Filament-specific Z offset adjustments?

I’ve seem it said that PLA should be squished a bit for the first layer for best adhesion, but don’t do that at all for PETG. I have certainly experienced having perfect PETG prints, then going to PLA and having it just not stick. This would imply that it would make sense to have a filament-specific Z offset in the slicer. I don’t see this in slic3r or prusa-slicer; as far as I can tell, I need to make a different “printer” if I want to do this. Does anyone know of a setting I’m missing? I’m not used to slic3r missing configuration settings; usually the complaint is that it has the kitchen sink full of settings! :wink:

I’ll probably do a new “printer” for PLA anyway, because I think I’m going to give up on PLA sticking to the PEI-coated spring steel bed I bought, or the PEI sheet on my second printer. I have plenty of happy experience printing PLA on glass; I can just set a sheet of glass on top and adjust Z offset for the thickness of the glass and go back to what I know for PLA. For PETG, on the other hand, printing on PEI is great.

I don’t use a Z probe; both of my printers are carefully “leveled” (trammed) with kinematic mounts, and they don’t go out of tram. That means I can’t just probe the current surface with (say) a bltouch. (I even own a bltouch; I just bought it right before learning about kinematic mounts, and I never bothered to install it.)

(I’ve made the mistake of trying to adjust my Z endstop for PLA after dialing in PETG, and losing track of my notes and taking longer than I’d like to get it dialed back in for PETG afterward.)

I run the same offset for PLA and PETG. That is for my printers that don’t have a BL touch now. That being said I do run different extrusion multipliers for PLA(0.96)versus PETG(0.93).The only thing I found is that with smooth PEI I sand the surface as soon as I get the sheet. I don’t care about a glass mirror finish bottom. I just want to have things not pop off. So I sand it with 220 grit and then wipe it down with isopropyl alcohol.

Is yours a textured bed or a smooth bed?

The corexy has a magnetically-fixed two-sided spring steel sheet; one side is lightly textured and the other is deeply textured. I have gotten PLA to print, though with some corner lifting, on the deeply-textured side. It still sometimes shifts. The Parametric phone stand is the only thing I’ve successfully printed in PLA on that surface.

The flinger has a thick smooth PEI sheet applied directly to the heated bed. I didn’t realize at first that I should sand it. PETG adhered easily to the entire smooth surface and was a horrible pain to remove. PLA didn’t stick at all. After sanding it with 220 grit and cleaning it with isopropyl alcohol, it’s easier to remove PETG parts, and I haven’t tried PLA yet since sanding it.

Sometimes a glass-smooth bottom finish is something I want, but what I care about more is that PLA on hairspray on glass sticks amazingly well, yet pops off easily when cooled. I have almost never had corner-lift problems with glass when printing PLA, even on my terrible first kit printer, and almost always have corner-lift problems with PLA on PEI. I have multiple glass bed sheets, and I can swap a new bed on and keep printing while the first one cools. Also, the freezer is just around the corner. I can set the glass sheet with the print on it in the freezer, set up the new print, and as soon as the next print has started, get the prior print back out of the freezer and the parts have released perfectly. I’ve missed that convenience since going to spring steel.

However, I might need a different clamping system. It was fine to use metal clips when I was holding the glass directly to metal; that wouldn’t bend the glass. But clipping it to magnetic sheet (corexy) or thick PEI sheet (flinger) I’d think it would potentially bend the glass a bit. I’ll have to set up a dial indicator to test that theory.

I print PLA on PEI. I just give it a quick coat of hairspray before printing and it adheres very well.

You know, I’ve done that so long on glass, why didn’t I try on PEI? :roll_eyes:

Though I can’t find the aqua net extra hold I used to use as a bed fixative any more. Need to find a new unscented hairspray now.

…It looks like it’s available again, at least on Amazon. I whiffed so many times trying to buy it locally. Multiple drug stores, Target, etc.

It makes sense to spray on the removable PEI bed on my corexy. Not so sure about spraying it directly on my flinger where the surface isn’t removable, though. I don’t like the idea of getting it into motors, on belts, into fan bearings, etc.

1 Like

I think the fact that you’re asking this question means that you’re thinking about Z offset wrong. Technically, your Z position should be calibrated to make your first layer exactly the height specified in your slicer. What you’re really asking for is a first layer flow multiplier, so that you can slightly over/under-pack the filament on the first layer to adjust the level of adhesion (unless you’re suggesting that the difference in thermal expansion of the nozzle and platform is different enough at different materials’ printing temperatures that the nozzle gap is wrong with one material when calibrated at the other temperature).

Heh. Sorry about the wording. I’m not thinking about it wrong. just worded it poorly.

I’m adjusting to a reality in which as far as I can tell, I don’t have a first layer extrusion multiplier configuration setting. I looked. I would love to find out that I’m wrong.

I calibrated for precise layer thickness. I admit I used calipers not a micrometer because calipers were more precise than the variations from surface unsmoothness, but I’m printing on top of an Alca-5 cast bed with a kinematic mount in order to to more consistent in layer thickness.

With a glass bed and hairspray, the same correct thickness worked for PLA and PETG, consistently. I very rarely lost parts to adhesion failures even on the loose junk gantry printer that I eventually disassembled, let alone the corexy where the calibration was much more accurate and the movement more precise.

I’ve seen a lot of advice, maybe bad, that due to bed adhesion characteristics being different between PLA and PETG, the most expedient workaround is tweaking Z height. I don’t like the lack of precision. 3D printing was my gateway drug to machining, where “holding tenths” means being within a ten-thousandth of an inch of a dimension, and I became one of those people with a 4-jaw chuck and carbide boring bars for my lathe to be more precise. I sometimes finish 3D printed parts on the mill or lathe.

I’ve ordered more hairspray (which I kept striking out on finding locally) to try with the PEI surfaces and PLA to see whether I can leave my Z settings correct.

Oh, and if my calculations were correct, the difference in thermal expansion is not meaningful relative to other sources of variation. (I’ve forgotten precisely what small numbers I came up with, but it’s not like the expansion coefficients for 6061 and 5083 aluminum were hard to find.)

Yes, the change in thermal expansion itself shouldn’t make a measurable difference. Depending on the thermal gradients, both between the top and bottom surface of the platform and center-to-edge, the differential thermal expansion can cause bowing of the platform, which I HAVE seen cause noticeable effects, especially when the heater is powerful enough (and the thermal conductivity low enough) that this bowing changes as the heater kicks on and off.

It occurs to me that this is a reason to use the thinnest cast plate that makes sense in an application. I had wondered whether going with 6.35mm plate was thinner than optimal for a 330x330mm bed, but I’ll expect that the thicker the bed, the more pronounced this bowing could be.

I do have a 750W heater, but I’m using PWM, not bang-bang control of the SSR, and just have an aggressive heat sink on the SSR to reduce the impact on lifetime. (Also TCO inline bonded to heater for safety.) So I’ve seen no evidence of thermal banding as sometimes shows from bang-bang bed heating.

On my assembled-from-spare-parts printer, the 500W heater is 200x200mm on a 240x270mm bed (couldn’t easily get a heater than fit) and I have seen that bowing effect slightly, temporarily until the plate temperature stabilizes. But once it stabilizes, the max temperature difference from the hottest point near the center to the edge is 2⁰C or less, and I don’t think this effect is substantial. I just let it soak for about a minute at temperature before printing.

I’m hoping that hairspray on PEI solves my PLA problem, but do you know whether any of the slicers have a first layer flow multiplier?

I’m not sure if any slicers offer that, but if they let you specify layer-change gcodes (and don’t run them before the first layer, you could put a code there that sets the flow multiplier (via M221) to 100%, then set a different flow multiplier in your start code, and that value would only be active for the first layer.

1 Like

BlockquoteI don’t see this in slic3r or prusa-slicer; as far as I can tell,


On my Prusa Mk-3S printer running 3.9.0, I have six “print sheets” to customize however I want, including the Z-Axis offset. You can select the sheet just before starting the print from the toggle wheel. The sheets can be renamed. I think it would be to better to be able to do this selection in PrusaSlicer, but I don’t see a G-Code command for it.

Cool, that led me to this:

I’m personally not running a Prusa printer because I like to DIY but that’s cool to know about, thank you.

Is their a filament specific initial flow rate?
I have not been printing for a while this year.

That’s what’s @Whosa_whatsis’s approach provides: more/less flow in the first layer by taking advantage of layer change custom gcode. :brain: