The Z-nut does have a lot, and I will upload my mod soon, since it seems to be working well. But I noticed that there is a good amount of rotational play at the motor to threaded rod mount, and this will add some too… especially with small movements.
The motion of the x-carriage is what deploys and retracts the probe.
There is a sliding cam piece that can be seen in the pictures. It gets pushed by a fixed “tab” on the max X side that is mounted to the aluminum extrusion upright to deploy the probe. It is on the left side of the screen at 0:16 in the video (though I think I may change this tab piece). And to retract the probe, the X carriage just has to move to X-zero and that sliding cam piece is pushed the other direction by the “x motor for spring” part. At 1:06 in the video, though the stepper motor is blocking it.
BTW, It is VERY nice to simply hit print and watch the machine do what it should. no more fussing with the Z height for a few minutes before each print, and then watching the first few layers, and maybe killing the job and re-adjusting…just hit print, and practically walk away.
And this is with the stock plastic bushings in the Z and Y axis (swapped the X carriage bushings to ball bearings when I went to a dual extrusion). But I am swapping all the bushings to ball bearings for hopefully more consistency very soon.
Sounds great. I’m still wondering how you got rid of the z axis hysteresis from the screw threads/nut? Mine has a LOT of hysteresis. Its not a problem when its always moving one direction but it would never work if it had to move both ways. Let me know what you did for that.
I saw a video on youtube where a guy started with a purposely unlevel bed then did a print from there. Do you have any video showing a print first layer in action where the z-axis is moving to compensate?
Thanks for posting
I would LOVE to get my hands on this! Any plans for making a version that will work with folks only using a single extruder?
Awesome job!
g.
Basically I Extended the Z nut in Z, up and down, to take advantage of the full surface it is mounted to, added another Brass nut, and removed one of the springiness loops.
See the picture below:
the 4 holes are for pins during assembly, so the spring does not push the lower nut too far up, and the upper nut is kept in position.
The number of extruders has no effect on this mod. It should work well with single extruders…though currently it does require my X stiffened pieces (just because the cam piece pushes on them). it would be easy enough to get to work with the stock TAZ parts.
I eagerly await the posting of the part files for this. 
g.
Nice work. Any chance of seeing you print a couple layers with your new system? Thanks for sharing!
I have a small video I am uploading. basically a continuation of the previous video. you can see the Z moving as it prints. But the Z does not need to move much on this machine. the bed on the TAZ is decently flat.
In the video I saw of another system that used this, the guy purposely made the bed over 1 mm out of level and then went through the automated leveling process and showed the first few layers printing. It was very impressive to see it printing a perfect first layer knowing the bed was off level.
I think he had to manually actuate his surface probe though. Your automatically actuated probe is very nice.
there are some good video’s here with another auto pop out probe. https://www.matterhackers.com/news/automatic-printer-calibration-update
That one uses magnets though, and has to raise the z really high to pop the probe out and put it away. I wanted to stay away from magnets just to simplify things, and I also did not want to wait for the Z to raise and lower.
My printer is working so well now, that I do not want to skew the build plate lots just for a video right now.
but here is the video of it printing just after the bed probe previous video. you can see the Z rod moving slowly.
http://youtu.be/c8dM1KpBXJ8
http://youtu.be/c8dM1KpBXJ8
Part on the left is before the bed probing. part on the right is after.
the big difference is that there are no more squished first layers, and the heigth is much more accurate. The part is supposed to be 6mm thick. Part on the left is 5.6 to 5.7mm. part on right is 5.92 to 5.98.
Just started uploading some of the files
First up:
Anti-backlash Z-nut and motor to M6 couplers (needed to make the Z-probing compensation accurate):
http://www.thingiverse.com/thing:117950
Bed probing parts:
http://www.thingiverse.com/thing:117957
Now you know, you need to modify this part to attach onto the bed probe/bearing bracket and allow a 40mm fan for PLA…
http://www.thingiverse.com/thing:86875
ps, I have Alibre and can save you a format of your choosing, the original is this set of files along with the Alibre version…
http://www.thingiverse.com/thing:63123
Also, and I couldn’t get these to print just right, but they looked to be pretty nice…
http://www.thingiverse.com/thing:22422
Alan
I did start a fan design…but the bed probe is now in the way.might be able to mount it on the sides though.
It looks like you need to print that anti-backlash part with support?
Alan
support is already built in. except for the holes. but they print OK without support. These printers can do a 45 degree angle with no support OK.
Could you post up the gcode you are using for the z nutsprings? We aren’t having much luck with printing one that will let the m6 in without a lot of cleaning. thanks 
I printed the parts on a dimension machine, so I do not have code to put up.
If you can measure the size of the hex (maybe only print a few layers up to the hex) I can modify the file and re-post it.
Thatd be great! The hex is measuring at 9.1mm using
http://devel.lulzbot.com/TAZ/test_parts/z_leveling/Z_NutSpring_1013_M6-116s-cc.gcode
and
http://devel.lulzbot.com/TAZ/test_parts/z_leveling/z_nut_spring_m6-116s-.50noz-t.ini