UM3E bed leveling - at back of the plate it's always printed too far away

[akke 0]

I have an UM3E with automatic bed leveling enabled but whenever I try to print a big object, the first layer isn't printed evenly on the plate.

 

Let me explain:

At the front of the plate the first layer is squished to the plate as it should. I think it's more or less a perfect first layer.

From the front to the back I can clearly see that the printed skirt gets smaller and smaller.

At the back of the plate it's barely staying sticking on the adhesion sheet.

 

I already tried manual leveling using the menu and using the calibrarion card. After manual calibration I set automatic leveling frequency to "Never" and tried again.

Yet, I got exactly the same problem.

 

So I wonder:  Is the UM3E still applying automatic bed correction when level frequency is set to "Never" (from a previous probe).

If so, can this be disabled? So I can still get it printed without relying on automatic leveling to function properly.

 

I also want to fix the automatic leveling but what can I do?

The spring at the back of the build plate is weaker compared to the ones at the front.

Is this an issues for the automatic leveling system? 

If I'm not mistaken the weaker spring at the back is normal for older UM3E printers. But Ultimaker doesn't speak about it on any web page so I don't think we, as a customer, were supposed to replace it with a stiffer one?

 

Any ideas or tips to fix this issue are welcome. Thanks!

 

 

[Enigma_M4 119]

Hi akke,

 

the problem you described looks like a print bed heavily out of level (tilted backward), so manual leveling should be done again.

Active leveling is not able to reposition the print bed, but only measures how far it is out of level and corrects it at the first 2-3 layers which means: the better the bed is leveled (manual), the less correction the activ levelling has to do (and the less out of level is the bottom of your print).

 

The "weak" spring should be no issue, so it's your choice to keep it as it is or to replace it with a new one (part nr. 1788, spring DR2160).

 

Check, if the distance of the bottom of the heat bed and the top of the build platform is about 14mm (which is needed for active levelling to work properly)

 

Regards

 

[akke 0]

Hi,

 

As explained, I already did manual leveling.. multiple times actually. It doesn’t help.

 

Any more advice?

[akke 0]

Distance is about 14mm so it should be good. Yet, it keeps printing like this every time.

 

I do use ultimaker adhesion sheets but that shouldn’t be an issue, right... 

[akke 0]

Just to make sure the spring isn't causing issues I've ordered the DR2160 spring to test with as soon as it arrives.

 

If anyone has anything else I can try, please speak 😉

[Enigma_M4 119]

Hi,

 

the adhesion sheets are no issue, neither for manual nor for active leveling.

 

When doing manual levelling, be careful not to apply any pressure (up or down) on the platform when adjusting the two front screws (happened to me: adjusted just fine with the calibration card, but when I took my hand off the screw and re-checked, the adjustment was not good and had to be redone).

 

The issue you describe is that the front side of the printbed is closer to the print head than the rear side.

Do the axles of the print head (especially the y-axis) sit properly in the sliding blocks?

 

Regards

[akke 0]

When doing manual leveling I always make sure to not touch anything at all (except the calibration card ONLY) and I'm positive that the friction between the nozzle and the bed feels the same at all 3 points before I hit the "confirm" button. So I don't think I'm doing this the wrong way.

 

I'll check the axles in the sliding blocks in a couple minutes when I get back home and report back.

[akke 0]

I don't see anything wrong with the axles. All 4 sliding blocks are fit perfectly on the axles. I tried pushing on them but it's not moving so to me it looks like it's perfectly fine.

[akke 0]

20 minutes ago, Enigma_M4 said:

The issue you describe is that the front side of the printbed is closer to the print head than the rear side.

 

I'm not totally sure about this. I'm wondering if it's not an auto level correction fault or something.

[Enigma_M4 119]

21 minutes ago, akke said:

 

I'm not totally sure about this. I'm wondering if it's not an auto level correction fault or something.

 

This could be easily checked by observing the z-axis when printing the first layer: if there is a ticking sound and motion of the z-axis when printing the first layer (or first layers), especially in y-direction, then auto level correction is active.

 

As I see you have checked everything and made no mistake when levelling. Unfortunately, I'm now at my wits' end :-(

 

Regards

[akke 0]

My UM3E's active leveling system is somehow broken. It doesn't always work anymore.

Sometimes it starts the active leveling process, the glass moves to the top but doesn't touch the nozzle. Then repeats the same process for the other printcore and comes back with a message telling me the height difference is incorrect and stops.

After repeating a couple times the printer starts it's printjob but with the issue described above.

 

I've ordered a new cable for the printhead as someone told me this could be the problem. I'll give it a try!

[Enigma_M4 119]

Hi Akke,

 

it could be worth doing two checks before exchanging the print head cable:

1. testing the capacitive sensor via maintenance menu

2. checking the sensor cable, as described in the following post:

https://community.ultimaker.com/topic/20852-um3-auto-levelling-not-working/?tab=comments#comment-205252

 

Regards

[akke 0]

When doing the captice sensor test using the maintenance menu most of the time it reports "GOOD" but sometimes it comes back with an error telling me there is some interference.

 

So I replaced the cable and the first few days the problem (where active leveling fails telling me the height difference is incorrect...) did not come back but then suddenly a few days later the problem was there again.

 

It looks like the cable wasn 't bad after all.

 

Now when I try to print sometimes active level fails, sometimes it works. I can sometimes fail like 3 or 4 times in a row and the next time it continues and starts printing.

 

 

But I do have underextrusion problems too now. I've not been able to diagnose it yet but I think the printer is now printing too far from the bed. This could be an active level issue too offcourse.

 

Could it be a broken printhead PCB? If I'm not mistaking this PCB has all the active level sensor logic and maybe I can try replacing this one? .. except it's not cheap 😞 ...

 

 

[Enigma_M4 119]

Hi Akke,

 

this sounds strange.

Of course it could be a defective printhead PCB, but I'd guess rather a broken cable or broken cable connector inside the print head (i guess the connection between the printhead PCB and the capacitive sensor, maybe just a loose plug), as the problem is only occurring sometimes (and this would be solved easier and with less expense).

 

I assume the reported interference is not caused by nearby electromagnetic fields (speakers / TV etc.) or being to close to the printer while active levelling (if I'm wrong, this really could be a source of errors, as the sensor is very sensitive for such disturbance, so keeping hands off the printer while active levelling is mandatory).

 

But, as said, I'm now very deep in the field of guessing. Perhaps someone else reading these posts is able to help?

 

Regards.

 

[fbrc8-erin 272]

It's rare, but I've seen a pinched front fan cable (the black and white one) cause electrical interference with the active leveling. To test this unplug your front fan and run active leveling and see if it works. (Note: don't run your print this way; running prints without your front fan can make the filament swell in your Print Cores.)

[akke 0]

I completely dismantled the printhead, checked the PCB but it all looks perfectly fine. The cable from the sensor to the PCB was fine too but the connectors are a bit loose. I think it's normal but just to be sure I soldered them in place to rule out this as a possible problem at least.

 

After putting everything back together the problem looked solved: I was able to perform active leveling 13 times but the 14th time it failed again. Damn!

 

It's strange. The problem is very random. Some days it works most of the time. Other days it fails most of the time. I have no idea what could be causing it and I don't think there's any other device nearby that could be causing interference.

 

There's a monitor 1.5m away and some other devices in the same room but nothing is touching the printer. The closest thing, next to the printer, is a formlabs form 2 printer sitting there with about 40cm space between them. But that formlabs printer is not even powered on....

 

I'm thinking about replacing the printhead PCB but it's more a guess, maybe it's just throwing away money? but how else can I find the root cause?

 

Also, I do think that auto leveling doesn't work as it should. Sometimes it looks like it's printing too far from the printbed and when I restart the same printjob it's better. I could disable active leveling and see if it still happens, if not: the problem is active leveling. If it still happens it must be something else.

 

But the fact that active leveling sometimes fails with that error about differences in nozzle height is something we should be able to fix somehow...

 

[p-kimberley 4]

@akke I had a similar issue with active levelling failing most of the time and was getting errors like "nozzle probe failed" and "difference between detected height...". Turns out that when I had reassembled the printhead during some maintenance, I had installed the front fan back-to-front. I was getting underextrusion issues as well, which as Erin suggested, can be caused by filament swelling due to cooler air not being drawn in from the front.

 

Once I turned the fan back around, the underextrusion stopped and I'm now able to active level most of the time. Still have the occasional failure, but not like before.

 

Check whether the label at the centre of the fan housing, is facing the rear. If it's visible from the front - the fan's the wrong way around.

[akke 0]

I checked the fan but it’s correctly installed.

A new printhead pcb is on it’s way to me to try....

[V3DPrinting 11]

I have the same issue on a UM3E.

The printer is running for nearly 8 days without any significant stops, just time to unload the print and restart a new one.

 

The fan is correctly installed, no sign of loose connection (no ER17 or ER18).

The next electric appliance is an UM S5 0,6 m away and had been there for a year.

[akke 0]

New PCB fixed the active leveling issue. Or at least: I have not seen the problem anymore since I installed the new PCB...

Yet, the printer is still printing too far from the bed at the back of the buildplate.

 

This is very annoying for bigger prints and I would love to hear what can be done about it ...

[tinkergnome 927]

On 10/13/2019 at 10:01 AM, akke said:

So I wonder:  Is the UM3E still applying automatic bed correction when level frequency is set to "Never" (from a previous probe).

If so, can this be disabled? So I can still get it printed without relying on automatic leveling to function properly.

 

Step 1: set frequency to "Never"

Step 2: do a manual leveling

 

-> active leveling is now disabled

 

hope that helps!