again print problems

hi all,
my mendel90 is giving me a hard time. i tried to print the bedleveling software, kisslcer and cura, both have problems with steps i think. somewhere in the print the x position with kisslicer and the y position with cura changes, see attached pictures.
what goes wrong there?
thanks, michael

i checked all the endstops, they are working and sending the l or h respond with m119. g28 is part of the start code.
i printed out the small 20mm box for calibration and that was beautiful.
then i was trying to print a lithophane from makerbot about 6cm by 6cm, stl model is okay, and somehow the printer was printing the model in the left front corner instead of in the middle. i was trying 3 times, changing the gcode export file on sd to .g and few letters, it did not print it in the center.
tried again the cube, again in the center.
slowly this is driving me insane ;-)

michael

i forgot, i used kisslicer. bed center x 100 y 100, printer size 200 by 200.

when shit hits the fan.....
the cube was printed in the center, but itlooks more then fer then a cube ;-)
the filament was blocked. the forth time on this new roll!!!!
attached is also a photo with a nice print.

The bed centre is 0,0 on Mendel90.

---

[www.hydraraptor.blogspot.com]

thanks nophead,
that brought the print to the middle of the printbed. i downloaded a configuration from somebody else, and i thought i should be right.
however, the next problem occured. it looks like the printer was missing steps. attached is a picture.

how can that happen?
travel speed is 100, extruding speed something like 50 or 60, so not too fast.

michael

The picture shows that you are missing steps in both X and Y. Do all of your prints behave that way?

Are you sure that your:
- belts are have the right tension? low bassy sound when plucked
- set screws are tight on your pulleys, and on the flats of the steppers
- stepper voltages aren't set too high? Did you tune them with the pots on the stepper drivers?

Fl0yd

i just find out that when i switch motors off and manually push the print table or the x axis, i does not go smooth. there are points where it is harder and softer. so that might be a mechanical problem. i tried to loosen the belt tension and it hels, but it is still a little bit uneven. are there acceleration variables that i canchange in the firmware?

thanks, michael

i found out it is not only the belt tension, it is also how it is distributed. the piece of belt from the print table to the y axis motor was too tight, although the rest of the belt was okay. so loosening the belt there for 1 tooth made the difference.
now he is running smoothly.
so again another test print....

michael

Uh? How can the belt tension not be equal all the way along? If the tensions on two ends attached to the carriage were not equal the carriage would move until they were.

Are you sure the Y axis bars are parallel? The short one is made parallel to the long one by loosening the screws and moving the carriage to one end, tightening those screws and then repeating for the other end.

---

[www.hydraraptor.blogspot.com]

I am having a similar problem with skipping prints. I have adjusted the belt tension to loose, tight, and various amounts in between. I also have the issue of "variable" tension when I turn motors off and slide the carriages. When the belt is removed the carriage moves smootly. It appears that the belt is stiffer in some areas and is more resistant to turn on pulley. Also, it has a tendency for a tooth to not fully mesh in the metal pulley and ride on the top of the grove. This makes for uneven prints.

I believe the print offsets are from the motor skipping steps. I have reduced the acceleration from 4000 to 2000 to 1000 and I still see this issue on the x axis, sometimes on the Y but its rare.

I can reproduce it easily by applying a little resistance to the pulley as its printing and the motor will quickly miss steps.

I have not tried to adjust the trim pots on the Melzi because my multimeter is not working and Im afraid of frying something.

Which direction (and how many degrees) would one turn the trimmer pot to safely give the motors a bit more juice?

The trim pots are set for 1.2A, which is plenty and towards the maximum the Melzi will supply without a fan. The motor should be quite stiff to turn by hand when energised so if you can stop it with little pressure that doesn't seem right.

I don't understand how the belt can ride out of the pulley teeth unless it is far too slack or the pitch of the belt does not match the pulley. I also don't understand how the belt can be stiff enough to make the pulley hard to turn. The force to flex the belt should be negligible compared to the torque available from the motor and the force to move the carriage.

Are you sure the idler has been adjusted so the belt runs in the centre and doesn't rub on the washers?

---

[www.hydraraptor.blogspot.com]

I should be clear, I can't "stop" the motor with my hand, only that I can make it miss steps and notice the print layer immediately being shifted with a little pressure on the pully while printing.

I will also clarify that the belt doesnt necessarly make it "hard" to turn, only that when sliding the carriage back and forth there are sections of the belt that seem stiffer than others, its a very uneven amount of resistance.

It is quite possible the belt is riding against the washers in the idler, I have adjusted it but I have also been playing with the different tension levels so I'll have to revisit the idler tonight. Perhaps that is adding the extra friction?

I will also take photo of the belt/ tooth / pulley mismatch tonight, it will be more clear.

Thanks.

so tbowren and i have the same problem. my english is not as good to explain these things nicely.
but after i removed the belt by lipping and turning it over the pulley and made sure to make it a little more loose on the part between motor and print table it worked. then the whole table was running smoothly.
it was not the belt touching the washer. in my case it did but with manually puttung the belt in the middle of the idler still the problem existed.
i think it is possible to have different tensions in the belt because the belt ist flexible and it has teeth and with the pulley and the idler and the turning of the belt, you have internal friction . that might be the reason. theoretically that would not be possible.

but i will check again and report if it was something different that changed.

michael

You should probably post some photos of machine and belts. Maybe there is some problem with assembly.

@tbowren.

It seems odd that you can't easily stop the motor but you can easily make it miss steps. That sounds more like the pulley is loose on the motor shaft. With the motor energised and stationary see if you can turn the pulley.

The belt is flexible so doesn't take significant force or friction to bend it round a pulley. The detent torque of the motor is more significant when pushing the carriage with the motor off. It also acts like a generator so the faster you try to move it by hand the more torque it takes. The belt should also be uniform so shouldn't have stiffer sections. If it jams hard against the fender washers it can add a lot of friction.

@neptunier,
If there is different tension in the two halves of the belt the carriage should move to equalise it. If it doesn't it must be jammed.

The belts should be tight enough to twang with a low note, like a bass guitar.

---

[www.hydraraptor.blogspot.com]

I had this same problem but to a smaller degree. When I ran the bed along the rails with the power off I could feel some unevenness in the travel. When I checked closely, the belt was occasionally climbing over the pulley tooth. I did think that my belts were tight enough, but when I tightened them more and was able to get a true low note when plucked the problem disappeared.

I would recheck that the belts are tight enough.

It's definitely worth checking the idler pulley after initial construction and once you've had the chance to run the machine through its paces a few times.
I thought I'd tightened mine up enough but maybe all the moving around to check all my other bolts, the vibration from the first few prints and general settling in can make what seems to be a tight connection work itself loose. My Y-belt was significantly slacker than I thought when I checked a week or so into ownership so I ended up torquing down on the idler bolt a lot more to make it stay - there's a slight depression in the dibond material now which tells me that it's both tight and less likely to slip, at least without being noticeable anyway.

today the printer was running fine, until now. it again lost steps. so my solution from yesterday was not so permanent ;-)

i attached fotos, maybe i did something wrong without knowing or recognizing. the belts make nice low sounds.

my other printer, a mendelmax was running today without any problems, so i thought it would be a good day, two printers running....

could anybody have alook at the pictures, thanks.

michael

here [www.youtube.com] is a video of the belt going over the pulley. he is moving up and down. that movement correlates with the ease of pushing the axis by hand. sorry about the quality, but you can see the movement.

@neptunier,
The belts need to be centred on the idler pulleys. For Y you angle the idler bracket so it runs in the centre. For X you shim the inside washer with something thin like an offcut from the PP strips. This is in the instructions.

The belt tensioner is fitted back to front. The only effect of that is to reduce the range of movement before it hits the fan bracket. It also looks like it is set for minimum tension, i.e. it is hard against the belt clamp.

The video just looks like belt is far too loose so it is able to ride up the teeth.

Edited 1 time(s). Last edit at 07/09/2013 05:06PM by nophead.

---

[www.hydraraptor.blogspot.com]

This picture looks the belt teeth don't go all the way to edge of the belt.


Is it an illusion or is a faulty belt?

---

[www.hydraraptor.blogspot.com]

Ok, I checked my idler and you were correct, it was rubbing against the washer. Also, the bolt that holds the x axis rod on the non motor side was loose. I wish there was an extra hole cut in the faceplate to make it easier to tighten that bolt.

Upon centering the belt and retighting all the bolts it moves quite smoothly now. I did a 3 hour print and it looks good.

One thing I will add is that I do still see the teeth trying to ride out of the pully when I make the tension too tight. It appears to me a belt with more rounded teeth would slot better, but that is not based on experience at all.

Thank you Nophead for you insight, I was so focused on the drive side of the pully I neglected to think about the idler end.

Edited 1 time(s). Last edit at 07/09/2013 06:40PM by tbowren.

The T2.5 belt has trapezoidal teeth and so does the pulley. They should fit together perfectly. They certainly shouldn't ride out when there is sufficient tension to keep the belt straight. Actually, they should continue to function with far more tension than the plastic parts can apply. I am wondering if there is something wrong with the belt.

---

[www.hydraraptor.blogspot.com]

nophead,
the teethof the belt go all the way through, so that is not the problem.
i tightened the y axis belt as strong as possible, but there is still this problem. the belt is running in the middle of the idler.

the belts of my mendelmax are much softer and they have less tensin but they are running fine. i dont have much experience with belts, but as tbowren and i seem to have the same problems, did you change the belts recently? maybe the belts are just too hard?

michael

No we haven't changed belt supplier since we switched from T5 to T2.5 a long time ago. We get it in 100m reels and cut them up. They do have steel wires but I would not describe them as stiff but they are stiffer than belts with kevlar reinforcement.

Playing with a piece here it seems fine, but if I put a sharp bend in it then it remains kinked. Maybe that has happened to it somehow. If you remove the tension can you see kinks in it?

If you are sure they are tight enough then all I can suggest is we send you some more belt, but it is probably from the same reel. Do you have the same problems with both belts?

---

[www.hydraraptor.blogspot.com]

hi nophead,
one more thing that came to my mind is that i had a problem with broken x axis holder on the left side. as a german coming from a tradition were machines were build to last a 100 years i do not feel completely comfortable with that solution ;-)
all the x axis and all the belt tension depends an a few sqare centimeters printed in the direction of the layers, which very easily breaks. a little too much tension and it breaks...
anyway, therefore there was movement in the left part of the x axis holder and that might have also attribuated to the belt problem.
so i put in these things you see on the photos which stabilize the thing a lot. i will exchange the ugly screws with headless screws.
i hope that makes it better.
but the problem on the y axis is still there. it is less with more belt tension but it is still there. if you move the print table fast by hand you recognize it. i doublechecked all screws there, nothing is broken ;-) so it should work.

did you ever had faulty belts? if yes i might be worth trying to change them.

michael

Yes that is a bit of a design weakness. It needs just the right amount of tension on the plastic parts to grip without breaking them. Your fix looks good for the front bar (but the one at the motor end is not required as the bar can rest against the box around the motor) but the back bar has the forces from the belt in the opposite direction.

No I haven't seen any problem with the belts in more than 250 kits but nearly every supplier we have supplies faulty stuff at some time or another.

Have you tried removing the tension to see if the belt is kinked? That is the only thing I can think of that explains the symptoms. The belt can't have some sections stiffer than others as it has uniform construction but the wires can be kinked which makes that section harder to bend round the pulley.

---

[www.hydraraptor.blogspot.com]

neptunier:

I really like what you did there adding the locking bolts to the axis. I have noticed that I cannot tighten the motor mount enough to make it stay perpendicular when I tension the belt. I even cracked one of the rod fasteners on the non motor side trying to tighten things up.

Right now the belt teeth are not slipping, but it is loose enough that I can see it changing tension when the motor changes speed. I think it may be giving me some wavy patterns and smoothed out edges in my prints. I would be nice to try a bit more tension and see if the waves go away.

Did you degrease the rods properly? It should be possible to tension the belt sufficiently without breaking the parts.

---

[www.hydraraptor.blogspot.com]