Print Shifting

Hey guys,

So I recently got my prusa i3 from folger tech up and running and had some really good prints. Roughly 10-15 hours of prints came out awesome. I had a couple of issues with prints shifting like one in 15 was good, but I recently have been getting all my prints to have this shifting issue. It seems to only be the x axis. I have tried moving around the stepper pots and it seemed to help a little bit but it went back to the same thing after a few minutes of printing. all of my belts are tight and my sets screws are tight. I haven't tried the acceleration or jerks settings because I thought that I would be able to fix it with the pots but that will be next on my list. Any ideas?

Thanks!

[youtu.be]

Did you try adjusting the current on the stepper driver for x? Sounds like it might be overheating.

I think so too because it does get quite hot. What voltage would you recommend using? I had it to the spec that folger tech reccomended. I think it was around .5-.6.

I run .68 on my Melzi board and fans on all motors. No problem with heat.

i think you are getting problems about Print Shifting then can contact with printer support because i had faced problems then i contact with printer assist really they had been solve my issue amazing services they provides best printer support services you can also contact

my folgertech i3 has exactly the same problem, I replaced the motor to motech 48mm and its working fine ever since, the stock motor runs hot no matter the voltage, all the 48mm nema 17 I run barely even gets warm let alone hot

put it simple, everything been replaced except the z motors which gets warmer than the x y motors I replaced and they barely move much and PSU which soon be replaced, it does not put out 12v after few months, it was at the start but after few months I had to turn up the pot

it would have been cheaper buying everything elsewhere separately, y carriage bends so you never get the bed levels, all printed parts are crap, I can print better than that, hotend clogs every 5 mins, eventually fall apart completely while trying to get it to print without clogging, z screw popping all the time, replaced with 10mm leadscrew and my own x motor mount and idler, ramp board blew up for no apparent reason, x y idler were awful, folgertech support didnt even have much clue about electronics when i contacted them last year, insufficient cooling on the direct drive that came with it, if you running the stock direct drive on x carriage, it be worst, since the stock motor skips on bowden drive, so all I had was the genuine e3d v6 hotend and a capacitive sensor, it would not matter however if there was nothing on it I dont think, I simply place my finger beside it with no force and it will skip. put a fan on it helps but the noise of that motor is unacceptable. heatbed doesnt hold temp even at 90c. I dont even use a heatbed anymore, I can print ABS, PLA and PETG on cold bed.

currently I'm building my new printer so bye bye soon to folgerSHIT, I would never go near them again. I should have stick to DIY, some of the parts I am using I have posted them on .thingiverse, I have new x motor mount and idler made again to allow me to use the big 10mm nut, the first set I made had the thread integrated, they still work after 6 months, but I tho it would be time to get the proper nuts in before they wear down, but not been posted as I'm giving up on this i3 soon, since the frame is starting to soften up soon it may not hold anymore, its the acrylic version not 2020, conclusion, it was the worst buy I had, the price maybe low, but so is the quality, I rate it super low if there was an option

Edited 9 time(s). Last edit at 07/13/2016 09:45PM by deaconfrost.

I have a similar problem: I’ve built a Prusa i3 from scratch: box frame, Greg J head extruder, Arduino mega 2560 with Ramps 1.4 electronics, DRV 8825 stepper drivers, Sprinter firmware and Pronterface and Sic3r software control. After considerable effort, I have finally been able to make prints—but not without problems. The X-axis does not make vertical sides, i.e. the edges look more like waves instead of straight up and down. (see picture). A typical print can vary 5mm in a 20 mm print in the X direction. The Y-axis is functioning fine and produces vertical lines. I have tried many things to see if I could correct this, but perhaps the most interesting result was when I placed the X axis motor in the Y axis connectors on the board and vice versa. In this case the Y-axis had the offset problem and the X-axis was fine. This seems to eliminate the belt linkage and motors as issues.

I have tried changing motor drivers, reducing the motor driver current, reducing the print speed, adding a fan to help cool the board. I have changed both the arduino and Ramps boards. All of this had no effect.

Any suggestions on how to fix this?

Skogi,

So I fixed mine but I don't know if our issues are the same. It seems like the issue doesn't have as much to do with the motor as the connection. I know that I had same issue, but actually turned out to be the motor connector to my nema motors. That doesn't seem to sound like the issue that I was having as mine was completely disabling any movement in the direction. I also added a fan to the actual stepper motor to try and cool the motor down because the motor can shut off for hundredths of second and continue printing, which seems to be the case for you. Those are my two suggestions as well as trying to tune the stepper drivers when you aren't printing to the lowest possible voltage that the motor will still be functioning at to see if that helps the cooling.

I wanna say that it is the connector for the motor because you seem to have ruled everything else out, but I don't know either.

I have the current on the X stepper driver about as low it will go and still operate (0.5V). Neither the stepper driver or motor seem to be getting overly hot. I don’t understand how the connector on the X motor can be a problem since the motor and driver worked fine when connected to the Y terminals on the board.

Could this be a clue to my problem? Even though the motors and gearing are the same for both the X and Y axis, after calibration the values in the Sprinter firmware line
#define_AXIS_STEP_PER_UNIT (x,y,z,e)
are different for the X and Y. X is 406 and Y is 200. Calculations show 200 is expected value. If this is a problem, what is the cause and fix? Is there something else in the firmware that is a problem?

Or could the issue be that I am using an old computer—with Pentium M chip. I have to use baud rate of 115,200.

if the motors are the same, same belt same pulley, then your firmware setting #define_AXIS_STEP_PER_UNIT (x,y,z,e) is wrong

Yes, the motors are the same type, and same belt is used and both axis have identical pulleys. As I stated the Y axis works fine and there is no offset in the prints in the Y direction. The value for the calibrated Y axis in the

#define_AXIS_STEP_PER_UNIT (x,y,z,e)

is 200. However if I place 200 in for the X value, the print head travels only 56mm when it is asked to travel 100. I would like to know what I need to do to reconcile this issue. Perhaps it will solve my x-axis shifting issue as well.

Do you have a picture of the belt path of the X carriage and the Y carriage perhaps?

Here are three photos of my Prusa i3 including the whole system (less computer) and close ups of the X and Y-axis motors, pulleys and belts. I put a red marks on the X-axis pulley and belt and since they are lined up, no slipping of cogs has been observed.

In order to see if the Ramps board X motor pins were somehow at fault, I reassigned the X motor control pins to the E_1 motor position on the Ramps board. The resulting print is very similar to the one shown in my first post in this thread, i.e. there is still X-axis shifting during prints. This experiment verified the problem is not in the X control part of the Ramps board.

What speeds do you have in your configuration.h file? That's a decent amount of weight on your X axis which makes me wonder if your missing steps when you move left and right. It would also explain why your step count is higher for X than Y, as based on your photo your X and Y counts should be identical. It also has me wondering what your Vref is set to on your stepper drivers for each axis.

Here are the speed settings that I have tried in the configuration H tab in Sprinter:

Step Initial
#define_MAX_FEEDRATE // (x,y,z,e) (400, 400, 2, 45) //mm/sec
#define_HOMING_FEEDRATE (1500, 1500, 120) //mm/min
#define MAX_STEP_FREQUENCY 30000
#define MAX_RETRACT_FEEDRATE 100 //mm/sec
#define_ACCELERATION 1000 //mm/s2 Also tried 500
#define_RETRACT_ACCELERATION 2000 //mm/s2
#define_MAX_XY_JERK 20 Also tried 10
#define_MAX_Z_JERK 0.4
#define_MAX_Z_JERK 5//mm/sec
#define_MAX_ACCELERATION_UNITS_PER_SQ_SEC (5000, 5000, 50, 5000) Also tried (2000, 2000, 20, 1000)
#define MINIMUM_PLANER_SPEED 0.05 //mm/sec
#define DEFAULT_MINIMUMFEEDRATE 0.0
#define DEFAULT_MINTRAVELFEEDRATE 0.0
#define_MIN_SEG_TIME 20000

I have reduced the voltage on the drivers a couple times –now both X and Y drivers are set at about 0.5V—any lower and the motors would not work. The AXIS_STEPS_PER UNIT setting did not appreciably change with the lower voltage setting.

If you increase the voltage for the X axis, does it still require the 406 in define_AXIS_STEP_PER_UNIT to move the same distance? I'd slowly increase the pot to 0.6, 0.7, 0.8, 0.9, etc...and see if your travel is the same for each move of 100mm with the 406 steps setting.

Here is data on the Voltage/Distance travelled. I used the X_STEP_PER_UNIT at 200 since this is the expected value if things were working right. The desired distance is 100mm.

Voltage/Distance(mm)
0.50/50
0.58/52
0.78/55
0.85/56
1.10/61
1.34/65
1.67/70

At 1.67V, the motor started acting flakey and the driver was obviously hot. It is obvious that by changing the X-driver voltage, a distance of 100mm cannot be achieved. The calculated X_STEP_PER_UNIT would range from 392 to 286 depending on the voltage used. It was interesting to find that no matter which voltage was used, it took 6.2 sec. to travel the distance shown.

Is there anything you can remove from the X axis to reduce the weight and see if it can travel the 100mm with the 200 step setup like the Y axis? Based on your voltage test, I'm still leaning towards too much weight on your X axis for the motor, or it's binding somewhere during its travel.

Summary of my problem: Home made Prusa i3 prints demonstrate a shifting in the x axis during prints that leads to “wavy” appearance in the X direction. The Y direction is fine. A clue may be that the AXIS_STEP_PER_UNIT is different for the X and Y-axis even though geometrically identical.

The prior post shows that using the calculated AXIS_STEP_PER_UNIT of 200, the distance the X axis moves varies directly with the voltage applied. In fact it is quite linear. I thought I’d better do a similar experiment with the Y axis. The AXIS_STEP_PER_UNIT of 200 was used for both the X and Y values in the experiments below. A 100mm move requested for the Y--results shown below.

Y motor with Y board connections
Voltage/Distance
0.2/100
0.71/100
0.97/100
1.16/100
1.33/100
For the Y motor and direction, the distance is independent of the amount of Voltage on the driver. To further investigate, I put the X motor connections on the Y driver on the board. Results:

X motor with Y board connections
Voltage/Distance
0.2/100
0.54/100
The X motor and the carriage assembly can move as it should (Yay!).

Y motor with X board connections
Voltage/Distance
0.68/52
In this configuration, the Y motor is now showing the problem demonstrated in the original X configuration (Boo!).

I have already demonstrated the problem persists when I changed Ramps and Arduino boards. I also tried unzipping a fresh Sprinter firmware and installed. No change.

I think this experiment demonstrates that the X-axis carriage assembly/weight is not the main factor in my problem, but thanks so much for the input. Any other ideas would be appreciated.

If the X axis works properly when you connect it to the Y driver, and the Y axis has the same issue when you connect it to the X driver, then you have a bad stepper driver plugged into X right now. Replace your stepper drive that is plugged into the X axis location on the Ramps board.

I believe my print waviness issue is directly attributable to the fact that the x-axis move distance is directly related to a slight change in voltage in the driver during the print. A slight change in voltage would change the distance the x carriage moves and could easily account for the waviness in the prints in the x-axis direction. The data shows that changing voltages on the Y driver does NOT change the distance the y assembly moves. If the x axis acted like the Y axis, my problem would be solved.

It was suggested to change out stepper drivers in effort to see if that would solve my issue. I used 4 different stepper drivers (designated red, green, yellow and blue-because I put a small dab of paint on them to distinguish between them) and asked the X and Y-axes to move 100mm. Below is the data.

Y axis
Yellow driver V=0.54
Distance 100mm

Red driver V=0.63
Distance=100mm

Blue driver V=0.75
Distance=100mm

Green driver V=0.76
Distance=100mm

Data shows good results for all drivers for the Y axis

X-axis
Yellow driver V=0.54
Distance =40mm

Red driver V=0.63
Distance=50mm

Blue driver V=0.75
Distance=60mm

Green driver V=0.76
Distance=44mm

Back to the drawing board—New suggestions?

X axis shifting issue problem solved! After seeming to eliminate all mechanical and electrical possibilities, the only possibility left was the firmware . As I indicated in my initial post to this thread, I have an Arduino Mega 2560 board with Ramps1.4. Following the video instructions to build this Prusa I3, I installed Arduino version 0023 on the Arduino board and Sprinter firmware from 9/3/2013. I downloaded fresh Sprinter firmware (apparently not updated since my first download) with no improvement. So I tried Marlin 1.1.0-RC6 (dated 5/4/2016), but it wouldn’t compile. But Marlin info indicated I needed to update the Arduino mega software to 1.0.6 or 1.0.5. –but after install still no Marlin compile. But after I downloaded the newest version of Arduino software 1.6.11, Marlin did compiled and I was able to download to my Arduino board. Test results showed the x axis moved the desired distance! And after recalibration of all axis, the test cube came out right on with no print shifting. I could find nothing in the Sprinter code to account for the problem—but there must something that and expert could determine.

I have learned much in this process. Not mentioned previously, was the fact I struggled because my extruder kept getting jammed many times. I was using 1.75mm PLA with an E3DV6 hotend with 0.4mm nozzle. I was having to torch the nozzle and hot end heat diffuser every other (shifted) print. This problem was solved by a trip to a local 3D printer manufacture, Aleph Industries, maker of the Lulzbot. There I bought a new extruder hot end (Lulzbot v2 Hexagon) with 0.5mm nozzle. The filament they recommended is 3mm nGen filament, a polyester with printing temp 230C, bed temp 80C. I had to use my prior heating element because the new hot end heater was for a 24V supply and I have a 12V power supply (it would not heat up in the time I had available). But the good news is that I have not had a nozzle/hot end jam since.

This is why spending a ton of cash on a printer doesn't pay (IMHO). You wouldn't have learned a thing had it worked perfectly. I maintain that all hobbyists "should" buy the $300.00 piece of crap (like I did), battle like hell to make it work and then build the dream machine. I recently looked back at my original posts about how to figure out stepper wiring, little did I know that that was the easiest problem I would encounter. Since then I've redrawn and printed and replaced almost every part. I resolved a stripping issue lately and found that a perfectly working printer was "almost" disappointing. Fixing it gave me purpose. The battle was the challenge. Weird huh?
Nice job man.