DELTA SPEED ⚡ CarbonShrike Delta 3D Printer Speed Testing | 800-1000 x 20-130k

This thing is so light, high accels don't even look cool anymore 😭
Pedal down to find the upper limits of my new delta design, now named CarbonShrike (the Shrike part will make more sense when it's a complete toolchanger). And no, these findings don't really translate into what's usable in real prints, just sketching out the corners of the envelope.
Keywords and stuff: Kossel Delta 3D printer, Ultralight printer, CPAP cooling, Toolchanging

[@lawrencemanning]

Nice to see the deltas get some love. :)

[@jamespray]

They need more, for sure!

[@ManjaroBlack]

Nice. My delta can only do 400mm/s at 120k.

[@jamespray]

Haha, "only"! Still ripping 👌

[@brokencreationlordmegatrol3037]

… v400? :)

[@ManjaroBlack]

@@brokencreationlordmegatrol3037 Super Racer

[@sivisjoe9877]

nice work!

[@jamespray]

Thanks!

[@arthurlugovoy3720]

Well done Mate! What is the effector and rod ends for this?

[@jamespray]

Thanks! This was my CarbonShrike V1 effector design, using Igus Iguball EBRM 4mm ball joints. I wasn't satisfied with the preload on the Igus joints, though (they slopped side-to-side), and I've since moved to a V2 design using MP Jet "short" M3 ball joints with a much more appropriate level of preload. I made it easy to tell V1 / V2 apart by using red for the toolhead locking tabs on the effector.

[@indianmusicjunkie]

Hey, how have you braced the individual towers to reduce vibration. I see the blue 'x'. Care to elaborate the setup

[@jamespray]

Correct, I use wire crossbracing to improve the frame rigidity (the blue is just covering made from drinking straws). My setup uses a single loop of wire that runs around printed anchors attached to the extrusions at top and bottom, with adjustment screws to allow tensioning the wire, as well as lockoffs to fix the wire in place once tensioned. You can get a better look by poking through my delta videos and I have an older version of the system posted to my printables (see link in my channel About page). Hope that helps!

[@indianmusicjunkie]

I saw the setup, thank you. I'm unsure if I'll be able to won't any steel wire locally and easily. Although I found some fishing line rated for upto 120lb. Maybe that helps.

[@jamespray]

@@indianmusicjunkie I'm using 1/16th inch steel cable from my local hardware store, so sourcing may not be as hard as you think :)

[@stefanguiton]

Excellent

[@ManjaroBlack]

Will you be selling any of your models? I could really use a floating extruder design.

[@jamespray]

I would like to make them available at some point, but it may be a while. I don't like to put out "half done" designs! (Also, the LGX makes lovely prints but it's quite heavy and not the quickest.) For the meantime, feel free to copy the mechanics / ask questions if you're not sure of anything.

[@ManjaroBlack]

@@jamespray ah yeah, ill have to do a lot of modifications to make it work on my Super Racer and OMG V2. I tried to pause and zoom in on some of your videos to understand the damping system. I'm not so sure I get it 😅

[@jamespray]

@@ManjaroBlack The current setup doesn't seem to need the damping too much, because I moved the extruder up almost to the carriage plane (this way it moves very little, just pitches/rolls). Basically it's just one part sliding inside another part with lithium grease for a damping fluid, and a light spring return. The "secret" is the link joints -- they are full motion at the carriage end, but vertical pivots on bearings on the extruder end. This means that the extruder is constrained in Z and pitch/roll: its height and angle is set by the carriages (instead of by the effector pulling on it via the bowden). The setup automatically keeps the extruder "aimed" at the effector and within a small range of the same distance from it, too. I found that this gave much smoother extrusion (the bowden is not in tension, so it never crushes on the filament by creating tight angles near the fittings) and it's much easier on the effector and the bowden fittings, too. My old system put all the links and the bowden in tension, and it worked, but it also broke a bowden fitting one time, and caused more print artifacting by bouncing, even when damped, AND it underextruded around the bed edges because the tension pulled the bowden into too-tight bends.

[@rplasticpirate3999]

Neat