Nighthawk

Nighthawk is a budget, beginner friendly core-xy printer designed to go fast!

CAD: https://cad.onshape.com/documents/3186eea6377bce7836e0401c/w/5553290dba011a4fef555910/e/45d615a5b9130724f59f9021?renderMode=0&uiState=67fb559247aff81f521ea9fc Note: I couldn’t find a CAD for 150mm stepper, so 200mm is in the CAD. Yes, I know it will hit the x/y axis.

Planning

Desired specifications:

150x150x150 build volume
Core-xy design, with linear rails for all axes and lead screws for z-axis screws
Klipper run on a Raspberry Pi 4B
Enclosed in a polycarb shell (1/8 inch thick?)
FAST!!! Possibly a sub 7 or sub 6 benchy!
Phaetus Rapido Hotend
Sherpa Mini extruder

Toolhead

For the toolhead, I started off with the aesthetic of the Apogee toolhead for the Ender 3 (https://www.orbiterprojects.com/apogee-ender-3-v2/), but since it doesn’t support linear rails (Ender 3 uses v-wheels) or the Sherpa Mini, that was really all I could use.

Fan shroud to house all three fans (CFD optimized) and built from scratch to fit the Sherpa mini (3 hrs)
Rapido hotend mount + linear block mount + BLTouch holder mount (6 hrs) I integrated all the mounts into one to save weight -> print faster!

BLTouch mount (Had to be built from scratch to fit the Sherpa Mini) (1 hr)
Final assembly (with an accelerometer) (30 min) Klipper needs accelerometer readings for input shaping (make printer go wheee), so I managed to (barely) add one in (faster! faster!)

I forgot about belts, redesigned v2 (3 hrs)
MGN12H rails are too bulky/heavy for this size of printer (plus, their geometry straight up doesn’t work with 1515 extrusions), downsized to MGN7H rails v3 (2 hrs)
Needed to slim carriage to avoid impact with joints (so that homing is easier, and faster!) (v4) (1 hr)

Gantry

First, I made an “x-axis” (I’ll import real extrusions later) (10 min)
Then, I had to get working on the two joints. I sketched up a basic sketch of their shape first, I tried to keep it minimal to save weight (20 min)
To determine holes, I modeled the belt paths before making the holes (20 min)
Then, I made the holes, and I have basic joints now! (10 min)
With some chamfers and fillets, this looks a lot nicer (5 min)
Uhh, just realized that I forgot about the actual blocks on the carriage joint, added them now (30 min)
Quick block cad of the drives, the stepper is mounted to the top because it doesn’t clear the height of the toolhead anyway and it’s much better for CG (20 min)

Sketched out motor holes (mounted to the top) (Tensioning should be relatively simple with these style) (20 min)
Made holes and shaft hole + some chamfering/filleting (20 min)
Made a quick idler pulley (5 min)
Mated the two steppers in, added screws, mated in pulleys and idler pulleys (1 hrs)
Quick sketch + block of idlers (20 min)
Chamfers + parts (20 min)
Modeling belts (1 hr, so many things broke :( )
Fully gantry assembled!

Z-Axis

Final assembly:

Built all all bed mounts (From build plate to extrustions), needs to be small (filament efficient) and light (while being strong enough to support the bed) (3 hrs)
Mounted the ball screw stepper next to the bottom-most extrusion for stability + save extrusions (1 hr)
Made a mounting plate for the ball screw stepper (30 min)
Munting plate for the ball screw and nut to the bed (spacing changed, also ugly geometry fixed) (2 hrs)

Frame

I wanted to keep it simple here, 15mm extrusions with a very simple architecture. I tried to use as much of the v0.2 architecture as possible because their frame kits are relatively inexpensive (3 hrs)

Electronics

My computer will crash if I load any more assemblies, so here’s the plan:

Mount a giant piece of polycarb onto the back of the 2 extrusions supporting the Z axis
Stick the PSU, RPi, SKR Pico, Buck converter, and whatever else I need on there. (Stick the psu as close to ground as possible)
Check BOM for electronics details, but I already have most of what I need other than the PSU and Pico

Motor improvements

Z Axis

On further thought, I wanna make sure the Z axis stepper I use is enough to go fast.

I’m gonna first try to estimate the weight of the bed assembly.

1515 extrusions are approximately 0.503 kg/m
100mm = 0.1m (why am I writing this)
3*(0.503 * 0.1) = 150.9 grams
3D printed parts are approximately 62.931 grams in ABS

Add an extra 25 grams for screws, ball screw nut, springs
Approximate weight for bed = 642 grams

Now, using physics (crazy), we can calculate the minimum amount of torque we need to move the bed in my setup, along with some RPM numbers

Let’s convert the values to load: F = m⋅g = 0.642kg⋅9.81m/s = 6.296N
Using some cool formulas, calculate the idealized torque: T = (6.296⋅0.008)/2pi = 6.283/0.05037 = 0.00802N
Convert to g/cm, usually used in torque curves -> 81.74 g/cm

Friction exists. So, we should probably account for 100-120 g/cm of torque.

I’m going to experiment with a different stepper, this one: https://www.aliexpress.us/item/2251832776347337.html?gatewayAdapt=glo2usa4itemAdapt

Unfortunately, there is no torque curve, so I need to find the estimated RPM output for 120 g/cm of torque.

Max RPM is likely between 250–350 RPM at 0.0196 Nm (200 g·cm)
Might reach 400–500 RPM at 0.0098 Nm (100 g·cm)
For 200 g·cm → Max RPM ≈ 300 RPM
For 100 g·cm → Max RPM ≈ 450–500 RPM
Linear speed (mm/s) = RPM × Lead (mm) / 60
Estimated linear speed = 10-16.7 mm/s

Alright, so for a Z axis, this is really fast, so I switched the motor out!

Gantry Physics

Now, for the gantry, here’s the torque curve for the stepper I’m using:

On recommendation, here’s some other torque charts I’m considering:

Kraken:

Wantai NEMA17:

Based on an estimated 0.8 amp average supply to the stepper motor, and assuming the weight of the toolhead and torque needed, I found that the Wantai NEMA17 had the best max RPM output compared to the others (including another variant)

Model here:

The cool thing is that, since it’s a NEMA17, it’s a drop in replacement! And even cheaper than the StepperOnline one I was using earlier!

Vibrations

I got a lot of ideas from here: https://www.youtube.com/watch?v=y08v6PY_7ak&t=789s I still need some way to decrease vibration from the flat extrusions to the ground.

I designed these quick mounts for more surface area to stick on the bottom of the printer:

Then, I did some research into different vibration dampening materials:

Sorbothane: This seems to be the best solution, a form of rubber that has high vibration absorption
Rubber: A little bit too unorthodox for my liking, doesn’t work great with concrete
TPU: This doesn’t really work because I want to maintain the rigidity of the surface mount
Concrete Paver: This will happen, but since it’s not part of the printer, I don’t have to worry about it (yet)

To be really sure, I found some quantitative data. Apparently, there’s something called damping coefficients (read an article here: https://www.sciencedirect.com/topics/engineering/damping-coefficient) that can help me find a good material. I also read through this paper: https://asmedigitalcollection.asme.org/astm-ebooks/book/1813/chapter-abstract/27849243/The-Relationship-of-Traditional-Damping-Measures?redirectedFrom=PDF (didn’t really help, but really interesting)

Sorbothane is about 0.5 (absorbs close to 50%)
TPU is anywhere from 0.05 to 0.5, but usually around 0.15-0.2 (I need consistency, so this isn’t great)
Rubber is close to 0.2 (again, a little unorthodox, and not great damping) We have a clear winner!

Since Sorbothane seems to work the best, I found some cheap sheets of it here: https://www.isolateit.com/products/sorbothane-strip-36-91-4cm-x-2-5-1cm-1-strip?variant=37396536819873&country=US&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&srsltid=AfmBOorRawhGNcH6DOdEWkVSQgJC5gtM31SIxVORnRBf-TZJYa7PYxjfMyw&gQT=1

Yay! Vibration-free printing!

Bom

This took the most amount of time, being a brokie. (15 hrs) https://docs.google.com/spreadsheets/d/19cffJtxG1fblSYRPSsD5pqwqxI_k1R2xgWYntvwAhY0/edit?usp=sharing