Switching froma 0.3mm nozzle to 0.2mm, Little Help?

I ordered a 0.2mm nozzle to test a few things out.
I switched over the nozzle (down from 0.3) and obviously the extruder can't push as much filament through the nozzle as it could.

My problem is that now my extruder is still trying to push as much filament through at the same speed... changing the steps per mm would change the amount of filament that is trying to go through, but the problem lies with the fact its pushing too fast.

How do I go about calibrating my new nozzle?
I'm assuming its to do with print speed and not steps per mm?

Think about what you are doing !

The steps per mm relate to the amount of un-melted filament fed in to the extruder per x steps of the filament stepper motor. Once the stepper motor is calibrated then the nozzle size is a great extent irrelevant. The amount of filament that comes out of the nozzle is determined by the nozzle size and the amount fed in. The amount to feed in is controlled by the slicer on the basis of the nozzle diameter set within it and this amount is embedded in the produced gcode file. I presume that you sliced the model and generated the gcode with 0.2 mm set as the nozzle diameter in the slicer. You CANNOT print a gcode file generated for one nozzle diameter with a nozzle of a different diameter.

OK, so i've left the extruder steps the way they were... changed my settings in the slicer so that it is now using a 0.2mm nozzle and set my layer height to 0.16mm (80% of the nozzle thickness)
Now all i'm getting is my extruder skipping steps.
It pushes the filament out during the printers start code:
M201 X1000 Y1000 E600
M92 E161.3
G28
G92 E0.0
G1 X8.0 F2000
G1 Y60.0 E4.0 F1000.0
G1 Y100.0 E8.5 F1000.0
G92 Z0.1
G92 E0.0

but after that it is just skipping steps.

My printer doesn't need and doesn't have an M201 in the start code. I'm also unsure why you are changing the steps per unit of the extruder with the M92 command when this is something that should be sorted and corrected in firmware as the nozzle size is irrelevant to this.

The amount of 1.75 mm diameter filament extruded OUT through a 0.2 mm diameter nozzle is roughly 76 mm for every 1 mm pushed IN. The second G1 command tells the printer to extrude 300 mm of melted filament and the third one tells it to extrude 646 mm. I have code in my files to retract the filament 5 mm at the end of a print job, and in the start code to put that back before commencing the print. All of the other extruder instructions in my gcode files deal with fractions of a mm. If all of the rest of the extruder instructions in your gcode are similar then I'm not surprised that it's missing steps because you are simply asking it to push too much filament through the nozzle.

The area of a 0.2 mm nozzle is less than half that for a 0.3 mm and a quarter of that for a 0.4 mm. As you go down in nozzle size it requires more effort to push the filament through the nozzle due to fluid friction and it is also much more dependent on having the correct fluidity of the melted plastic (it becomes more important to have a suitable temperature). The smaller the hole you are trying to force the plastic through the more fluid the plastic must be.

In addition to the above the, thinner the layer that you are printing the more effort that is required to force the filament through and the more that any error in thickness will effect the extruder pressure. If you are printing with an increased layer width on the first layer this will also put further pressure on the extruder.

Why don't you get the printer working properly and become familiar with its workings and capabilities (including the effects of changing various slicer parameters) with a larger nozzle before trying the smallest size ? I can manage to print 38DP spur gears ( 1.5 mm tooth depth) to an accuracy pretty close to that of machined gears, and I use a 0.4 mm nozzle.

Edited 1 time(s). Last edit at 11/14/2016 12:39PM by Supermec.

Basically I was provided this start code by a user on another forum. The reason I am trying a 0.2mm nozzle is because I was also suggested by someone else.
Another post in this forum describes my problems with my printer extruding WAY too much filament making my 1mm walls 1.8mm (which i believe you have found), so I have tried everything to try and fix it, a 0.2mm nozzle was the next thing to try.

I would love to learn more about why things are going wrong, hence why I am asking questions. I have never had any extruder problems printing the parts I was, it has only been recently that I have tried printing more details that I have struggled (especially parts that fit into eachother)
Hence the reason for trying these new things.

Some of the problems arise from how the slicer deals with objects. If you print a cube or a cylinder with the slicer set to a solid bottom and no top with the shell thickness set to 3 x nozzle diameter with 0% infill then, after trimming the very slight spread that you get on the first layer, the cube or cylinder should measure up to close on the required dimensions. The thickness of the shells (the walls in this case) should be pretty close as well to the 3 x nozzle diameter.

You notice that I said pretty close - it won't be exact because the slicer assumes that a line of filament goes down with a rectangular cross section, whereas the actual shape is governed by the nozzle, the layer below and the fluidity of the plastic, the inner and outer edges of which do not form vertical faces. If this was not the case you would not be able to see the lines in the faces of a print. For a 0.4 mm nozzle and 0.2 mm layer thickness the top and bottom of the new layer will be slightly below 0.4 mm wide and its width at mid height will be about 0.48 mm. You are effectively squashing a round jet of filament into a wide oval with flat top and bottom. You can reduce the width by reducing the amount of filament extruded, but the down side is that the new layer is pressed less firmly against the previous one which can cause layer separation in thin shells.

If you now design a model that is a hollow box or a hollow cylinder with a designed wall thickness of say 6 x nozzle diameter then this should still be OK with the shell thickness set to 3 x nozzle diameter. This is because the slicer treats the object as having two shells with the shells being the inner and outer three rings.

The problem comes when you go to a designed wall thickness below two shell thicknesses as the slicer is not entirely sure how to treat the object - is it one or two separate shells? Since there may be other parts of the object that are thicker you obviously cannot set the shell thickness too thin as the shells would be too fragile. This is why you may have to design internal walls thinner to achieve the required result.

Round objects designed with a CAD package are not truly round. They are regular polygons that form a close approximation to the required circle. The more faces the better the approximation, but the more information that is required to define the object - there is a trade off on this. All objects are modelled as an accumulation of rectangular or triangular faces, although the triangles may be very very small. If an object is required to be absolutely circular then it has to be printed over size and machined down. A hole would have to be printed under size and drilled or bored out.

The effects on the size of objects are that those forming a plug swell slightly whereas holes contract slightly. There are formulae for round holes depending on whether they are printed vertically or horizontally as the allowances are different for the two. Basically you design the hole larger than required but the end result turns out the correct size if the allowances are used.

You will find that there is a varying amount of trial and error before you get objects that will fit together to close tolerances. Bear in mind that if a metal shaft is turned on a lathe to fit a bored hole and the shaft is only 0.0001 inches too large it will not go in. Most respectable lathes operate to a higher degree of accuracy than our 3D printers.

OK, I understand that a printed object at 2.0 is not going to fit into a hole of 2.0. But When I print things at 1.00 extrusion multiplier (with the nozzle diameter set to 3mm, which is not the nozzle I have returned to) and the extruder steps are calibrated properly, I have a model that has 5 different holes and a peg inside, each increasing the distance of the peg from the hole by 0.1mm from 0.2mm to 0.6mm. At an extrusion multiplier of 1.00 (100%) I can not remove the peg from any of the holes.
Only at an extrusion multiplier of 0.36 does this become possible to take out all pegs (i have tested this meticulously).

This is the reason that I have been trying a smaller nozzle and searching for tips from every corner of the internet.
The only solution I have found to print objects that can fit inside one another is to reduce the extrusion. This compromises the structure of the items I am printing.

The holes will be smaller than you have designed them because they are not truly circular nor will the pegs be the size that you have designed them. I already remarked about the pillow shape of the filament layer so in the fit of the peg in the hole there are four pillow effects involved - the inside of the hole on each side and the outside of the peg on each side - lets say around 2 x 0.08 = 0.16 mm. Now if the material of the peg contacts the material of the hole they will stick. Although the amount of adhesion will be low by the time it is considered all round the circumference and over the length of the hole it will probably prevent removal of the peg. On top of this there is the exactitude with which the hole is modelled in the STL file fed to the slicer which will require further clearance over and above the 0.16 mm.

The formulae for adjusting hole sizes to get accurate results are as follows - where X is the required finished size of hole and Y is the hole size that has to be modelled to obtain it (X and Y both in mm)
for vertical holes Y = 1.0155X + 0.2795 and for horizontal holes Y = 0.9927X + 0.3602. These formulae are for PLA and also take into account the shrinkage of the model on cooling. These give pretty good results but may need a final tweak depending on your filament. These formulae were obtained from a professional 3D printing site.

These adjustments are a fact of life when 3D printing and have to be lived with. The beautiful print of the immediately usable adjustable spanner taken straight from the printer do not apply to our type of machine, only to the much more expensive laser fuse type printers.

Don't make your tests too complicated. Deal with one variable at a time. Start with a round metal rod of known diameter and have a go at designing a hole to be an exact fit for it using the formula (tweaked if required). Then have a go at printing a round plastic rod to be an exact fit in a round hole of known diameter in a piece of metal - the formula will not help you with the rod but start by assuming that its design diameter will have to be 0.08 mm smaller than the hole into which it will fit and take it from there. Remember what I said about the thickness of things in relation to shell thickness.

Once you have solved each of these you will be able to print parts that will assemble together accurately. I recently printed some gears for use with my Meccano, some were for use on round axles and some on the more modern axles that have three flats on them. After only a little experimentation I was able to print gears that were just a nice gentle push fit on their respective axles. I have seen videos of people trying to print these marvelous all in one assemblages of interlocking gears, only to have gear rings crack after cooling due to the shrinkage stresses and lack of clearance. PLA will shrink by about 0.0025 to 0.003% on cooling (in other words something 100 mm long will shrink by about 0.3 mm).

Once your machine is performing consistently and you are familiar with its capabilities it is easy to make the required adjustments to model designs to incorporate the necessary tolerances. I have been printing some smaller than normal road curves for my grandchildren's Thomas road and rail set. I have now got these printing so that after removal of the supports and the trimming off of the brim these will immediately click fit into the original parts.

Don't under estimate the significant forces that can be generated in a model during cooling these can be quite considerable, especially when printing any large amounts of 100% infill. Make 100% sure that all the settings in your slicer are correct before printing including the input units. If you have already had a go at amending your firmware then I would recommend that you adjust the filament steps per mm in the firmware or in the Eeprom if you have one. Once you have it right it should apply over all nozzle sizes. The only adjustments you will then need to make in the slicer will be to allow for variations in actual filament diameter.

It takes only three things to obtain good results - thought, time and patience.

Before this goes to far of the original topic:-
Yes its speed related not the steps. the area of a 0.2mm nozzle is 0.44 of that of a 0.3mm so reduce all your print speeds. Take the current speed and multiply it by 0.44
This will only get you to the same situation as with the 0.3mm nozzle. If there was an issue with 0.3mm you will have the same issue with 0.2mm

I have just switched from a worn 0.33mm (measured) nozzle diameter to a 0.26mm diameter nozzle and had to adjust speeds to more than half to achieve proper printing.
e-steps stayed exactly the same. prints are perfect.. Remember, since each nozzle is slightly different in height recalibrate z offsets. I am using a Kossel mini delta which is the easiest most reliable machine I have. good luck