mounting screws.
Since only the stepper motors require grounding there is no other design purpose for the "faston" ground connection to the frame. Even if you have a suspicion that the crash/restart problem is potentially caused by another issue, this problem of build versus design intent should be corrected. Anecdotal evidence strongly supports the conclusion of the forums that the motor grounding is the source of the issue. I do not believe that there is a single forum poster that complained of this crash/restart issue for whom the issue was not resolved permanently by grounding at least the X-axis motor.
My $0.02,
Steve
RE: Core One crash/restart
Just experiencing this on the BenchLighthouse demo file ... started on the first print attempt about 5mm up the print. Second attempt is is starting to fail doing layer one. So whatever it is, it is changing - something mechanical, like something is loose and binding.
I'll go back and read what this 'magic bus' fix is and try to apply is... but not real happy to see this type of issue on a new printer.
RE: Core One crash/restart
Does anyone have photos of the motor side connections made for this mod?
Or better, anyone have links to schematics of the printer: specifically the control board to motor paths? They'd be handy to understand why grounding a motor case might be helpful.
ps: early in my past life I was an electrical engineer and frequently called on to study ground path issues on higher end electronics. And yes, simple connections are often not so simple, stray currents are always there, and taming them is almost an art form.
RE: Core One crash/restart
Does anyone have photos of the motor side connections made for this mod?
Or better, anyone have links to schematics of the printer: specifically the control board to motor paths? They'd be handy to understand why grounding a motor case might be helpful.
ps: early in my past life I was an electrical engineer and frequently called on to study ground path issues on higher end electronics. And yes, simple connections are often not so simple, stray currents are always there, and taming them is almost an art form.
The four wires going from the xBuddy board to the motor are only two pairs of current loops for the coils (which are, of course, isolated from the motor chassis). So the motor chassis, which is in electrical contact with the pulley and can hence collect static charge from friction on the belt, needs to be grounded separately.
The easiest and cleanest way is to ensure that one of the four motor mounting screws makes electrical contact with the CoreXY frame. (The frame is reliably grounded through a ground strap.) Remove a screw and scratch off the paint below its head from the CoreXY frame. Or just insert a toothed washer with sharp edges which can penetrate the paint.
RE: Core One crash/restart
I don't think that having the electronic schematics of the printer will help here. My theory (also being an electrical engineer): The high speed movements of the belts will charge the motor(s) which is (are) isolated from the chassis. At a certain moment, the static voltage of the motor compared to the chassis will be that high causing an electrical discharge between such motor and the chassis. This discharge will have the control electronics malfunction/reboot. So make sure you get the chassis of both XY motors well connected to the frame of the printer. Easiest way is to do this is to "Remove a screw and scratch off the paint below its head from the CoreXY frame. Or just insert a toothed washer with sharp edges which can penetrate the paint." as indicated by @jurgen-7. And if you have a multimeter, please check the low resistance between the motor chassis and the printer frame using the resistance measurement function. Resistance should not be higher than a few Ohms.
By the way, since I have applied this fix, I have had no rebooting issues anymore. This while I have been printing big parts almost daily since then.
RE:
On the face of it, both X and Y motors are floating on my printer; as are the Z motors (and those look to be floating almost intentionally). By floating, I mean not connected to anything. Also, I'm not sure of many Prusa printers having the motor cases intentionally grounded. I also can't see any immediate adverse effect to the motor housings floating. Maybe static buildup and then an arc through the paint? But that seems like a long stretch. I might try to measure the motor case voltage, just to see if any static charge is building. Again, other Prusa printers have floating motors and belts ... not life this is the first with that configuration.
I printed several small tests, then the full rocket engine, but the engine mostly on stealth because it was pretty noisy. Printing the lighthouse today is where my fun started. Long diagonals for infill 'seems' to be where the crashes happen. I'll play with ground the motors tomorrow ... but honestly, I'm not expecting success. I am also not disassembling anything to do the grounds, everything is semi-accessible from the chamber side.
Coming at this from another direction, I might also craft a large oval that fills the print bed and use that as a sample print. If it fails, then the sample Lighthouse model gcode won't be suspicious.
On another note, I think I did see this 'crash' one other time on a simple 20x20mm vase mode cube ... but it was before the door sensor got turned off and it happened when I opened the door. The printer went into what I can only call a power fail reset. One more piece of the puzzle.
RE: Core One crash/restart
What about a spark between the motor chassis and the wires driving the motors? Personally I am convinced this is a static discharge issue. Also because it occurs often but quite random. And after "grounding" my motors to the frame I haven't seen any of these issues anymore.
RE: Core One crash/restart
As others have reported here and as I have experienced as well, long diagonal prints going from 0,0 to 250,250 (not exact coordinates, just an example) cause the printer to eventually crash. The conclusion is static build-up on the x motor housing. Grounding it solves this problem. Doing so doesn't require any disassembly beyond removing a screw and either scraping off some paint or adding a star washer. Both my X and Y motor housings are grounded to the frame and I haven't seen the problem since.
RE:
On the face of it, both X and Y motors are floating on my printer; as are the Z motors (and those look to be floating almost intentionally). By floating, I mean not connected to anything. Also, I'm not sure of many Prusa printers having the motor cases intentionally grounded. I also can't see any immediate adverse effect to the motor housings floating. Maybe static buildup and then an arc through the paint? But that seems like a long stretch. I might try to measure the motor case voltage, just to see if any static charge is building. Again, other Prusa printers have floating motors and belts ... not life this is the first with that configuration.
I printed several small tests, then the full rocket engine, but the engine mostly on stealth because it was pretty noisy. Printing the lighthouse today is where my fun started. Long diagonals for infill 'seems' to be where the crashes happen. I'll play with ground the motors tomorrow ... but honestly, I'm not expecting success. I am also not disassembling anything to do the grounds, everything is semi-accessible from the chamber side.
Coming at this from another direction, I might also craft a large oval that fills the print bed and use that as a sample print. If it fails, then the sample Lighthouse model gcode won't be suspicious.
On another note, I think I did see this 'crash' one other time on a simple 20x20mm vase mode cube ... but it was before the door sensor got turned off and it happened when I opened the door. The printer went into what I can only call a power fail reset. One more piece of the puzzle.
I am not sure what "puzzle" you are trying to solve here? Many users have reported the crash issue; all of them have reported that it only occurs on large prints with fast movements; all of them have reported that grounding the motor chassis fixed it.
Static buildup from fast and extended belt movements seems entirely plausible to me. The spark discharge path could either go through the paint layer to chassis ground, or inside the motor into the coil connections. Again, it seems entirely plausible to me that this can cause the firmware to crash.
What else do you need to know, or what is keeping you from just grounding the motor chassis and solving the problem?
RE: Core One crash/restart
The puzzle is finding the root cause. We're all guessing it is static from the belts. But no one actually knows. Sort of like the murder mystery where everyone thinks the butler did it. I agree with the feedback points to static, but I also wonder if simply taking the box apart and putting it back together doesn't provide a temporary fix.
I've just completed my ground the motors task, and a print is in process. Yesterday the stops were within a few minutes, 6mm or build. So I'll have some feedback pretty soon.
If the issue does go away, it brings up the next problem grounding the motor cases presents: arcing through the motor bearing is a very real problem. Arcing damages the bearings, every time it happens. Yes, bearings are floating in oil and not electrically connected to the case, so all grounding the motor case does is move exchange the stop problem for a motor life problem. The best solution, if ESD is the issue, would be to ground the motor shafts or pulleys in the system or install ESD brushes contacting the belts... but so far I'm not finding suitable brushes.
RE: Core One crash/restart
The puzzle is finding the root cause. We're all guessing it is static from the belts. But no one actually knows. Sort of like the murder mystery where everyone thinks the butler did it. I agree with the feedback points to static, but I also wonder if simply taking the box apart and putting it back together doesn't provide a temporary fix.
The evidence is good enough for me, especially since the fix it points to does actually work. And I wonder how you would go about pinpointing the problem even further? Try different belt materials? Connect an electrometer to the motor chassis to confirm a charge buildup?
I don't think there was much "taking the box apart" in the more recent successful troubleshooting efforts: Remove one screw from the motor mount, insert a toothed washer, fasten the screw again. How could that "provide a temporary fix" -- other than by grounding the motors, which is assumed to be the permanent fix?
If the issue does go away, it brings up the next problem grounding the motor cases presents: arcing through the motor bearing is a very real problem. Arcing damages the bearings, every time it happens. Yes, bearings are floating in oil and not electrically connected to the case, so all grounding the motor case does is move exchange the stop problem for a motor life problem. The best solution, if ESD is the issue, would be to ground the motor shafts or pulleys in the system or install ESD brushes contacting the belts... but so far I'm not finding suitable brushes.
On my Core One, both motor pulleys have permanent low-resistance connections to the motor chassis, no matter how I move the belt. So the oil film in the bearings does not provide significant isolation, and there should not be any voltage build-up on the pulleys which could cause arcing.
RE:
So far no fails in the first print... fingers crossed.
My motor shafts read several k to case, and probably higher if I measure while the motors are spinning. Perhaps that'll get lower as the motors wear (new printer).
As an aside, I have a request into LDO asking if they provide any internal arc protection for their motors. If they respond, I'll post it here.
RE: Core One crash/restart
As an aside, I have a request into LDO asking if they provide any internal arc protection for their motors. If they respond, I'll post it here.
That may be interesting for general context. But you are aware that Prusa no longer uses LDO motors?
RE:
No, I wasn't aware. Who are they using today?
RE:
No, I wasn't aware. Who are they using today?
I have not seen that mentioned anywhere. There was a bit of a controversy around the motors early this year (I believe):
An ex-Prusa employee had started his own boutique 3D printer company, https://ston-3d.com , and stated Prusa's move from LDO to "some AliExpress special" as an example that Prusa is forfeiting quality for cost savings. Prusa published a statement somewhere (was it Jo Prusa himself on Reddit?) saying that LDO had developed quality problems, and that's what prompted Prusa to move away from them. But they did not say what brand(s) they are using instead.
RE:
4 hours in, no crashes ... and I also used a drill bit to scrape the paint away under the internal star washer.
RE: Core One crash/restart
Firstly I offer my thanks to the person who developed this idea and suggested a fix, also to those who followed up on it.
I too experienced this issue - I was trying to print a tool tray that used the full x/y of the print area. I tried twice and had so many 'crashes' that the print did not even complete. Prusa help did not suggest the idea of grounding the stepper motors but, finding this thread, I implemented the fix. I have just finished printing the tool tray. It has printed perfectly. I used the 'star washer' method of grounding both steppers motors driving the x / y axis. I suggest that it is a better method of grounding to use 'internal' and not 'external' star washers. My reason for this is that the head of the stepper mounting screw is shallow and thus having the sedations on the inside gives better crushing contact when tightening. I used a twist drill by hand to remove some of the powder coat - again, using internal star washers is more likely to make better contact with the bared area. I tested my motor / frame connection before and after installing the washer to check that I had acheived proper grounding. many thanks to all on the forum!
RE: Core One crash/restart
I have been seeing the same problem. Prusa support sent me a replacement Nextruder cable and LoveBoard. I replaced them (which is a pain in the a$$ when the printer is completely built) and it worked for about three days and the problem came back. Any news?
RE: Core One crash/restart
I have been seeing the same problem. Prusa support sent me a replacement Nextruder cable and LoveBoard. I replaced them (which is a pain in the a$$ when the printer is completely built) and it worked for about three days and the problem came back. Any news?
Have you tried just ground the motor housing like everyone else has been doing?