A full Tap probe build where the nozzle itself homes Z

Oct 26, 2026 Last reply: 2 days ago 3 Replies
Jump to replies (3)

The detail that jumped out at me first: this printer homes Z by tapping its own nozzle against the textured PEI build plate. No separate probe body touches the bed. The maker describes Tap as both a Z endstop and a bed mesh probe in one, and in testing the printer auto-homes exactly that way, with the red optical sensor LED lighting up as it triggers.



The assembly is a satisfying chain of small steps. Printed parts in CF-ABS from 3D Printing Canada get support material removed, then brass heat-set inserts are pressed in with a soldering iron. Two 6x3 mm neodymium magnets are glued into small printed magnet holders with cyanoacrylate applied by toothpick. An MGN9 rail with its carriage is bolted to the printed center mount, and that mount goes onto the X-axis MGN12 carriage while the GT2 belts are tensioned through the integrated slots. A printed retaining tool holds the carriage during the transfer onto the mounted rail, which seems like a smart safeguard given how easy it is to lose bearings when a cartridge comes off a rail. The front plate, upper bracket, and optical sensor PCB get screwed together, the Stealthburner toolhead is reattached, and the magnet retention screws are loosened so the magnets can pull into alignment against the front screws.



Two things stand out as honest trade-offs. Manual lifting shows 1 to 2 mm of vertical play, and the maker stresses the tolerances are very tight and rates this an advanced-level mod, agreeing with Voron's own assessment. The BOM and price estimate had not been released at the time of the comments, so cost is still an open question, though the maker says it is not very expensive in the grand scheme of things.



The maker calls Tap the best Z endstop and bed mesh probe they have ever used, and the belt routing, in their words, worked very nicely.



Project by Maple Leaf Makers. Watch it on their YouTube channel:

formatting link


The printed retaining tool for the carriage transfer is the cleverest bit here. Sliding a carriage off a rail and chasing bearings across the desk is the classic way to ruin an afternoon, so having a printed part hold everything together during the move is a small addition that removes the worst failure mode of the whole install.

How is the 1 to 2 mm of vertical play handled in practice? If the nozzle is the thing tapping the plate for homing, I would expect that slop to matter for repeatability, yet the maker reports clean auto-homing. Is the play only present under manual lifting and not during an actual probe move, or does the magnetic preload take care of it?

Since the maker has tried inductive, Klicky, and Klicky NG probes before this, it would be useful to see a direct probe-to-probe comparison on the same printer: mesh consistency and Z offset drift over a few weeks. That would give anyone deciding between Tap and an inductive probe real numbers instead of impressions, especially with the BOM still pending.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required