Share:
Share
Notifications
Clear all

Print Bed Stability

hazmatt99
(@hazmatt99)
Trusted Member

Hey guys!

I'm having a hardware problem and I'd like some tips nailing it down.

So below I have a picture of the results of running my auto-leveling program on a TRONXY XY-2 PRO (the values are just readings taken by the Z-axis sensor at various points on the bed, it uses the bottom left corner of the print bed as a reference point, so -0.13 means that point is 0.13mm lower than the bottom left corner)

image

And these are the results from running it again, right after the first one, without touching anything else on the machine. All 3 corners (or the reference corner and one other) shifted by AT LEAST 0.25mm in different directions. With a layer height of 0.2mm, you can imagine how this could cause problems lol.

image

I spent some time getting the bed leveling as perfect as I can and then this kind of thing happens. I thought it might just be a faulty Z height sensor but I'm not entirely sure since the effect on the print quality shows a similar change. Below you can see that even within the same first layer the height varied quite a bit at different times in the print. (Red arrow = skirt, Yellow arrow = print before I noticed it was printing to close and adjusted the corner)

image

I'm reaching out to tronxy support as well but I was curious if you guys had troubleshooting tips to get a more stable print bed. It seems like I need to adjust something for every new print.

For context, the XY-2 PRO is built very similarly to the ender 3, except it comes mostly assembled and it features a z axis sensor which I think detects the metal in the print bed.

Could be an eccentric nut is loose? maybe the z axis lead screw is messed up? Maybe it is the sensor? Maybe I need stronger springs for the bed? Should I switch to a glass print bed? I'll take any advice you got this is driving me up a wall lol.

Quote
Topic starter Posted : 06/08/2021 7:01 pm
Pablo
(@placherre)
Member Moderator

Hi Hazmatt99!

Glad to have you back…

I am sorry to hear that you are having issues with your printer. I will help you with this problem.

I apologise for the late response. But I was doing some research regarding the level sensor, which is different to the BLTouch I am familiar with.

When I verify the data from the auto levelling values, I notice a kind of pattern. In both cases, I notice the distance to the bed decreases from the back left corner (if watching the printer from the front) to the front left corner and the back right corner. Therefore, I assume there is an issue related to the mechanical parts. Both arrows in the following picture show that pattern.

image

I would recommend you to check the eccentric nuts placed in the bed and the gantry. If you check the guide wheels, you will notice that some of them have an hexagon spacer, that is the eccentric nut that tightens and fixes the bed and the gantry, in order to prevent them from being loose and wobble when moving (due to accelerations). If that is the case, then the distance between nozzle and bed won’t repeat after every level control. In the following pictures I will show where are located:

image

 

When looking at the manual, I can’t tell if the bed has two eccentric nuts. But, since it is similar to the Ender 3, I suggest you check both right side wheels; because both should have this type of nut. I recommend you not to adjust them too tight, because you will end up having issues with the stepper motors and the timing belt; due to the increased resistance you applied when tightening. The final adjustment should be enough that you can move the bed or the hot end with your hand without any extra effort. Use the same criteria when checking the gantry nuts.

When I was searching about the sensor, I didn't find much information. But, I could say the sensor detects a metal surface, which is the build plate. Unfortunately, I wasn’t able to gather more information about it. I would recommend you check both screws if they are tight. Because it might come loose and affect the readings afterwards. Also, if the lower face of the device is not parallel to the bed this could also affect the readings.

If the springs are not fully compressed, you shouldn’t have issues with that. Because they have room between coils in order to work properly; but make sure the adjusting knobs are not loose. If you notice that, the spring will be loose; thus, affecting the level of your bed. 

Just for now, you could stay using the flexible bed. If you decide to change to a glass bed, that will be decided upon surface finishing for example. 

But, before going further, I have a few questions:

  • When you refer to the print quality, is it only in the first layer or only when printing the whole model? In order to check if the problem is related to the Z-Axis movement.
  • After performing the manual bed levelling, do you adjust the Z-Offset? In that case, is the auto levelling is not consistent either? Regardless the actual difference is below 0.8 mm, take a couple of minutes to check this value.
  • Are you performing the auto levelling with the bed or nozzle heated?

Once again, sorry for the late response…

I am looking forward to your response.

Kind regards,

Pablo

ReplyQuote
Posted : 09/08/2021 4:18 pm
hazmatt99
(@hazmatt99)
Trusted Member

Hello Pablo!

First I'll answer your questions:

-First layer is where the problem is

-I haven't done manual leveling in a while because I was trying to get the auto-leveling to work because it's faster and I'm lazy. Question for you, Is the auto leveling supposed to tell the machine to make adjustments during 3D printing? So like in the top left corner of the picture you have shown, would it read -.13 and make an adjustment of +.13 when it gets to that corner? I figured it was just letting you know how much each area was off by so you could make adjustments but wanted to confirm.

-Yes I always do auto-leveling with the bed heated, when I do manual leveling or z offset adjustment I do bed and nozzle heated. I just don't do auto-leveling with nozzle heated because the nozzle doesn't come into play here it's just the sensor and the bed.

 

Ok so here's what I did following feedback from tronxy support and my own curiosity.

I noticed that pushing on the print bed in certain spots was making a noise, after some investigation, I found that the bearing what shifting slightly within the wheel on the front right print bed wheel. I tried adjusting this as much as I can (that eccentric nut was a little loose) but that issue persisted. (Also yes the print bed has 2 eccentric nuts)

I concluded though that this wasn't that severe of an issue because if I push on any of the print beds for any of my 3D printers it will cause the bed to shift fairly significantly so maybe part if it was just me putting pressure on the print bed accidentally? Of course in the case where I just ran auto-leveling back to back without touching anything else that wasn't the case but maybe the nut being loose was the cause in that case. I may have made it too tight though so I'll have to go back and check that.

TBH I'm also unsure how I should be adjusting the left and right slicer nuts. The side of the support bar opposite the z-axis lead screw (im assuming this is the right slicer) was able to shift up and down slightly but I don't know how much pressure is too much for those. It's especially confusing because the lead screw provides extra support to the other side so I wouldn't really be able to tell if that one is tightened properly without disconnecting the lead screw.

When I brought up that the z offset may shift between prints on one of them, tronxy support said it's possible that "the leveler has an induction error", I haven't responded to them yet but I'm curious what your thoughts are on that problem and how to confirm it or fix it.

 

Ok so far for general stuff for improving stability.

  • Make sure all the nuts are tightened to the point where there's no play but there's no extra resistance when manually moving the part.
  • When leveling, do manual level > Z Offset > autolevel? There a reason for this specifically?
  • Make sure the springs aren't super compressed and that the leveling knobs aren't loose.
  • Make sure the screws are tight on the leveling sensor

Good so far?

ReplyQuote
Topic starter Posted : 09/08/2021 7:09 pm
Pablo
(@placherre)
Member Moderator

Hi Hazzmat99!!

Thank you for your detailed response!

The main idea of the auto-levelling is to forget about doing manual levelling. Despite this, it is a good practice to check the bed level manually in order to verify that everything is in place and there wasn’t any movement of the bed. Regarding your question, yes. The auto-levelling tells the printer to adjust the Z-Axis height in order to compensate for the variations measured along the bed.

The importance of checking the manual levelling also is related to the adjustment of the Z Offset. This measure compensates the distance of your nozzle to the bed (in the order o a piece of A4 paper); thus, the auto-levelling will have a reference when performing this action. In overall, the Z Offset sets the 0 for your automatic bed level system. That is why it is important to check this parameter. In your printer, select Level (green anchor icon) and wait for the auto-level to finish. Once it is done, go to Z Offset and wait until the nozzle is placed in the middle of the bed. Next, try to place a piece of A4 paper and move up or down the Z-Axis until you are able to move the paper with some resistance. I would recommend you to move the nozzle height in 0.1 mm or 0.01 increments. Finally, when you find the correct distance between nozzle and bed, click the Set Z as 0 icon. Follow this procedure even if the points measured in the auto-levelling are under 0.8 mm. 

In conclusion, I would suggest you to perform the manual bed levelling, then the auto-levelling and finally the Z-Offset adjustment. Then, verify with auto-levelling once again.

Also, I would recommend you to perform the auto-levelling with the bed cold (ambient temperature actually). Since the technology of the sensor is based on magnetic detection, temperature could affect the readings of the probe. Also, the manual doesn’t mention to heat up the bed when doing this procedure; although it is a good practice to set the components to working temperatures for a good bed level. For example, when using the BLTouch probe you can heat up the bed when performing the auto-levelling due to its technology (Hall effect).

When you press the bed, you will notice it bends. That is because you are pressing over the springs (if I am not mistaken of what you said). In order to check if the eccentric nuts are tight enough and the bed is not loose is to try to move the under part of the bed (where the wheels are fixed). Maybe there are two or more issues affecting your bed levelling. By adjusting the eccentric nut properly, you have eliminated one of the causes. If the nuts are too tight, you will find a resistance to move the bed with your hand. After a couple of time of tightening and loosening the nuts you will find the correct adjustment. When I did it the first time, I had to adjust the nuts a couple of times before finding the correct adjustment. 

Regarding the gantry eccentric nuts, it applies the same criteria. Make some minor adjustments to those nuts until you can find the correct adjustment. You are right when you mention that the Z Axis rod adds more stiffness to the whole gantry. If the eccentric nuts are loose, you will notice some small movement when applying some force with your hand. On the other hand, if they are too tight, you will notice some resistance when moving the gantry (by turning the rod from the coupling). Before making further adjustment, apply a little force with your hands and check if there is any movement where the wheels are placed. I am sharing the following video showing how to check the eccentric nuts in the gantry (it also shows the bed): https://www.youtube.com/watch?v=GsEdU8ZtI6U . This video was made using an Ender 3, but it will work fine for you due to the similarities with yours.

Posted by: @hazmatt99

When I brought up that the z offset may shift between prints on one of them, tronxy support said it's possible that "the leveler has an induction error", I haven't responded to them yet but I'm curious what your thoughts are on that problem and how to confirm it or fix it.

About this answer, I believe it is quite vague. If there is an induction problem, therefore there is a fault in the probe. Hence, you will have to replace it. Another meaning could be related to some electrical interference affecting the readings made by the sensor. In that case, the recommendation of turning off the bed heater when doing the auto-levelling will do some of that.

Posted by: @hazmatt99

Ok so far for general stuff for improving stability.

  • Make sure all the nuts are tightened to the point where there's no play but there's no extra resistance when manually moving the part.
  • When leveling, do manual level > Z Offset > autolevel? There a reason for this specifically?
  • Make sure the springs aren't super compressed and that the leveling knobs aren't loose.
  • Make sure the screws are tight on the leveling sensor

Good so far?

Regarding what you mentioned, I only made one change, accordingly to the following comments:

First bullet point: Correct.

Second bullet point: Manual Levelling > Auto-levelling > Z-Offset Adjustment.

Third bullet point: Correct.

Fourth bullet point: Correct.

Let me know how it goes and if you have any other reply from TRONXY support.

I am looking forward to your response.

Kind regard,

Pablo

 

ReplyQuote
Posted : 10/08/2021 1:43 pm
hazmatt99
(@hazmatt99)
Trusted Member

Nice, ok so I'm going to quickly confirm your info, and then see if I can get your input on a few other related hardware issues I've noticed.

FIRST:
Steps to level a bed:

Manual leveling (hot nozzle and bed)
Auto leveling (ambient temp)
Z offset, what temp? Ambient?
Auto level to confirm (ambient temp)

Confirming, so the eccentric nuts should only be tightened enough that applying force or torque by hand doesn't make it wobble, so it should ONLY move if I push in the direction it would naturally move on the track and this movement should be smooth and not require much force? Pushing in any other direction shouldn't make it shift at all? Use light pressure in all cases I'm guessing.

Few other things I've noticed:

I may have made a mistake and messed with the tightness of the fixed nuts and bolts in addition to the eccentric ones, will this cause problems? Is there a way for me to get back to the correct adjustment for these? I noticed after messing with them that no matter how much I turned the eccentric nuts they wouldn't grip the track so I troubleshot it until they did grip the track but now I'm worried that some of these fixed nuts may be too loose or tight.

Some wheels I'm able to turn by hand and it will cause the bed/gantry/print head to move along the track, I assume this is a good thing and means it has a tight grip (maybe you can tell me if that means its too tight). I did notice though that on the right side of the gantry, 2 of the wheels share this characteristic and one doesn't, this leads me to think that the gantry may not be level? There a way to confirm/fix that if its the case?

I'm getting a sound and some slight vibration coming from the print bed belt near the motor, the sound is like if you were scratching a zip tie with your nail. I'm assuming this means that all the screwing around I did with my wheels messed with the belt timing and tension?

Thanks again!

ReplyQuote
Topic starter Posted : 11/08/2021 3:49 pm
Pablo
(@placherre)
Member Moderator

Hi Hazmat99!

Posted by: @hazmatt99

Steps to level a bed:

Manual leveling (hot nozzle and bed)
Auto leveling (ambient temp)
Z offset, what temp? Ambient?
Auto level to confirm (ambient temp)

The Z-Offset will be done in ambient temperature, since it is one step further from the Auto-Levelling. Regarding Manual Levelling, it is correct to do it at working temperatures (hot Bed and Nozzle). Finally, when performing Auto-Levelling again, do it at ambient temperature. I recommend you to check the Auto-Levelling at ambient temperature because the heat from the bed might be affecting the sensor readings.

Posted by: @hazmatt99

Confirming, so the eccentric nuts should only be tightened enough that applying force or torque by hand doesn't make it wobble, so it should ONLY move if I push in the direction it would naturally move on the track and this movement should be smooth and not require much force? Pushing in any other direction shouldn't make it shift at all? Use light pressure in all cases I'm guessing.

What you described is correct. That is the best way to verify the correct adjustment of Bed and Gantry.

Regarding the fixed nuts and bolts, make sure they are not loose. If they are, tighten them firmly; but do not apply to much force. In order to get a better approach in this issue, would you post a video and some pictures? I will get additional information to what you described.

I am looking forward to your response.

Kind regards,

Pablo

ReplyQuote
Posted : 11/08/2021 4:35 pm
hazmatt99
(@hazmatt99)
Trusted Member

Sure! Buuuut you probably won't hear from me until next week, I'm about to go out of town. Won't even really have time to make these adjustments unfortunately.

ReplyQuote
Topic starter Posted : 11/08/2021 4:50 pm
Pablo
(@placherre)
Member Moderator

No problem at all! We will keep in touch when you get back... 😀 

Have a safe travel!

Kind regards,

Pablo

ReplyQuote
Posted : 11/08/2021 5:06 pm
hazmatt99
(@hazmatt99)
Trusted Member

I have returned!

I got to try your bed leveling steps, nearly perfect on the first try, I'm gonna see how long it lasts or if long term stability is still an issue. One of my printers was a little odd and ended up pushing the nozzle closer to the bed towards the right side, this was reflected a bit in the autoleveling, the right side of the platform was significantly lower than the rest of it so maybe it over compensated?

20210820 064618

I'm fairly sure all the extra issues I was talking about are resolved or just non-issues with the exception of the y axis motor grinding.

I tested how level the gantry is and I guess all of my gantries are actually higher on the side opposite the lead screw? Which I assume was done on purpose to compensate for the print head weight?

The wheels are tight enough nbd there.

Here's a video of the motor grinding (ignore the crinkling sounds that's just the cable sleeves)

Lemme know what you think!

ReplyQuote
Topic starter Posted : 20/08/2021 12:26 pm
Pablo
(@placherre)
Member Moderator

Hi Hazmatt99!

Welcome back… 😀 

Posted by: @hazmatt99

One of my printers was a little odd and ended up pushing the nozzle closer to the bed towards the right side, this was reflected a bit in the autoleveling, the right side of the platform was significantly lower than the rest of it so maybe it over compensated?

This explanation refers to the picture of the print? The auto-levelling maintains the separation between the Nozzle and the Bed at all times along the Build Platform. When the probe takes all the measures, it takes in account the height differences and compensates the Z height for the first layers in order to keep the separation between the Nozzle and the Bed. Just to make sure, I would recommend you to perform twice the auto-levelling in order to verify the repeatability of the measurements. Since you have everything nice and tight. Also, it is important to verify the Z-Offset because the printer uses this value to adjust the distance between the Nozzle and the Probe (by setting a starting point or 0).

image

Regarding the gantry level, it should be even on both sides. I would suggest you check the distance between the X gantry beam and the top beam (like in the following picture). in order to give a dimension to that difference and see if there is any action to be taken. Since the assembly is rigid and the Hot End weight is not significant, this doesn’t affect the structure. But it does affect the acceleration and jerk configuration when setting up the model in the Slicer (this is another story). Use a Vernier caliper if possible, in order to achieve a measure with 0.1 dimensional accuracy (a digital one will make the reading easier). If you don't have one, you can use a ruler.

CR 20 Pro   X Gantry Measure

Getting back into the stepper motor, the source of the sound seems to be the roller bearing of the motor. Also, it could be the bearing of the small pulley right next to the motor. I would suggest you loosen up the timing belt until the pulley of the stepper motor is free from the belt. Then, turn the shaft of the stepper motor and check if it makes the same noise. In this case, it might sound quieter, since it doesn't have the timing belt tension. If it makes the same noise, unfortunately you will have to replace the stepper motor.

To put you in context, every motor has two bearings that supports the shaft and allows it two spin, adding the least resistance possible (in order to prevent loss of efficiency). From what I heard, the bearing seems to be broken. by broken I mean it has an excessive wear that causes a misalignment of the shaft. This wear might be affected by a high tension of the timing belt, forcing an extra load on the bearing. In the following picture I will show the exploded view of a NEMA 17 stepper motor, showing both bearings.

NEMA 17 Stepper Motor   Exploded View

The NEMA 17 is the standard used for stepper motors for CNC machines, 3D printers and any equipment that needs controlled movements. In the back of the motor you will see the Model Number. Please send me a picture of that and I will show you how to order it.

I am looking forward to your response.

Kind regards,

Pablo

ReplyQuote
Posted : 20/08/2021 10:44 pm
Share:
Share

About us

We’re a bunch of 3D printing and design nuts. We just want to make great 3D design available to everyone. 

Our mission

Good design takes time, great design takes a process. You can learn this with us, to build your skill set in this rapidly expanding market. 

Learn with us:

© Copyright: 2019 io3dp.com
Close Menu