Questions like: what determines printer resolution?

[JimmyD]

Hi

I'd like to build a 3D printer using existing electronics, preferably open source software and hardware. I see a lot of printers out there use RepRap controller/driver boards. I have a few questions I was hoping someone could help me with to see if they would work for me:

* What is the biggest determinant of a printer's resolution? Assuming I want to use stepper motors instead of servos, is it the number of positions the motors have? I would imagine the size of the nozzle and the software would also have an impact.

* Is it common for a printer without closed-loop error correction to have errors compound over the span of a print? Is there anything that can be done about it?

* Speaking of software, if a printer's hardware theoretically had the capability to print nearly infinitely small resolution, how small of an increment of movement can G-code specify to it?

* Can anyone recommend any good controller boards I could use to make a printer that could print down to 50 microns? How about down to 10?

* I have found only 1 example out there of someone who used a Raspberry Pi as the "brain" in their printer. Does anyone know of any others? I'd like to

Thanks very much for any help. I'm really looking forward to understanding how to make an awesome printer!

[Traumflug]

* What is the biggest determinant of a printer's resolution?

1. Steps/mm of stepper motors. Servo drives aren't infinitely accurate, either.

2. Stiffness of the printer frame. That's why all the good models come as box with (partially) closed sides.

3. Stiffness of the driving mechanism.

* Is it common for a printer without closed-loop error correction to have errors compound over the span of a print?

Properly calibrated, steppers keep their position virtually forever.

* Speaking of software [...]

* [...] controller boards [...]

Software and controllers can achieve any precision you desire. Easily down to 1/1000 micron if your hardware were capable of that.

Nozzle size determines the size of the smallest printable feature. Precision is independent on this as long as the relation is reasonable. Printing at 1 micron resolution with a 10 mm nozzle may not work, printing with 1/10 resolution with a 0.5 size nozzle certainly does.

* I have found only 1 example out there of someone who used a Raspberry Pi as the "brain" in their printer. Does anyone know of any others?

They also have a separate or backpacked controller. A RPi simply plays the role of a PC. So far controllers with attached LCD are simply more in fashion, even knowing that handling a display dents into a controllers performance.

[Sebastian Finke]

To add to the above:

Most printers are really pretty good in terms of positional accuracy and imo easily as accurate as proper CNC machines. Ultimately the problems come down to the extrusion process itself. Melting filament, squashing it through a tiny hole and mashing it against the HBP or previous layer is literally a shot in the dark when you consider all the factor tht can affect it including, but not limited to: filament diameter, nozzle temp, ambient temp and surface quality being extruded on to. Variations in these and your accuracy is out the window depending on your tolerance, of course. Now throw in vibrations from the printer, knocks from outside sources (kids, pets, etc), corner ringing...

You could probably get repeatability of +/- 0.1mm, possibly better if you take the time to calibrate and XXXXX-proof everything.

[curious aardvark]

Now throw in vibrations from the printer, knocks from outside sources (kids, pets, etc), corner ringing...

Variable temperature and humidity in your workshop.

Considering all the things that conspire against them - 3d printers really are remarkable machines.

[Davo]

1. Steps/mm of stepper motors. Servo drives aren't infinitely accurate, either.

2. Stiffness of the printer frame. That's why all the good models come as box with (partially) closed sides.

3. Stiffness of the driving mechanism.

Properly calibrated, steppers keep their position virtually forever.

Software and controllers can achieve any precision you desire. Easily down to 1/1000 micron if your hardware were capable of that.

Nozzle size determines the size of the smallest printable feature. Precision is independent on this as long as the relation is reasonable.

Very well said.

[Traumflug]

Now throw in vibrations from the printer, knocks from outside sources (kids, pets, etc), corner ringing...

This nicely describes why stiffness matters. A sturdy box won't notice when you knock against it.

Maybe that's a bit hard to believe, because many common 3D printer designs do not exactly shine regarding to it. For example, the Prusa i3 gained a lot of stiffness in the X-Z-plane compared to the i2, but lost most of it in the Y-Z-plane. i3's with additional struts exist, though.

[JimmyD]

Thanks for all the feedback folks. I've got tons o' questions, and this really helps me to get going. So as long as my steppers are decent, and I have a pulley and belt system (assuming that's what I'll use) with very little play, I should be okay making a non closed-loop or "PID" printer and be able to reasonably expect few errors from that part of the machine? Awesome.