# PETG nozzle clearance and extrusion multiplier

I've started printing PETG recently and I'm happy with results so far, awesome strength and good looking (except for stringing). But I've noticed that PETG prints better with more distance nozzle-plate than usual, and under-extrusion make parts looking better than both normal/over-extrusion.

• What distance nozzle-plate is optimal for PETG? (i.e. the distance between nozzle and build plate during calibration)
• What extrusion percentage is optimal for PETG?
• Your second question seems a bit strange, any value lower than 100 % will cause under-extrusion, so what is optimal? Any value different than 100 % means that you have not calibrated the nozzle well. – 0scar Jan 25 '19 at 14:19
• Is it solely the clearance on the first layer that's a problem here? If so, adjust the Z-axis stop. If the nozzle appears to be running into every layer, then as @Oscar suggests, your "underextrusion" value is really a proper calibration for the nozzle in use. BTW: what is "dirty" ? If you mean burnt bits, just drop the temperature. – Carl Witthoft Jan 25 '19 at 15:25
• @CarlWitthoft Your comment about the temperature makes perfect sense in light of of the OP's mentioning of stringing also! – 0scar Jan 25 '19 at 15:45
• Clearance is machine, not material dependant. You also speak about initial layer thickness, not clearance, which is space between parts – Trish Jan 25 '19 at 18:50
• @FredoCorleone I have assumed that with "clearance" you meant the paper distance when bed levelling, also known as Z offset, maybe it is better you make that more clear. Both questions are now unclear. Please update by edit, thanks! – 0scar Jan 26 '19 at 8:30

Here is the mental framework that I use to reason about PETG: In a nutshell you want to avoid nozzle contact.

Unlike most other plastics, PETG sticks to hot brass really well and every time the nozzle moves through material it will pick up some of it. Material around the the nozzle then sticks to a random place creating a blob. It can also cook, turn transparent brown and drop into the print. Investing in a plated nozzle or silicone socks helps but doesn't eliminate the problem completely.

Now to the questions.

### 1) Nozzle Distance

Distance to the plate has to be such that the plastic is laid down precisely without the nozzle dragging through the material (remember, avoid nozzle contact). Precise lines require the build plate to be level and the flow perfectly calibrated. If nozzle is too low and/or the layer is over-extruded then PETG will stick to the nozzle and rip the lines off the plate again. Inspecting the first layer is required for best results. I like to print a layer test pattern after the flow has been calibrated and tweak Z offset in 0.02mm increments until it's perfect.

With many other plastics it's ok to have a large amount of "squish" in the first layer as it helps to work around minor leveling issues. This is where the cookie-cutter recommendation to raise the nozzle when printing with PETG is coming from.

### 2) Extrusion percentage

Flow has to be near perfect. Down to one percent perfect. Even a slightest over extrusion and some of the excess material will end up on the nozzle when it makes the next pass. Under extrusion isn't great either as this can lead to holes and affect overhangs where thinner strands of a previous pass may not be enough for the next line to stick to.

There are two critical parameters: diameter of the filament and extrusion multiplier. This is how to determine the settings:

1. Measure filament diameter. I use an average of ten measurements over about a meter (yard) of filament taken in multiple orientations.
2. Calibrate the extrusion multiplier using a method described in Prusa manual: I print a 40x40x40 cube in vase mode with extrusion multiplier set to 1 and fixed extrusion width (e.g. 0.45mm), measure the wall thickness in three spots on every side, average the results and compute the correction factor.

I perform flow calibration for every new roll of filament.

• To calibrate flow percentage I make a mark on the filament itself, about 10cm then I extrude 10cm within the move axis panel and then divide 10cm for whatever it has extruded. Is it good? – Fredo Corleone Jan 30 '19 at 14:53
• @Fredo Corleone - this is a different process called E-steps calibration. It is a prerequisite for a good setup. However it's done once when you assemble the printer. The process described here is fine tuning for every individual roll of filament. – anttix Feb 4 '19 at 23:54
• @fredo-corleone The reason additional calibration is needed is that e-steps calibration handles the length of filament going into the machine. However the volume that is extruded also depends on the shape (how oval it is) and diameter of the filament (measured first to get close). – anttix Feb 5 '19 at 0:02
• Also where did you enter that number? This should go to printer firmware and not into slicer. – anttix Feb 5 '19 at 0:10
• I've just finished calibrating the extruder stepper, now it extrudes exactly the right amount! And I'm currently calibrating the flow. It seems that my printer over-extrude a bit (printed wall is around 0.45mm while line width is 0.4mm). – Fredo Corleone Feb 8 '19 at 14:01

I have printed literally kilometers of 2.85 mm PETG filament on various 3D printers, and frankly, I do not share your opinion on an increased calibration distance/offset (like using thicker paper when levelling you build plate or increasing the Z offset by G-code M851). I even lower the default first layer height in Ultimaker Cura (0.2 mm prints fine). I am aware that on the web there are folks that do increase the calibration offset, or increase the first layer height, but that should not be necessary on a well tuned printer with sufficient first layer adhesion (e.g. printing on glass with 3DLAC).

Furthermore, the best extrusion multiplier for printing PETG is 100 % on a well tuned extruder for a constant diameter quality filament brand.

On my Ender 3 Pro's I have found the following works well (also remember settings can be effected by different brand/quality of filament):

• Bed to nozzle 0.2-0.3 mm,
• Multiplier 100 %,
• Nozzle ~230 °C/bed 70 °C,
• Speed 50 mm/s.
• Cooling off first few layers but from there cooling and retraction is part specific.

If you use retraction, it may help to slow it down to 25 mm/s and adjust retract distance if your using Bowden tube or direct drive. Last, a must have, a can of hair spray, works great and less expensive than the glues.