Bearly a MK3S
See what I did there?
Here's my recently finished MK3S+ clone printing a 13 minute benchy. The only Prusa bits left are the EINSY, screen, Y-carriage/heatbed and some of the motors, so MK3S inspired I guess.
Bear 2.1 frame
MGN12 rails in standard sizes 350 and 400 mm
RPi 3B running octoprint + octodash on a 7" Longruner touchscreen
Heatsinks and forced air cooling on the stepper drivers
All the movements are my own designs. I rotated the y motor and have it mounted to the cross braces for additional stiffness. The carriage is clamped top and bottom to the linear rail block. I eliminated the 2nd y rail, it was just adding drag and there just aren't enough torsional forces to justify it. I have the Z and Y rails attached directly to the frame, no plastic in between. The x rail is bolted straight to the z blocks, again, no plastic. The x rail is flipped from the normal orientation so the block is on the back side. This puts the extruder COM much closer to the rail. I've also chosen the BNBSX extruder and adjusted the height so the belt pulls on the COM.
Although it was never the purpose of this build, I've been working on the #SpeedBoatRace challenge. I've completed prints below 10 min, but they looked awful, so I'm working my way down without sacrificing too much quality. Currently in the 12-14 min range.
I've identified 2 major bottlenecks. The first is the heavy y-axis. At accelerations above about 8k mm/s^2 it starts missing steps regardless of how much tuning I do. The other big one is the EINSY board. Playing around with the gcode I discovered that PrusaSlicer is no bueno for speed. I couldn't run y above 180 mm/s and 3k mm/s^2 with PrusaSlicer code. When I switched to slic3r with pretty much identical settings I could bump those numbers way up. The code from slic3r is much simpler, pointing to the EINSY as a limiting factor.
So I've got a 48 mm stepper on the way for the y axis, as well as 0.9 deg x and y to test out. I'm also about to install a stiffer y motor mount and tensioner assembly I'm printing now. As far as the EINSY I've got a Fysetc spider in the mail with TMC2209s. Once I get that I'll be making the move to Klipper. While I'm waiting on parts I think I'm going to experiment with a printed y-carriage. The weight saving would be huge over the aluminum one, and it shouldn't be a problem making it rigid enough with polycarbonate or CF filled nylon. I'm thinking that bed temp could cause some dimensional changes, but adequate preheating should help with that.