How do you define 3d printer resolution?

[Zak Kus]

I got into this debate with a friend. 

From my understanding, it seems that when you talk about resolution, it's basically the minimum layer height you can print with, which then would just be expressed as something like 100 microns.

My buddy wants to refer to things in terms of dpi, like 2d printers are. He claims the (more business-level) machines he's looked at online use this as a metric.

Are their different standard measurements for different types of 3d printers, or maybe between hobbiest and "professional" grade printers?

[Thomas Charron]

It depends on the type of printer, really. Some printers publish a
'dpi', some don't. Generally, they publish DPI to make claims for
marketing people to understand. One could extrapolate a dpi sort of
measurement by calculating how much travel each axis goes in a single
microstep, and say that's the resolution it's capable of.

Layer thickness is really the 'Z' part of it. Look at it this way.
If I'm using .1mm layers, then I've got logically, 254 DPI in Z. I'f
I'm using .5mm layers, I've got 50. But then again, a Dimension SST
1200 prints fine resolution .254mm and gets really good results.
Newer Makerbot printers seem to be able to rival that my having much
thinner layers, but much 'simpler' extruders.

Really, different printers are good at different things. If I want
something HIGHLY accurate, but not good for say, bolting things to, I
might use a different printer then the same parat, which I want to put
a bolt thru.

--
-- Thomas

[notstarman]

The other factor is the extrusion width. Most machines are very good
at positioning the tool so they can draw two lines very close together
but they deposit plastic lines many time larger then than the
positioning resolution. In milling this is called the kerf of the
tool. On my extruder the effective kerf is .5mm. This limits the
smallest feature that can be created on my printer to being larger
than .5mm.

[lassikin]

dpi is meaningful when talking about some of the other type of 3d printing techniques than fdm.
line width/height and precision of it's placement is more meaningful with fdm printers.

so if you're buddy has been watching some z corp commercials, or for machines like pwdr, then dpi is meaningful(the print head is an inkjet printer head anyways).

-lassi

[Elbot]

You are BOTH correct! Remember it is a THREE D printer. That means dpi = x and y axes while layer height = z axis. All three together gives you the most accurate description of a 3d printer's print resolution. However, professional grade powder and laser sintering printers might have a layer height so fine that x, y, and z axes are the same, so using dpi to describe all three might be ok. On hobbyist 3d printers and older professional printers, the layer height is bigger than the x and y axes resolution. Therefore, it would be dishonest of a salesman to tell you the x and y resolution. It would be more honest to give you the biggest of the 3 numbers only by telling you the layer height only.

In other words, you could use layer height to describe the resolution of any 3d printer while you could only use dpi to describe professional grade 3d printers with a much higher resolution. Now, these are generalizations because if you've been following the 3d printer projects on kickstarter.com, you'll notice that hobbyist priced 3d printers are now capable of laser liquid plastic lithography. Prices are coming down rapidly. Even Makerbot with the Replicator 2 is moving away from plywood boxes due to the tendency of these plywood boxes to expand, contract, and screws loosening. This affects the positional accuracy if you want to print with a layer height of 100 microns or less. I have no problems printing out layer heights of 200 microns with my Replicator 1, but I haven't tried 100 microns with ABS with my Replicator 1 yet (my printer has been broken for 2 weeks and I just fixed it).

[Jetty]

I don't know about other printers, but for the Replicator 1 and
Thingomatic, that
statement is incorrect.

The layer height is governed by the resolution of the Z axis, which
is 4X more accurate than the X/Y resolution on the Thingomatic /
Replicator 1.

Of course, mechanics of the extruder, Z axis, and skeinforge setup
might mean that you
can't achieve that resolution in practice.

However, I've been down to 10 Micron layer heights (0.01mm) on the
Replicator 1 easily in ABS,
which is smaller than the resolution of the X/Y axis.
In theory 5 Micron should be possible too (although I haven't tried
yet) and would likely need PLA.

[Zak Kus]

10 microns? i would love to know how you do that (might be an awesome forum post if it isn't already)

One of the reasons I was arguing with this buddy is that he is an expert wood worker (says there are few things a printer could make that he couldn't build better), and hates lined look of my prints (which i've kinda learned to like). I would love to do a print or two of something at ridiculous resolutions.

[Doogiekr]

Even at .1mm or .05 the lines get really hard to see... And if he is a wood worker, i am sure very few of his final products get out the door without being hit with some sandpaper... Take a bit of sandpaper to an ABS print and the results can be quite good.

[Zak Kus]

That was sorta his point, if he has to sand it anyway then its not the magical make things device that he expects it to be. 

[Doogiekr]

Ah, I see... well I have had good results at .1mm and .05mm layer heights with default setting on my Rep1

 

The only thing I adjust is the filament diameter, number of shells, and infill depending on the object I am printing.

 

The key for me for lower layer heights has been learning exactly what my machine likes for the HBP to Extruder distance (mine is where a .11mm feeler guage has a slight drag on it at any point on the build plate. The lower the layer height, the more critical this gap becomes.

 

I have also found that printing closer to the back of the build platform help a bit with vibrations which can induce some issues into prints at lower layer heights.

 

I have not attempted .01 layer heights (10 microns) but I too would be curious what Jetty does to get that done.

 

I rarely print below .15 these days just because of the amout of time it adds to the print, and I dont see enough of a difference to make it worth it for me... but for a proof of concept print I would like to do something like the Yoda bust or something with alot of detail at .01 just to see what it looks like =)

[Joseph Chiu]

Do you print the entire thing with such fine layers?  Or do you turn on skin and use the fine layers on the exterior surfaces only?

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.

To view this discussion on the web visit https://groups.google.com/d/msg/makerbot/-/C7KK_ivXTYQJ.

To post to this group, send email to make...@googlegroups.com.
To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.

[Doogiekr]

I could never get skin to look as good as if I just printed the whole thing at low layer heights... not sure why... it might just be something I didnt set right.

 

Does not matter much since I dont print at really low layer heights anymore... .15 to .2 is nice enough for me and fast enough to not drive me crazy =)

 

I may mess with skin some more now that my bot is printing well 99% of the time and I rarely have to mess with anything at all

[geneb]

On Sun, 7 Oct 2012, Zak Kus wrote:

> That was sorta his point, if he has to sand it anyway then its not the
> magical make things device that he expects it to be.
>

Print him a herringbone gear, then hand him a dowel and some sandpaper and
tell him to have at. :)

g.

--
Proud owner of F-15C 80-0007
http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.
Some people collect things for a hobby. Geeks collect hobbies.

ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.
http://www.scarletdme.org - Get it _today_!

[Dan Newman]

On 8 Oct 2012 , at 2:48 AM, Doogiekr wrote:

> I could never get skin to look as good as if I just printed the whole thing
> at low layer heights... not sure why... it might just be something I didnt
> set right.

Skin does a linear interpolation between layers and as such works better
for layers for which the changes are, well, linear (straight lines, planes). It doesn't
give satisfactory results on surfaces which do not lend themselves to linear
interpolation.

Dan

[Doogiekr]

That makes sense... Maybe why skin never worked too well for me.

[Cymon]

My problem with skin is that it didn't take the cool factor into account between interpolated layers. So if I was printing something small (which is where I feel the increased resolution could be best employed) every other layer looked like crap.

So, yeah, skin has a long way to go. I love the idea, but it leaves much to be desired.