Love it.
I like that it naturally means the end of the lead screw. So you won’t overrun it.
Watching it clear the endstop the cycle before makes me think it isn’t perfectly tunable. But I think that is just in my head.
For a 1 or 2 start, the paddle would come in and could miss the endstop metal clip and but either the mounting screws or the endstop body. A different endstop design might fix that.
Like maniacal evil genius laughter, or like watching a tiktok of a turtle farting type chuckle. Not sure if you are good or evil genius yet.
Jim’s fix handles this and gives us a manual axis mover as well. Point it up and trigger off the bottom of the disk. No debris to worry about either.
You can’t tell, but those helicopter pieces are threaded, which helps it grip. It also means there are four ways to install (along each of the four starts). Three of the starts don’t work because it collides on the cycle before triggering.
Adjusting means unscrewing (like a nut) and then re-tightening the clamping screw. Since it is traveling along the threads when you do that, it stays in the right phase relative to the endstop, so it will still clear and then trigger.
If the helicopter piece were smooth instead of threaded then you would have to gamble when you make an adjustment.
Maybe the endstop could be mounted at an angle so that the metal switch arm is facing somewhat, but not fully upward. Then even with a 1- or 2-start screw it might be assured to hit the endstop switch before crashing into the endstop body.
By the way, these are 150 mm screws and after moving the helicopter pieces up a bit, I have about 92 mm of Z travel.
Might I suggest what the part look like?
On second thought, if the switch is facing upward, the amplification is still possible if the underside of the cap is not flat. It could be a helix with the same pitch as the leadscrew, so it can’t trigger or collide along the helix, and then a steeper section where the trigger happens. Depending on the radius where it hits the switch, the axial contact could retain the high amplification.
With the switch facing up, there is more room for the body of the switch to hang downward along the side of the Z stub. It has to clear the coupler when Z is near the bottom, but it’s definitely possible and maybe not difficult to find room.
It doesn’t even need to be a full helix, just a ramped nub with a protrusion sized for the leadscrew pitch plus some margin.
Edit- but I’m not sure that’s necessary. We’d be using either the roller or spring arm in it’s designed way. We’d be down in the sub-mm/steps range where the repeatability of the endstop and the calculation down in the firmware is more a factor than the multiplication of the motion.
This is probably worth testing. The roller may even have an advantage here with a spinning disk as the contact point.
True, it technically doesn’t have to be a helix.
I was thinking with a helix most of the surface is “not moving” (relative to the Z stub) as it turns.
Mainly I was thinking it can’t crash into anything when it’s “not moving” but another side effect is that with the switch positioned properly, you can overrun the endstop and it will be held in the activated state for most of a full turn. That means you can overrun by a lot which in turn means that for the seek portion of homing you can go at extremely high speed without worrying about crashing.
In the original helicopter concept in the video with tangential contact, it crashes less than 0.1 mm after triggering!
So are endstops really necessary? on my mpcnc I start it where 0,0 is. can you do that with the lowrider. Make some gages that go under the carriage and hold it at 0. Then start the machine. Bang you are now at 0,0,0 until you do the Z calibration with the bit in
by gages i mean printed blocks, both the same size
or just make stops built into (somwhere) that the carriage rests on for zero.
Not 100%, but squaring the Z is WAY easier.
Having a known zero point where the gantry is definitely level to the spoilboard and being able to find it again makes for repeatability of operation that is a huge advantage for CNC.
Looks like I may want to give up the 2 start lead screws, but I’m OK with that with good homing.
Is no Z endstops possible? Assuming Z couplers were fastened to the stepper shaft consistently using the YZ guide, then, could Z homing down to negative height until both steppers are skipping steps result in level enough gantry? M666/offset to compensate during initial calibration for assembly caused Z level issues.
To avoid bit contacting spoilboard, would need to home X first and rely on a margin gap around the spoilboard, or, make a hole for bit to sink into in some XY location homed to first. Bonus marks for Z probe calibrating in that spot while Z leveling the gantry.
If you ever pick up a big job the next day homing to Z max is critical.
Home to Zmin at xMin,Ymax where Ymax is 1" beyond spoilboard Y length?
“Home” as in crash down until steppers skip, at XY pos where bit doesn’t hit and gantry doesn’t hit spoilboard/workpiece. Pros: no endstops, and max Z travel possible. Cons: Less accurate than endstop trigger, but good enough given gantry span?
That means the whole beam, core, and router need to be able to hit Z min. That is a lot of wasted space to save $2 in microswitches. If we say it can handle 100-105mm it has to be able to hit the entire volume.
Is the bonus the “ping” sound that the endmill makes when you change the angle in the hole levelling the gantry? 
Though if you are going to keep a home position, a couple of blocks to rest the gantry on was a standard for the LR2 with no end stops.
Resting blocks at Xmin,Ymax and Xmax,Ymax that can also be used to Z Level at Ymax? Sounds better than making calibration hole in the table. Saving $2 is a bonus, but saving assembly and operation time, minimizing effort to reliably calibrate is the main goal.
OK, no more posts from me until I have something moving… Want to see my RC2 cutting stuff before I print updated parts.
Precision parking blocks does not sound easy and there is no way to do that without loosing some cutting area. I will be adding Z endstops, parking blocks are not going to be an option from me.
FWIW, I’ve not used Z endstops on my LR. I use blocks at both ends before energizing it, and I move them before I cut.
But the people want it.
@jamiek, am checking out Beta folks’ wiring setups… Looks like your JackPot is mounted near center of the gantry. Is that a temporary location, or are you planning a center mounted drag chain above/below the gantry? If drag chain below, then only Vac hose would be on top?
I am somewhat undecided but I will probably keep it in the center. The motor wires from both ends can reach the center with no extensions necessary.
Also a drag chain that’s roughly half the length of the X axis can route the wires from the center to either end. With the controller at one end, the X wires and drag chain have to be the length of the entire X axis.
It’s a tiny advantage but at the moment I don’t see an advantage to putting it at one end. Currently I’m accessing the table from the front (Y-min) primarily and not from the sides (X-min or X-max) so it would not be more easily accessible if it were on the side.
