Settings change mid brim?
The first two lines of brim are probably the best 1st layer I've ever put down. On the third line it all goes out the window. It's not like the brim is over, just checked Cura and brim line count is set to 20.
And it's the whole way around the 300mm build plate. So it's not a level issue. Have y'all ever seen anything like that before? Wonder if that's part of why my prints keep failing.
236 Replies
Print failed the next loop after this as it pulled the brim off the printbed.
Hold on gotta trim it
Here we go
Running the print without a brim now, just to see if it does it in general or on a brim specifically
Brim is irrelevant
I know there are settings per layer, is there a setting somewhere based on line count?
You can see the even layers on the right from where it laid down on the print bed, nice and even, to the left hand side going stupid
filament? Settings?
It's ASA, Cura Generic Profile, .2 on a .4 nozzle, 240Ex /110 Bed
Just trying a calibration cube now see if it does it on there as well.
are you doing this enclosed? If so, have you pre-heated the chamber and let everything settle before printing?
Affirmative on both. Tried letting it go for the full first layer totally enclosed, and then removing the front so I could catch a vid of it happening.
how long are you waiting after starting to heat the chamber and before printing?
~10 mins. Pretty sure this is an extrusion issue, no?
no, 10 minutes is not enough
that's your problem. Things are still moving, so as it prints the offset is changing
https://www.printables.com/model/230716-ratrig-v-core-3-extrusionless-x-gantry-mod
you can see from the graph, it can take like an hour to settle
Hmm, for the record, I don't think this is the issue, but I'm more than happy to give it a shot.
Just so we are on the same page, preheat bed to 110 for an hour in the enclosure and try again?
yes
Just find it hard to believe it "moves" at almost the same line count, at different models at different spots on the bed, that much. But hey, I'm learning lol. Got no problem admitting I was wrong lol
Will report back
look at the graph, 0.4mm in 50 minutes. That is enough that when the first layer offset is so critical
I'm not really debating the science there, just the fact my issue happens (just went back and counted on the last two prints) on the 4th line count.
Just seems wayy too consistent to be thermal fluctuation
But hey, we are testing that now
Does RatOS time out heaters by default if there no print queued? Like do I need to turn it back on in 30 mins or so?
I don't remember to be honest
No worries, I'll just have it pulled up during lunch and keep an eye on it
Mythbusted
Issue persists
For the record, the heater never timed out over the hour either
Same line count as well
Lift up your Z offset by 0.05 increments : to much material is extruded for too little height. It doesn't show until critical extra volume is reached, after a few lines, then it produces waves. I'd say lift by at least 0.2 total. Brims and skirts and not ideal to tune Z-offset. Try printing a serie of 1 layer only rectangles, adust on the fly, save_config
The problem persists Brim/NoBrim
I noticed it starting witht the brim, so I made the post. In the middle of troubleshooting I tried a model with no brim and got the same results
This example is a no Brim model
critical extra volume I am interested in this, are you saying flow changes after a certain line count?No. I mean there's enough room to fit the marginal excess material until critical point every X layers.
The model in which you tested this, can we see the result? Are the perimeter lines ok and just the infill lines an issue? Lots of variables in play still - need to start narrowing things down if we can
Strange issue for sure... Not convinced it's just a z offset problem
May just be the photo but those lines look pretty wide for 0.4 nozzle
I did loads of Cali cubes. I'll take some.photos when I return to work.
Alright so random thought about this issue.
1. Hotend melts ASA at 190/200.
2. The bad printing that happens after a few lines
Theory, what's the chances that my thermo is reporting a really off temp reading to RatOS. That would explain the low melting temp of ASA and the wacky flow being heat creep cause nozzle is actually way hotter, 40c more than reporting?
I forgot all about getting you some pics. I'll try to get them when I get back to the shop.
Here's a high resolution for all of them and the associated factors. If you want to see one specifically up close let me know and I'll get one for you
@sherbs ^
How could I verify the filament against the nozzle temp?...
I guess I could take the same roll of ASA home and set my ender to 200c and see if I can melt it. Probably could do the flow test to see a numerical comparison between the two. Stands to reason that if I cannot melt the ASA at home at 200c, my thermo on the RatRig has to be reporting wrong. Is that line of thinking correct?
why can't you use a thermometer?
For inside the chamber?
you are testing if the hotend is getting to the temp the thermistor is saying it is. nothing to do with chamber
Everything on the outside seems to be too shiny for IR/Laser
Hold on terminology
I meant the inside of the nozzle "melting chamber" my bad
is the inside of the nozzle and the outside of the nozzle really that different in temp?
i have no clue ¯\_(ツ)_/¯
I don't know how efficient the insulation is on the heater cartridge itself
Wait a minute, can't I meter resistance off it? Like against know. Room temp. Lemme look that up.
I encountered exactly the same thing two days ago for the first time ever after switching to Orca Slicer. Never had this issue with PrusaSlicer.
In my case nothing to do with Z height or thermal changes of bed or gantry.
1. Are you running OrcaSlicer?
ratty-blush•16mo ago
I had always assumed that the reason for doing heat towers was to tune THAT specific printer in for THAT specific filament, because the difference in heater, hot end block material, nozzle material, where the heater is in relation to the nozzle and where it is in relation to the thermister… means it is all relative for your setup and that the temp is NOT the temp in the middle of the nozzle’s melt area, just the temp where the thermister is and up to you to figure right temp (likely in the filament manufacturer’s list range).
I'm not, I've seen similar issues with SuperSlicer (sorta close) but also within Cura, which is what I'm testing all these out on
I mean yeah, makes sense, but that's a long way off. 10-15c difference? Ok yeah, but possibly 40? Idk man
ratty-blush•16mo ago
Yeah, 40c is way off. Dunno.
https://discord.com/channels/582187371529764864/791048619997659176/1140657501554102352
Ok back to this discussion
The datasheet for the thermistor says 25c = 100kohms, plus/- 3%
When I have RatOS heat up to "25c" according to the user interface, the resistance of the thermistor was 73.3kohms
At a resting room temp (days of no temp fluctuation) the resting temp according to Laser Thermometer on the outside of the heating chamber was 23.3, and at that temp the thermistor was reading 111kohms. Im assuming I need a new heater cartridge/thermistor, is that the general assumption based on those readings?
deep-jade•16mo ago
Just read through this whole thread. Fascinating problem. I think you’re on the right track with the temps (only other thing I can think of that would be so consistent is slicer extrusion widths, i.e. external perimeter flow, infill flow, etc).
With how far off spec those readings are, I’d say the thermistor is bad. The heater might be fine. Do you have another thermistor or hotend around to swap in for testing? Depending on shipping times for new thermistors it might be worth bringing the ender hotend in and plugging the thermistor into your RatOS board (don’t swap it, only plug in the thermistor to read the ambient temp). If the readings are the same in Klipper then something else is going on. If shipping is fast, screw it, order a pack of thermistors. They’re cheap.
Thanks for the input, it really helps.
I suspected the slicer, but under defaults (Unless someone has ones that want me to try) from SuperSlicer it did it, so I moved over to Cura and it also was repeatable. I may try to grab another heater but apparently the replacements are sold out? Idk I'll keep looking
Wait wait
Ok so 1 I think I have a thermistor issues ok cool.
BUT
This is one I was trying to print off for work today and at least limp it
What I didn't notice before
Is this time I ran my finger over the print.
It's like ringing, except z
You can see the bands in the print
Like if this was an ender styled printer with the wheels, this would make sense.
What doesn't make sense is what it picks and chooses when it happens, but when it does happen, it happens consistently
So I've tried two different slicers, with different models. The only thing left common is in RatOS, yes?
Because z is the 3 steppers, it has to be moving the steppers up and down?
How could I even test that when the movement is that small
The UI skips around too much to verify
Maybe a dial caliper as it prints would be easiest
I've also moved the models around on my build plate, to see if it was a hardware issue with the linear rails
But if that was the case, I'd have the ridges throughout the whole model, not just infill/some.walls
It's kind of hard to get a picture of it, but you can kind of see it here
The straight lines from the pov are slightly raised
Well actually now that I think about it I guess my issue is a difference in layer width? I am unsure of whether the movement is too low, normal, too low... Or normal, to high, normal
Or is that randomly the extruder sides to extrude more than it's supposed to resulting in it smooshing out the sides
Well I guess that wouldn't make sense, Guess it has to be bed movement then
Tomorrow I'm going to test changing the bottom layer to see if that makes any difference to concentric, If that fails I think I may try reflashing a new SD card maybe there's some data bottleneck happening
Does it correlate with your bed mesh?
111kOhm at 23.3°C is within tolerance.
How have you measured this resistance at 25°C?
You cannot simply connect the multimeter in parallel, while the NTC is connected to your printer.
Multimeter measures resistance by injecting a small current and measuring the voltage drop.
Your printer mainboard also injects a current, meaning they will add up and produce an invalid resistance reading.
What you can do is measuring the voltage. It should be:
V(T) = 3.3V * R_NTC(T) / (R_NTC(T) + R_Pullup) = 3.3V * R_NTC(T) / (R_NTC(T) + 4.7kOhm)
Then just insert the R_NTC depending on the temperature. You can calculate values e.g. on this website, by using the correct parameters of your NTC.
https://www.lasercalculator.com/ntc-thermistor-calculator/NTC thermistor calculator
Calculate NTC thermistor resistance at a target temperature.
Thermistor was unhooked at the time of measuring resistance, so isolated away from the board. This is good info and I'm gonna save that site for sure.
Hmm that's easy enough to test. I can just # that portion of the config yes?
Also using this same calculator, using my other measurement at 25.1c, it's now out of tolerance. Min resistance should be 96.5kohms on the calculator, while realworld meters 73.3kohms
Either this calculator is incorrect or this data sheet is. Just inputting temps to compare them to the sheet, it's off. For example doing 200c for target temp gives me a min resistance of .487kohms, where the data sheet expected resistance is .4452kohms, which is already below the minimum.
You're right: B is fitted from two points only: resistance at 25°C and 85°C. The entire R_NTC(T) curve apparently can't be perfectly represented with B only.
I guess you would need to build some interpolator using the whole table, if you want to calculate temperature from the NTC resistance.
If it's only for finding out the resistance for playing around, you can feed this tool with some other points and get the B/beta:
https://www.ametherm.com/thermistor/ntc-thermistor-beta
Ok, thats weird then. I'm still confused by the fact that it reads the correct value at 22.3°C but not at (supposedly?) 25°C? 
Can you verify that it actually has 25°C?
Or try to expose the NTC to some different temperature (refrigerator, basement, idk) and measure the resistance + temperature using the laser thermometer.
If the NTC itself is ok, there could be an issue with the NTC readout circuitry on the mainboard and/or configuration.
Can you verify that it actually has 25°C?
Or try to expose the NTC to some different temperature (refrigerator, basement, idk) and measure the resistance + temperature using the laser thermometer.
If the NTC itself is ok, there could be an issue with the NTC readout circuitry on the mainboard and/or configuration.All the real world temps I confirmed by infrared and by laser, at least on the outside of the heater cartridge (not the silicon sock, the actual wall) Each temp was measured 3 times to make sure there wasn't an anomaly
Today I will attempt a print with no bed mesh to rule that out, but just to add further confusion. I had to run a print overnight for a vehicle in need this morning.
Notice the very outer lines lay down fine, then we have issues for the next 2-3 lines, and the returns to normal.
Even heading towards the center is interesting to note that when the issue happens, they are all pretty uniform in spacing/size, but then the spacing/sizes changes when the pattern changes to a different set of spacing/sizes
Is the Bed_Mesh_Clear supposed to 0 out (delete?) The auto generated bed mesh values at the bottom of printer.cfg? I tried the command, but see no difference in the printer.cfg
no idea, I don't have bed mesh saved
Ah, I'm assuming if I clear out the whole section, I can delete that auto generated portion then?
https://github.com/jscancella/myKlipperFiles/blob/main/vcore3-printer.cfg
as you can see, I don't keep any auto generated stuff in my config (I prefer to put it where it is needed in the config)
I am unsure how to tell it not to mesh at the beginning of a print since macros.cfg is a read only. I am unable to comment out the Mesh section of the Start_Print macros
Thank you. I changed that to False a few minutes ago, but understood that as loading the stored mesh (which now I see is a different thing entirely)
Going to start a test
Just tried changing the extrusion on a whim, but it still has the same issue.
Are you changing extrusion widths dynamically or something?
Also worth noting I did move these models to different spots on the printbed
I am not actively doing such a thing, no
Unless that's a config hidden somewhere?
in your slicer
One moment.
Is that the actual name of what I'm looking for dynamic width?
what slicer are you using?
Atm Cura
But I have replicated this on both Cura and SuperSlicer
ok, I only have a really old version of cura installed
No worries. I just went through the hidden settings, and I have nothing via Width, or Dynamic/Dynamically enabled
sorry if you answered this before but I didn't see in your early posts - what extruder are you using?
(I'ma mess this spelling up) Phaestos Rapido UHF
no, that's your hotend, what is the extruder
<:homer_doh:584710153773776896>
I read that three times too
Bondtech lgx lite
and do you have a reverse bowden tube setup on it?
To be clear, thats Spool into Bowden to extruder, correct? If so yes.
can you take a picture of it please? I wonder if you are using tubing that is too tight and it is causing friction issues
1 Wouldn't that affect the print as a whole, even as the print head is moving around at the max angle on those circles?
2 no I don't.mind at all, gimme a few
3 I'd be more than happy to bypass the tube as a whole just to rule it out.
I forgot to post this other test for the sake of keeping track of what's been done.
This was still with bed mesh enabled, and bottom layer still set to default Lines. Notice the outer lines are clean, and then one section towards the right hand side, diagonal
I apologize I cannot move the printer at the moment to get some better pictures.
First pic is looking down
yeah, I would remove the filament from that tube and feed directly down into extruder as a test
Big swoop up into the loom. The second look to the hotend has piano wire in it to support it
like pop a hole in the top of that cardboard and then just feed filament directly down into extruder
Yup will do. Had to step out of the office for right now. Will report back
No Bowden tube, still persists
(forgot to turn the extrusion back up tho)
I would try prusaslicer and make sure arachine is turned off. That infill just looks weird like it is trying to dynamically do something
I have not used prusaslicer, but I'm assuming it's not prusa only? I'll download it here in a minute.
Alright setup and sliced. I left everything as default preset.
Issue persists
I'm really thinking it's something in RatOS
Or
Random thought, what's the chances of my SD card going bad? It gets a few lines in and starts to bottleneck data (idk how data transfer actually work, just trying to make an example) so the first bit is fine, but then it starts to bottle neck and the extruder skips a tick, resulting in under extrusion? If it misses every 3 ticks wouldn't that recreate this pattern?
Oh wait I missed the last bit of this.
Let me look to see if that was disabled by default.
It was not.
To be clear, Arachne perimeter generator is what you want disabled? @blacksmithforlife
Oop
One sec
Classic
Ok going to reslice
Maybe it's just placebo for wanting to find a fix for this, but I swear I can hear a difference in the tone of the stepper motors, right where discrepancy happens
Issue persists, with model moved along build plates, outer lines still clean, Arachne is disabled.
I'm trying to take these pics in a way you can see the lines easily in the glare
What's crazy to me is how similar these are, but they are NOT precisely the same
Thanks for this btw, so I wasn't stuck digging through the settings.
what infill type are you using?
Monotonic, default
What is your infill speeds?
and in your filament gcode, what is your PA?
SET_PRESSURE_ADVANCE ADVANCE=0.035for example
160mm/s for infill speed.
Let's see. Filament under Custom Gcode, says
SET_PRESSURE_ADVANCE ADVANCE=0.03
I'm stepping away from the office for a bit, but go ahead and ask questions/suggestions and I will try them asap
I let one finish just for the sake of seeing what else may be messed up.
@blacksmithforlife I'm pretty sure I already ruled this out since it happened across multiple models but for troubleshooting sake would you prefer if I drop the model It's self here to either be printed or to be looked at in a slicer
No, I don't think it is a model issue.
If you unplug your stepper motors from the main board (so it doesn't fry it), and pull on the X gantry, does the head move from side to side? I am beginning to wonder if the belts aren't tensioned correctly or there is a skew in your gantry. Since it takes both steppers to move in X and Y, perhaps it is trying to move directly across but is actually having issues going straight
I will certainly try that when I get back to the office.
Pulling on x results in straight movement. Thought we already ruled out movement/hardware as the issue persists at different spots on the printbed, and at no constant intervals. Like if a belt was missing a tooth, it would always mess up in the same spot
Free to correct me if that logic is wrong tho
When you were moving it left it looked like it also moved backwards
rotate your model 45° in the slicer (placement on the bed) and see the diff in the pattern,,,then 90° and repeat
This...is something I haven't done.
I'll make sure to give that a shot when I return to the office Monday.
It did, yes. But I pretty sure it's because I was trying to do it one handed. I will also try this again Monday.
deep-jade•16mo ago
Did you verify whether the thermistor is reading correctly? Since you’ve got a laser temp measure, try heating the nozzle to 100C and measure the nozzle tip with the laser measure. Repeat with higher temps, 220C and 240C. If it’s close then the heaters and sensors are likely fine.
Looking at the top and bottom of your test prints I think it’s very likely an over extrusion problem. The rippling/dashed line appearance of the bottom and the blobby/ridged top surface both indicate too much plastic is being pressed down. Have you calibrated e-steps and the extrusion multiplier? I see you’ve changed filament but be aware different filaments flow differently so you should recalibrate the “extrusion multiplier” when changing filaments. It would also be a good idea to do a temp tower. What type of filament?
deep-jade•16mo ago
I urge you to do some tuning and follow a guide like this (while written by a Voron user, much of it applies generically to Klipper printers): https://ellis3dp.com/Print-Tuning-Guide/articles/index_tuning.html I think there might be some common causes for your issue that will be resolved by following a tuning guide.
Agree with the above... Pictures show obvious over extrusion... I read through the thread and didn't see any mention of having tried tuning probe offset? , extruder steps? (extruder config in general - are you referencing the right extruder in your printer.cfg?), absolute v relative extrusion mismatch?
I'd suggest testing with a 0.3 layer height, will be much more forgiving than with a 0.2 height.
I'm not familiar with cura but I know SS allows you to get pretty specific with widths & flow rate... Might be worth switching back if cura doesn't give you that control.
The patterns you see are a result of trying to push way too much plastic into the volume of space that is available... Either way too close to the bed or there is just something very broken with extrusion related settings.
Wow,... That noise when you move the lot... that can't be good... sounds like something is rubbing or your bearings are to tight/loose?.
Also,... Somebody mentioned it here already, but my best guess here is your Z-Offset,... I always tune my Z-offset by letting SS generate a 200mm diam cylinder of 0.25mm height. Slice it with 1 or 2 perimiters. put the speed in mainsail at 25% so you have the time to adjust the z-height. Then move the z-height (in your case I would think UP) with very small increments. Particularly when the printer is starting to infill, you can clearly see the difference (taking into account the the printer buffers 1 or 2 moves, so change is vissible after 2 moves or so). Once the ridges disapear, slowly bring back the speed in 10% increments. This teaches you also if your first layerspeed is not to high.
I have no issues rechecking zoffset probes steps and such. However, if it was z, my issues would be persistent throughout the whole print. Same with estep/extrusion. I have not changed any sort of dynamic extrusion that would change mid infill or mid wall count.
https://discord.com/channels/582187371529764864/1135579117249638501/1142096570653868063
^ this included various higher temps, I just didn't list them in the post because it had already been verified.
Also I have *not changed filaments, just for the sake of not changing any variables. Although this will be happening soon as I'm about to run out of this roll.
I have changed slicers, which for note have different default extrusion multipliers for ASA, which I did not change. Even with different extrusion multipliers, my issue persists at almost the same line count.
This was also a known working printer before, but since this is at my office I'm only able to work on it at certain times.
Before this thread even started It had been so long since the last time I worked on it, I couldn't remember which variables I had changed and which I had not, so I elected to just flash my SD card and start from scratch. So everything RatOS is at its defaults except the variables to identify my printer and I have done a extrusion rotation distance calibration
This was at an 8 perimeter wall line count. At 30mm/s print speed. You can see the outer layers laying down perfectly, then at the 4th line count (outside in) it goes crazy. Same speed same settings.
https://discord.com/channels/582187371529764864/1135579117249638501/1135590124827189280
*when starting infill
This example has 2 walls and the rest is supposed to be infill. You can see where it starts messing up, fixes itself for a few line counts (outside to inside) and then goes back to the issues. That should have all been the same infill settings.
https://discord.com/channels/582187371529764864/1135579117249638501/1142152675803267224
List for troubleshooting
Move print head for isolated movement with steppers unhooked
Rotate model 45degrees and see if the pattern changes with it
Noises when moving the print head (bearings too tight)
Check against a .3 layer height
Oh dang, I can't pin that.
This I will double check when I get back to the office.
Can anyone take a sample I can compare against?
Alright I think that addressed everything so far.
deep-jade•16mo ago
Inconsistent extrusion (as seen in these prints, some lines look okay but they’re not consistent width or thickness) can be caused by bad estep/extrusion settings. In this case, it looks like over extrusion which can appear fine in spots but build up and worsen other areas. The infill might be especially bad if the slicer is trying to do it with fewer lines (wider extrusions) or faster. Can you upload the gcode file? It would be worth recalibrating the extruder and printing a test 30x30x0.25mm square. Top and bottom pics please.
I absolutely will Monday when I get back to the office. Anything to figure this stuff out lol. Boss been breathing down my neck trying to get this fixed
To be clear, you want my model, already sliced? Last time someone asked that they were actually wanting stl. Just wanted to be sure.
deep-jade•16mo ago
The sliced model please, the gcode itself not the stl. This can help troubleshoot as simulators can show a lot of info from the gcode files.
This is the last model sliced with PrusaSlicer per Blacksmithforlife's recommendations
This is the last model slice with Cura
@skrat19
Here in a minute I will check the movement before hooking the steppers up.
As for the movement. I CAN move it left to right forward back, without it going an off direction, but requires definitely more force.
If I'm only using a fingers worth of pressure, we move at a diagonal, like towards the end of the video.
Here's movement^
Video of movement ^
Who else
I'm trying this now. Thanks for the idea.
I forgot about the last half of this message. When the current test is finished, I will redo RotationDistance for the extruder and do your cube.
Mmm,... listened with headphones on... (also earlier clip)... doesn't sound to bad actually,... maybe a bit rateling, but nvery similar to mine...
Should I loosen?
I also have the adxl, But me personally cannot make sense of the graph.
I can send the last I did? I've made no adjustments since.
I rotated this one 90degrees
Hell maybe it is hardware
Those look to be in the same direction to me
I'm printing your square now. You are just wanting the bottom layer, no infill correct?
deep-jade•16mo ago
Use the same settings as the others but make the whole object 0.25mm tall and 0.25mm layer height.
Ah I didn't change the layer height tho. When it's done I'll rerun it.
Sorry, peeled it up while it was still hot so curved.
@skrat19 ^
.25 tall and .25 layer height.
deep-jade•16mo ago
This is definitely over extrusion. Z offset/bed mesh could cause this look but the top surface of the taller prints look similarly rough.
If you’ve calibrated the extruder then next is calibrating the extrusion multiplier.
deep-jade•16mo ago
Try this technique: https://ellis3dp.com/Print-Tuning-Guide/articles/extrusion_multiplier.html (It has the advantage of negating first layer/Z offset issues affecting the calibration.)
Well that's all for today. RatOS has decided to not let me connect to it via IP or host. Gonna shut it down and resume tomorrow. Maybe it'll be in a better mood then.
1. See if I need to loosen belts via graph
2. https://discord.com/channels/582187371529764864/1135579117249638501/1143262088555417683
Upload your printer.cfg please
And I'd suggest going to max layer height for your nozzle size... I think 0.4 nozzle? If so try 0.3 layer... The taller your first layer, the less effect you'll see from being too close to the bed
Even just the skirt in that last photo looks awful. Can you also switch to a single skirt loop?
deep-jade•16mo ago
The calibration method I linked will minimize affects of being too close or far from the bed on the first layer. I think no amount of Z error would cause the very rough top surface in https://discord.com/channels/582187371529764864/1135579117249638501/1142179634897768631 Thus my deduction that the extrusion multiplier is not calibrated correctly.
Unless EM is set to 200%, I'm not sure it's that either... Need to fix the first layer first, then we can see about the upper layers... I agree z-offset isn't the major issue here but it's a factor given that pretty much all of the most recent test prints are single layer prints
@sherbs this was a full print.
I'll check all that out when I get to the office today
Yep... And the whole thing looks massively overextruded... But, gotta start with layer 1
I'll have to put this on hold for the moment. I cannot connect to the printer anymore for some reason, gonna spend my lunch figuring that out I guess.
Turns out Sd card had ever so slightly moved. So mainsail would never load but pi would connect to the network.
Anyways
Since the general consensus is over extrusion. I was just curious what it would look like, So I reprinted the square and set the Extrusion factor in RatOS to 70%. The part I find interesting is that you can see "spots" in the lines.
Going to back up now and calibrate to the guide @skrat19 linked. Going in order now. @sherbs one moment and I'll get that printer.cfg loaded
(oops forgot pic)
@sherbs Anything I have changed recently while troubleshooting this I have datestamped in comments.
just fyi
Finished this one.
Working on this one now, but the guide wnts me to do pressure advance first
(Dang I forgot to tell it not to mention you, mb)
Working on extrusion multipliers now. Although now for some reason PruseSlicer is telling it to prime twice.
you probably have the start gcode in the printer gcode and the filament gcode. Just put it in the machine gcode
This is exactly it. I first copied it over to the wrong section, then put it into printer. I assumed it would just be moot since it was the same settings.
no, those are two different settings
Ty, now I know.
Fwiw, can I just leave it blank?
Or does it need the default preset code
leave what blank?
The filament specific Gcode, since I overwrite the defaults putting the printer start in on the wrong section
given the name, it is for filament specific gcode stuff. As you can see in mine, I have the z offset and the pressure advance as that is different for each filament. The machine gcode is for your machine that applies to every print (which is why you have the start print stuff there)
Trying to calibrate extrusion multiplier now.
1. If I have done the zoffset wizard, using the paper like usual, and my print is not sticking to the plate on initial layer, should I assume that the extrusion is automatically too low and is not putting enough filament to actually stick to the bed?
2. If I turn on pressure advance (which I thought was already enabled but was not) I get something along the lines of
max_extruder_cross_width has been exceeded error. Any ideas? Upon initial Google, didn't see it on the Klipper docs and its not in my printer.cfgFwiw, my *inital problems seem to be almost gone.
Please do a proper z offset calibration, the paper method is only to find the starting point.
1. no - the paper method just gets you close. You need to do a test print to get it perfect where you are adjusting it on the fly
Maybe I worded my question wrong.
If you had your initial layer dialed in for zoffset and extrusion multiplier of 1.0, but then changed your extrusion multiplier to for example .7, would that be enough of a change to not stick the initial layer?
Could be possible. But for EM tuning you need the filament to properly stick, first
I would recommend to revert modifications to the extrusion modifier, widths and so on to safe defaults (e.g. Ratrig Profiles in PrusaSlicer).
Then do a z offset calibration by printing e.g. a 100x100x0.3 cube and live adjusting the offset
With this, proceed with the EM tuning
I'm really only trying to change one variable at a time as I seem to have an issue that is difficult to pin down. Which is why I was asking if EM made a difference for initially layer, because of it shouldn't, then I need to up my variable of EM since I had a known good z already. Does that make sense?
Tbf I'll do anything to get this fixed, but I've also been troubleshooting this for a while in this thread. I would hate to reset all of my progress by missing a variable.
Here is my .25 cubes. https://discord.com/channels/582187371529764864/1135579117249638501/1143252191033954354
Scroll down a few messages to see the results.
Well, EM affects all layers. So obviously also the first layer.
But you cannot tune EM based on the first layer. Also „good Z“ can change. So please do a proper Z calibration, not eyeballing it with paper.
So is the
probe_calibrate tool "eyeballing" it?Absolutely
Well if that's worthless, can you please point me to some documentation I can follow.
Or actually it’s not eyeballing because the tool has no eyes.
The procedure is simple: print some object (cube with 0.3mm is enough, but it can be higher and just cancel it after first layer). While it’s printing the first layer, use the buttons in the web interface, to adjust the z offset
You want the lines to overlap a bit, but not too much. I think here should be images of what you want, or maybe on the Prusa blog
https://ellis3dp.com/Print-Tuning-Guide/articles/first_layer_squish.html
I understand you cannot do it on first layer. I'm using ellis3dp extrusion cubes.
But you can't get to a higher layer of your bottom fails. Hence my original question.. if you have a known good z and it's not sticking, Do I automatically adjust EM back up and test again, or lower z to make sense for the lower EM.
If you make substantial changes to EM, you may have to retune Z during the process
First Layer Calibration (i3) | Prusa Knowledge Base
The First Layer Calibration is used to calibrate the distance…
Here is a similar guide. The Prusa printers have some built in program which prints a meandering line, but imho it’s better to just printer a large rectangle.
I like to use a phone flashlight to better see if there are gaps while printing.
This doesn't make sense. Why would you change multiple variables when trying to troubleshoot/isolate something.
Alright I'll get it started.
If the variables are „connected“ to each other, it’s impossible to tune one without the other. There is no way around this
But the Ellis cubes try to decouple the top part from the bottom using infill , so imperfect z offset shouldn’t be an issue, as long as the part still sticks.
In this case, just do it once afterwards to „perfect“ the offset
After doing all the cubes, we are still here
Still looks like too close to the bed. Have you tuned the z offset?
...yes.
If we need to, I will drag the nozzle on the plate and then raise .2 if I need to
Or .1 or whatever the number is so that we can all be on the same page
I have done this. Zprobe offset is now set to whichever has the most thickness. The thickness of ink, or .005
Whichever one of those is higher
TIL Sharpie ink has the thickness of 30 microns, as compared to the smallest amount RatRig can move at .005, or 127 microns.
(I was like man he's really yelling at me lol)
your probe should be like a 1mm above your nozzle, not 0.005
No
Hold on
I was told specifically not to do the paper test earlier
this has nothing to do with paper test
your probe should be physically higher than your nozzle
that is what z-offset is
..yes I understand that.
then something is wrong if you offset is set to 0.005
That's my tolerance, not the actual position of the nozzle
Let me finish
Lol
Bottom line is its calibrated as much as this machine will let.me
I literally can't get any tighter tolerance
tolerance and z offset are not the same thing. What is your z-offset set to?
for instance mine is currently set at https://github.com/jscancella/myKlipperFiles/blob/main/vcore3-printer.cfg#L203
Im starting to wish I never got this printer 😔
Yes I understand tolerance and offset are not the same. I keep being asked to calibrate z offset. What I am saying is my z offset is calibrated at 1.170, and THAT calibration is within .005
THAT calibration is within .005that doesn't matter
Just trying to make sure we are all on the same page
we are saying - 1.170 is probably wrong
That's not what I typed.
Yes I understand tolerance and offset are not the same. I keep being asked to calibrate z offset. What I am saying is my z offset is calibrated at 1.170, and THAT calibration is within .005the 1.170 is too close to the bed
This is a relative measurement, what do you mean?! Regardless of actual amount of offset, if the probe triggers at the time the nozzle is at the bed, it's the same. Could be offset of 100mm as long as it's all relative
Or am I misunderstanding what z offset is.
I'm happy to be corrected. Just want this thing to work
what? Your probe triggers before the nozzle hits the bed. The difference between when your probe senses the bed and your nozzle actually touching is the z-offset
so if you change that value, you will in effect change where your nozzle is actually starting in relation to the bed. Decrease the value and the nozzle will be further away. Increase the value, and your nozzle will be closer to the bed
Ok. Here's the thing. I'm gonna nuke the z offset on settings.
And start over, again
why are you nuking them?
you just need to adjust them
try setting it to 1.1 instead of 1.17
My probe triggers at 1.025 mm away from the bed.
Z offset is about the distance between the probe triggering and the nozzle height from the bed. Doesn't matter how far the probe is away from the bed
Ok lemme do it this way.
Do you want me to find z offset manually or with RatOS wizard
neither, you want you to go into your printer.cfg and set the z offset to 1.1 instead of 1.17
then restart klipper and go and print the exact same gcode you did before
Ok I can do that.
Zoffset set to 1.1 in printer.cfg and there's no change in the UI for the zoffset section
Please send your config
and there's no change in the UI for the zoffset sectionwhat? it should just display z=0
That's what I'm saying
I am so confused by what you are saying
Did you do the test print?
cutting thie video now so i can pos it
You want it again? Give me a moment.
sorry, I didn't remember that you already sent it
It's fine lemme finish this bit
I'll send a fresh one
This is the old config, it looks okay so far (no broken SAVE_CONFIG sections or something)
Just to make sure: you are changing the z_offset down here?
Correct
the download renamed it, its named correctly in the printer
no, reprint this and show us the bottom. That video just shows the beginning of the print
i cant
you don't see that print failing?
This was after the z offset change.
ok
What is this print surface you are using?
Smooth side of PEI sheet
The stock ratrig sheet?
ratrig doesn't come with a smooth side
Isn't the backside just bare steel
Have you used the other side yet?
I think it's fine if you just swap it, wipe it a bit with isopropanol, then start printing some flat box...
Yes I have, gave up and tried the other side.
Already flipped it and starting the same print
Yes I cleaned it with 99%
still need to fine tune on the other side, and the new offset could be too low. So don't give up quicklyI've been doing this for months, no point giving up now
Seems to be sticking. I'm gonna let it run for a bit. After this I'll try something with a flatter surface to see the first layer better.
Ok, but please choose a flatter surface that is actually sticky 
I don't know what this means. You meaning my object or printbed
I meant the printbed
other objects are fine, I like to print a huge rectangle
quickest-silver•15mo ago
This side of the bed is not meant to be used as a print surface, at all, please don't do that.
Use https://docs.ratrig.com/commissioning-guides/v-core-3-1#z-offset to calibrate Z offset