[BUG] First layer on MK4S is too low when using stock settings

[BlueFyre]

Printer model

MK4S

Firmware version

6.1.3

Upgrades and modifications

None

Printing from...

Octoprint

Describe the bug

The first layer when printing using stock profiles is too low on a large portion of the bed

How to reproduce

Print a sheet on the bed (250x210x0.2) using 0.2mm SPEED and Prusament PLA settings This can be reproduced using any of my sheets and I have tried various brands of PLA including Prusament

Expected behavior

Perfect first layer

Files

STL and 3MF attached: Sheet.zip

This is printed with stock values (230 C and 40 mm/s)

[BlueFyre]

Also to note, I tried printing this with other settings and sheets to compare. Lower temperature and speed did help but it's still not perfect

Smooth sheet, 215 C, 20 mm/s

Satin sheet, 230 C, 40 mm/s (Ignore the top right corner as the sheet is just cooling and detaching from the bed)

Satin sheet, 215 C, 20 mm/s

Satin sheet, 215 C, 40 mm/s

[BlueFyre]

I also noticed there's a related enhancement request #3306 that seems to "fix" this issue by making a 0.04 (positive) offset but that doesn't seem like something a user should need to do to

[frasier2k]

I can also confirm that after the upgrade to MK4S + HF Nozzle the first layer is too low. After setting a +0.04mm offset in the slicer it is ok. I used a satin sheet.

[Shushuda]

I can confirm this. Here are my results, the broken corners are just testing layer strength, used Prusament PLA Azure Blue, MK4S with a regular obXidian Prusa nozzle (NOT HF):

Smooth, 230C 40mm/s

Smooth, 215C 40mm/s

Smooth, 215C 30mm/s

Smooth, 215C 20mm/s

[simao]

I have the same issue. This happens on MK4 and MK4S (two different printers). Tried new load cell, new thermistor, new PEI sheet, multiple brands of PLA, always the same. Photo is with prusament.

[Shushuda]

I forgot to mention that my issues started after moving to Input Shaper firmware and profiles (which supports my findings regarding temp + speed settings). I've since then replaced the load cell, replaced the thermistor, tried multiple smooth sheets (I own 3), upgraded the printer to MK4S. Issue persists.

The only way to stop it in my case and get a perfect first layer is move back to 215C 20mm/s settings used before IS was a thing.

[m00dawg]

Somewhat relates to https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3306 That feature request would be a workaround by being able to set live Z on the printer and have it persist instead of having to put it in the slicer, which is what I'm doing now. That works but I'm worried I'll need specific values for multiple MK4s (which has prevented me from swapping any more MK3s to MK4s). Trying to dig through the source code for where I can simply add a load/save into the eeprom. Made some headway but not a lot.

Notably, it looks like Prusa tried to add in multiple sheets for the XL (and that feature does save live Z) but ran into issues based on code comments. I suspect that is one reason we don't have multiple sheets or savable live-Z offsets on the XL or MK4.

[qlery49]

I got the same issue does someone know how to solve? ![Uploading IMG_0539.jpeg…]()

[JoeScotty]

Same issue. Workaround by adding 0.05mm Z-offset in prusaslicer. If there is some load cell result differences from individual loadcells, there should be a setting at the printer, where the offset could be stored permanently.

[n-dony]

Did any of you experience Z-Homing failure ? https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3392 Could Interference during probing, offset the reading by 0.05 mm and a wrong MBL to this issue ?

[BlueFyre]

I haven't had any issues with z homing. It works normally/reliably to my knowledge. I can actually print the entire sheet repeatedly and get the same result each time so it seems like it's consistently producing the same mesh as well

[peterbenedicic]

Same problem, contacted Prusa chat, they sent me new load cell and hotend thermistor, completely rebuilt Nextruder with new parts, but issue with first layer rippling is exactly the same. As other already pointed out, its the same no matter which sheet or filament you use (including prusament). So probably not mechanical problem, something changed in firmware? And yes, +0.04 correction solves it.

[JoeScotty]

For getting some clarification I printed a a sheet with all original prusa profiles containing 3 square lines with a layer height of 0.2mm and extrusion width 0.8 (the only change against original settings). Printer is an MK4S.

The print looks like

Than I measured the thickness of the lines using a Mitutoyo micrometer getting a 7x7 result matrix. Since layer heigth is set to 0.2mm, the values should be cloe to this value. In the matrix i used µm instead of mm (0,2mm = 200 µm)

The result:

The matrix includes average per line and per column as well as the average over all: 0,167mm

I think, the offset is caused by load cell characteristics / bed pressure during test and just needs an offset. The offset can either be set in Prusa slicer at all printer profiles (separate for every nozzle size) But since its a hardware / firmware issue, the correct location is the firmware.

I found the M206 command where offsets for X, Y, Z can be defined. Tried to store this setting in the EEPROM, but at the next test print the EPROM value were set to default again. Think the EEPROM content is kind of protected, since changing values without the proper knowledge can ruin your printer.

Spoke with prusa chat: They know about this problem and a solution for entering the Z offset via display is in discussion. And I should get an answer, if it is possible to change the offset.

[andrem1987]

Same problem, contacted Prusa chat, they sent me new load cell and hotend thermistor, completely rebuilt Nextruder with new parts, but issue with first layer rippling is exactly the same. As other already pointed out, its the same no matter which sheet or filament you use (including prusament). So probably not mechanical problem, something changed in firmware? And yes, +0.04 correction solves it.

Same for me. I contacted the support, they sent me a hotend thermistor and a load cell as well, but that didn't improve anything. Aftet setting a positive 0.045 Z-Offset the first layer was perfect again. The possibility to set a persistent Z-Offset back again would be a good workaround for now.

[BlueFyre]

I did some additional testing with not doing the probe at a lower temperature This is using a satin sheet with probe at 230 C and first layer temperature at 230 C It looks perfect to me so I am thinking perhaps it has to do with thermal expansion of the hotend

I don't think this is a workaround since I had to set the nozzle to temp, wait a while for the ooze, clean it up, and then print Also I have the MMU so no filament is loaded until purge. The ooze was some residual from the previous print It's more or less just extra data that might help point us towards a solution

[simao]

I did some additional testing with not doing the probe at a lower temperature This is using a satin sheet with probe at 230 C and first layer temperature at 230 C It looks perfect to me so I am thinking perhaps it has to do with thermal expansion of the hotend

I don't think this is a workaround since I had to set the nozzle to temp, wait a while for the ooze, clean it up, and then print Also I have the MMU so no filament is loaded until purge. The ooze was some residual from the previous print It's more or less just extra data that might help point us towards a solution

This is somewhat consistent with my experience. If I reduce the temperature, the diference of temperature to 170 is smaller as well as a slower flow, so it reduces ripples. When I decreased the probing temperature to 160, 150, 140 to reduce oozing, this made the ripples worse, presumably because of a bigger difference in temperature?

If the firmware knew which nozzle is being used maybe it could account for this.

The only way to consistently get rid of the ripples for me is to reduce the first layer temperature by 10-15 degrees, depending on the filament. Raising the z-offset causes other problems in other parts of the build plate, like small holes next to the edge lines.

[JoeScotty]

@BlueFyre Interesting theory regarding the thermal expansion!

I just checked the magnitude of the thermal extension of brass using the calculator: https://goodcalculators.com/thermal-expansion-calculator/ The entered values were a length of 20mm (hot zone), temperature difference of 50 °C and brass material. The result was an extension of 0.0187mm, which is about half of the Z-offset value 0.04, which was posted above several times. This explains about 50% of the cause of the to small first layer height!

I think, the workaround with entering a Z-offset in prusaslicer at printer settings (and save this a custom printer) is a better then calibrating with high nozzle temperature and having to deal with the ouzing and ripples. The first layer MUST have the correct height as a first step!

The solution is the possibility to set a Z-offset in the firmware by Prusa. Up to now I'm still missing the promised answer from the prusa chat technician.

[BlueFyre]

The only way to consistently get rid of the ripples for me is to reduce the first layer temperature by 10-15 degrees, depending on the filament. Raising the z-offset causes other problems in other parts of the build plate, like small holes next to the edge lines.

Reducing the first layer did seem to help a bit for me as well (as shown previously) but it's worth calling out that this creates added knock on effects. My other layer temperature is usually set higher than the profile (245 C) in order to avoid the glossy/matte filament transitions and the clog detection going off. Depending how fast it's printing, the temperature was too low to achieve the requested flow. However, there's no waiting to transition from first layer temp to other layer temp which means it is more likely to go matte or clog. It's probably not waiting to avoid ooze in one part of the print

Re: Thermal expansion and offsets If it is thermal expansion then it makes me wonder when that offset would need to be adjusted. For example, if we increase the print temperature, does it mean the offset changes? Given the first layer speed we'd more likely run into over extrusion first for PLA. However, it also would be worth thinking about what happens when we swap the temperature for different materials

Anyhow just sharing my two cents and hoping someone smarter will come up with a better solution to all of this

[BlueFyre]

Additional idea:

• Collect a large set of offsets by testing combinations of nozzles, nozzle types, probe temperatures, print temperatures
• Generate a formula to determine the appropriate offset (eliminate extra variables as needed)
• Add a persistent way to adjust that offset manually

Hopefully the hardware is consistent enough such that this can be done

[Mike-FUT]

MK4S / 6.1.3 My first layers look the same. Doesn't matter if I use Prusament PETG, Pruasament PLA or any other non-Prusa filament. Flow calibration doesn't improve it.

Below is a picture of a PLA Silk print where it's extremely visible. With other materials it's the same but less visible.

But... even my first layers look awful I actually can't complain about height. My base layer is set to 0,2 mm in the slicer. I recently did a print of 24 mm rount objects that I filled the whole bed with and added a 5 mm brim around each of them. After removing the brim circles I measuered the thickness of the 5 mm brim circles with a Mitutoyo calliper and all measurements were between 0,20 and 0,22 mm

[Gman2606]

Same thing here 0.04 offset makes it perfect again, at first I was sure it's thermal expansion (I still do) But support says they took it into consideration in the profile's and they're a little clueless as it seems now..

[Gman2606]

The only way to consistently get rid of the ripples for me is to reduce the first layer temperature by 10-15 degrees, depending on the filament. Raising the z-offset causes other problems in other parts of the build plate, like small holes next to the edge lines.

Reducing the first layer did seem to help a bit for me as well (as shown previously) but it's worth calling out that this creates added knock on effects. My other layer temperature is usually set higher than the profile (245 C) in order to avoid the glossy/matte filament transitions and the clog detection going off. Depending how fast it's printing, the temperature was too low to achieve the requested flow.

However, there's no waiting to transition from first layer temp to other layer temp which means it is more likely to go matte or clog. It's probably not waiting to avoid ooze in one part of the print

Re: Thermal expansion and offsets

If it is thermal expansion then it makes me wonder when that offset would need to be adjusted. For example, if we increase the print temperature, does it mean the offset changes?

Given the first layer speed we'd more likely run into over extrusion first for PLA. However, it also would be worth thinking about what happens when we swap the temperature for different materials

Anyhow just sharing my two cents and hoping someone smarter will come up with a better solution to all of this

About thermal expansion, offset will be calculated in comparison to the probing temp and added through out the whole print for example I'm printing at 230 the first layer so Z height=calculated offset. If I move to 240 in the middle of the print then Z height=new calculated offset+layer number sum.

I don't know if it's already happens probably does. But maybe someone did some bad math...

[feyded1020]

I thought I was going crazy, I'm having similar issues with waves on my first layer. Using original parts and an ObXidian Prusa nozzle on my MK4S with Kingroon PLA on generic PLA preset.

Printing temp is 230 for first layer, with 220 for all other layers. Scratching my head in regards on what can be done to achieve the claimed 'perfect first layer'.

[mix579]

I really hope this is numero uno on the list for Prusa software engineers as it complete destroys the myth of "perfect first layer". There's not one print anymore where I don't have to adjust live z manually. I don't know if it was upgrading from Mk4 to Mk4S or a recent firmware upgrade that started it. But initially I didn't have problems with my Mk4 printers, at least with PLA (PETG always needed a -0.012 adjustment). Now even PLA gives me the waves so nicely documented in this thread. The Mk3.5 I kept on the other hand--perfect first layers.

[feyded1020]

Here's my current print at 225C with Kingroon PLA with ObXidian Non HF Prusa Nozzle on my MK4S. This is with zero offset adjustment.

[feyded1020]

Reprinted by changing offset +.04 as mentioned in this thread. This is the result, and it really can print a perfect first layer, just not out of the box like its competitors can... Prusa please fix, or figure out where the discrepancy lies hardware wise!

The last pic was the only glaring issue I saw, and that's where the first layer 'should' have started its print.

[JoeScotty]

In my experience the most important parameter for reaching a good first layer is the thickness in the complete bed area. The thickness will vary depending on the XY positions of the bed, since the bed is not really flat. And bed levelling does not completely balance this. A certain tolerance level is usually accepted.

Best reliable method I found is to print a square line test sample consisting of single lines using a smooth build plate. If you have two or more lines side by side and the layer thickness is to low, the filament ist pressed upwards beside the nozzle tip and the measurement results in a too highvalue. The line width should be around 0,8mm if using a 0.4mm nozzle to make sure the line is squeezed on top and bottom side and have no higher shoulder. Compare my post above.

Than measure the thickness using a good caliper or a micrometer and compare the value with the set first layer in the slicer. If the thickness is in a sufficient tolerance field, then its recommended to play with other parameters like extrusion rate, bed temperature, nozzle temperature, print speed, bed cleanliness.

[mshipman]

I've been pulling my hair out over this exact issue for days, ever since performing the MK4 -> MK4S upgrade with the new HF nozzle. My MK4 always had perfect first layer performance, but after the upgrade I was seeing exactly the pictured rippling on every first layer (mostly using Hatchbox PLA with the provided profile in PrusaSlicer on a new, smooth PEI sheet).

Visually, I thought it looked like the classic first-layer-too-low scenario, but that's not supposed to be an issue with the load cell... so I spent hours dropping temperatures, tweaking the flow rate, slowing the printer down, and trying different materials (usually resulting in about 50% improvement) but the underlying problem still remained.

I finally relented and hardcoded the +0.04mm Z-offset in the slicer, and it's back to printing a perfect first layer, just like it did with the original non-HF nozzle.

It'd be great if Prusa could solve this all in software, but if that turns out not to be possible it'll be vital to have that option to hard code an offset in the firmware, again.

EDIT: I feel the need to add that for me, this issue is only with the 0.4mm HF nozzle included in the MK4S upgrade kit. If I install the original 0.4mm nozzle that came with the printer into the same heat block and use the non-HF version of the MK4S profile in PrusaSlicer, the machine goes back to its previous performance (a flawless first layer on default settings, at least in PLA since that's the only material I've been testing with).