Something wrong with first layer!


#81

@goldensnake

That is exactly what I was getting with S3D. Adding .01 z-offset [gcode] and increasing first layer height to 110% [layer] seems to have solved it for now.

But yes, exactly the same thing. It calibrates as normal, but then prints as if it is not level at all, with the head too high in some places and too low in others.

I don’t think it is in the machine firmware update because it started before that, it is something in FLUX Studio.

I am curious if selecting one of the other slicers as the default, either Slic3r or Cura, before importing gcode will make any difference. I’m just guessing, but the problems seem to have started around the same time Cura2 was brought in to FS, so maybe selecting a different slicer will default the machine into the old calibration???


#83

I’m having the same issue, on some places the nozzle is too high and in other places is to close to the bead.

I’m using simplify3D and except for this issue my printer is printing very good pieces


#84

Actually if feels a bit random in the latest flux studio.
Most of the time calibration seems not bad (about 0,05 variation if I measure the brim / skirt) but I had a few prints where is was so bad that there was nearly nothing in some places (it actually looks like the nozzle was scraping on the glass).
Then again I only did smaller print so far so it migh be worse on a biger surface like 100 x100 mm.

@proclaim On that note does the flux use mesh bed leveing?


#85

I think that I have a cure for the first layer on my printer. I just raise a glass plate 0.15mm by using a small piece black tapes on two rubber pads where a first quadrant that always over extrude (see on pictures). In addition, z offset is 0.1mm. There is no more clunky noises from filament extruder while printing a first layer.


height=“500”>

Update:
The lesson that I learn from the thread is a FIRST LAYER that does affect a quality of a print. A boat printing surface is now perfectly smooth. Here is a comparison between old and current 3DBenchy test (a brown (old) vs a red (new)).



#86

Does everyone have to clean the nozzle with the pin after every print now?

Also I had issues with the first layer when I told the software I had upgraded the extruder. When rolled back to the original setting it was fine.


#87

It’s not in my case. I just need to wipe out all burned excess glue or filament which a nozzle scraps it on a glass bed while printing a first layer. Since I fix my first layer problem, it’s clean. I don’t know how much it’s worse of your first layer as mine but it’s always bad on all of the FS and firmware revision. I positively think that some thing to do with calibration routine. All I do is raise a glass bed to 0.15mm on where has over extrude layers (red rectangle) + 0.1mm z-offset; my first layer is like a thin sheet of paper of 0.15mm thick.


#88

Good idea Luan, I am going to try a piece of tape also, I’ll let you know if it helps. In my case, the red area you show in the picture is higher for some reason and the nozzle scratches the glass at that region. Rotating the glass platform does not help either and it is always there. I usually compensate for it with Z offset also but I will put some tape on the opposite side to see if it does reduce the scratching.


#89

Your problem is exactly as same as mine. Let me know if it works.


#90

Same problem here. Calibration succeed with no problems but while printing the first layer, the noozle in a region is so close to the plate that the filament doesn’t extrude. In other area is more far. (I’m on Fs 0.8.2, gcode generated with S3D)

Since this problem seems to affect more users, I wonder if Flux team has been warned. @proclaim any thoughts?


#91

Would you be able to try my method (see my previous post) of raising a bed,a glass bed recommended, about 0.15mm and test it? It a learning experience that might be working or not from others printer.


#92

This weekend I’ll try and I will let you know. Even if it works, I hope Flux team will fix this if it’s a recognized problem.

Thanks a lot for your help


#93

I know that but Flux team has been muted on the issue. I’ve to have my own solution.


#94

Unfortunately, this can happen with just about any printer, especially deltas that use FSRs under the bed. It’s a great way to go about it, but it can take a little bit of fine tuning. MHackney just posted a blog about probing height, if you’re curious.


#95

I’ll give that a try also… finally got around to ensuring my print head is not clogged (which I suspect there might have been a bit of as that darn cleaning pick did NOT want to go in until I used one of my other eBay driller type cleaner picks get it started) and that the fans appear to be working fine. Just applied the latest machine firmware update that had been sitting around waiting to be loaded, and set the printer off do do a 170mm cylindrical-ish (ish because it’s a cylinder with flats around the perimeter… silly tinkercad!) print, and I also have a inconsistent first layer, using the cura defaults. I’ll switch to S3D, as I was intending to get that tweaked and tuned, and I’ll work on getting that to work for me. I’ll upload a picture of my layer pancake for your viewing pleasure a bit later if you so wish to see it :wink:


#96

Uh oh. This is exactly what I was referring to a few days ago when I mentioned on FB that I stumbled into something with the first layer issues.

I was finding the same issues playing around with the magnetic plate and just by rotating it, and/or rotating the model or moving it to another spot on the bed I could get a perfect first layer and good print. So the issue is definitely in the calibration routine and it changed sometime after 0.6 versions of FLUX Studio. I agree with you Luan, this is something in the code for sure.

I’ve switched back to glass plate for last several prints and most have been smaller so I haven’t had the issue so much, but I think it is the same two FSRs that you guys have noted. I will print the 120mm pancake tomorrow and then try to add the adjustment tapes and see what happens.

How many pieces of black tape = 0.15mm? Or how thick is black electricians tape?


#97

I use electrical tape (~0.15mm). I tapped on top of FSRs, which are covered over extrude area. Thanks @Jimustanguitar information about probing z height and relations with my fixed that make more sense to me now. I just add/compensate a trigger point below a bed (which is a constant number in a firmware)+ a thickness of a piece electrical tape to correct a Z Height of my printer. I also have another 0.1mm z offset on gcode. However, I have to have both hardware and software z offset (0.25mm) that’s subtracted from the Flux z height. The first layer problem is closed on my case. Happy printing.


#98

@goldensnake, here are my test results, which confirm your findings.

First picture: 1 layer thick with Z-offset=0.1mm and Layer_0=0.25mm, 1st-layer-print-speed = 25mm/s, FS v0.82 and Cura2, no height adjustment tape. Layer was damaged while taking off glass platform, however next to the coin, by the torn hole, is thinner than the rest of the wafer. Still, the result is acceptable by my standard since I don’t print 1 layer structures.

Picture 2 below: Added 2 layers of Scotch tape on opposite side of the coin where the print was thinner. Print settings are the same but raising the opposite side made the coin side even thinner. Notice: I forgot to add more glue stick in between printing, and one side shows poor adhesion.

Picture 3 below: Similar to Pict #1, no tape adjustment, same Print settings, but changed to Cura engine (I changed to Cura as I am more comfortable with it), also left the print on the plate to avoid damage. Print shows thin quadrant right of the coin (black glass showing from underneath). As Pict #1, result is uneven but acceptable for 1 layer. The reason

Picture 4 below: Same Print settings, Cura engine + 2-layer Scotch tape adjustment. Most uniform print so far for a 1-layer wafer.

Conclusion: I am not an expert but I think most probably Flux Calibration does not create a height correction mesh (matrix) or maybe it needs more data points to compensate for platform slope from edge to edge.

I would also like to add that, even without the tape adjustments, I find my printer very reliable and perfect for my needs since I don’t print single layer structures (they are also a nightmare to lift from the printer). By the 2nd and 3rd layers almost all imperfection disappears in general. However, it would be interesting if @proclaim or someone from Flux could comment on the Calibration method and the leveling issue.


#99

Test #4 is your best result out of four trials. :ok_hand:


#100

Yes, it is and I am very satisfied with the uniformity of the layer, thanks for the earlier tip.


#101

This is interesting.

This shows really clearly that it is not a circle. Was it supposed to be?

I know a very slight difference in the build plate causes huge errors. I had not really though about it too hard. It took me a while to clue into that.

Is no one using the flux magnetic build plate?