NOTE: This thread is unedited since its original posting, and the fixes detailed here are now kept updated in a Community garden project:
Check it out on the Community Garden page, the details here are outdated!
The router motor is inside of a metal sleeve, and the sleeve has a wide slot cut out of the side. There is a rectangular cutout in the router motor body that travels up and down, visible in this cutout. A spring-loaded clip pushes a rectangular-shaped tab into the rectangular-shaped cutout in the motor body. This clip is what the z-axis adjustment screw (technically a ‘worm’) pushes up and down when the screw rotates.
These are not precision-machined parts, and either the cutout in the motor body is too big, the tab on the clip is too small, the hole in the clip has problems, the worm can slide up and down parallel to the router body - or a combination of all four.
With my Ridgid R22002, the problem was the hole in the clip used by the worm. In testing, here’s what happens when we zero the bit and drill two 10mm holes: (numbers rounded for illustration)
- The bit is 10mm above the surface, The clip is perfectly perpendicular to the router body (level).
- The z-axis is told to move 5mm down. The worm side of the clip moves 5mm down while the router body doesn’t move.
- The clip is now tilted “down”, with the worm side 5mm lower than the router side.
- The z-axis is told to move 10mm down. The bit is now zeroed, just grazing the surface. The clip is still angled 5mm down, lower on the worm side.
- The z-axis is told to rise to its traverse height of 5mm. The worm side of the clip is raised 5mm, and the clip is now level. The bit is still zeroed on the surface of the wood.
- The z-axis is now told to drill a 10mm hole. The worm drive turns enough to move the worm side of the clip down 10mm.
- The first 5mm of downward travel is used to angle the clip down on the worm side, with no change in the depth of the bit. The remaining 5mm push the bit 5mm below the surface.
- The worm now retracts 15mm, 10mm to clear the surface and 5mm to get to the safe traverse height.
- The first 5mm is used to make the clip level, perpendicular to the router body. The bit depth has not changed and is still 5mm below the surface.
- The next 5mm is used to angle the clip up, now with the worm side 5mm higher than the router side. The bit depth has not changed and is still 5mm below the surface.
- The final 5mm pulls the bit up 5mm. The bit should be 5mm above the surface but is zeroed on the surface.
- This process will repeat in the second hole.
This is the best-case scenario. Here’s where it gets worse:
Shortly afterward, you’ll notice the holes are 5mm too shallow. Adjusting the zero point 5mm down will be a mistake - the bit will drill 10mm-deep holes, but carve a 5mm deep path through the material in-between holes when it should be 5mm above the surface.
Moving the bit up until the bit is flush with the surface will leave the worm-side of the clip 5mm higher than the router side. Zeroing here will mean a 10mm travel of the worm before the bit moves. Going from a 5mm traverse height to create a 10mm hole will then be 15mm of worm travel:
- The first 10mm will move the worm-side of the clip from “up” to “down”, with the bit still 5mm above the surface.
- The final 5mm of worm travel will zero the bit on the surface, when it should be 10mm below it.
- Telling the z-axis to rise to traverse height will then put the bit perfectly 5mm above the surface and leave no hole at all.
This is the part where you start going crazy.
Both of these pictures are with the clip in the same place, and with finger pressure on the router tab either way.
@arnoldcp Did a great write-up on how to bore out either side of the clip, leaving a small lip inside, and gluing speacers into the bored-out hole to make it wider.
Here’s my problem: I don’t have the facilities to precisely machine out the hole in the clip exactly at a 90-degree angle and to a certain depth so the spacers have a supportive lip to support them.
A few trips to Lowe’s later, and here’s what worked for me:
Play in the worm housing
The worm is not secure in its mounting. It can move parallel to the router body by a few mm, and you can “rattle” it by hand, up and down.
The top of the worm fits into a larger hole on the upper mounting. The bottom of the worm has a smaller diameter end which goes through a smaller lower hole and is secured by a C-clip on the other side. When pressed all the way down, there is ~2mm gap between the lower hole surface and the c-clip. This movement needs to be eliminated so the worm cannot slide in parallel with the router body.
Disassembly is easy - use a (preferably non-rusty) small screwdriver to pop the c-clip off. Put your free hand over the top so it doesn’t go flying, and try not to stab yourself with a rusty screwdriver.
A 12-pc bag of 5/16-inch Hillman zinc washers from Lowes is $1.28. The barcode is 0 08236 59004 3, and my receipt shows a Lowes item number of 58125.
The washers fit around the non-threaded part of the worm. Placing one washer on the top of the lower mount and installing the worm presses it between the end of the threads on the worm and the lower mount. This puts the groove that holds the c-clip about 0.25mm inside the hole of the lower mount. This is enough to still get the c-clip in, but under pressure - leaving no room for travel in the worm.
I chose to put the washer on the thread-side of the lower mount, because I didn’t want to be pulling the threads into the lower mount for an extended amount of time as they’re rotating. (I didn’t want to make the threads “drill out” the lower mount)
Play in the worm-side of the clip
The goal here is to keep the clip perpendicular to the worm and router body at all times. The easiest way to do so is to make the hole longer and have it fit the worm as closely as possible so the elongated hole can’t tilt on the worm shaft.
The threads on the worm have a 1/2-inch diameter, and the clip has a 1/2-inch hole for it. The clip’s hole has 1/8-inch of clearance all the way around its perimeter, making for a 3/4-inch area to work with.
A Hillman bronze machine bushing from Lowes is 1/2-inch long, has a 1/2-inch hole through the center, and has a 3/4-inch diameter flange on one side They come in packs of one for $3.91, and it has a Lowes item number of 1772.
Attaching this bushing to the side of the clip will extend the length of the hole from 5/8-inches to 1 1/8 inches, where the latter 1/2-inch of the hole doesn’t have any other mechanisms to present any additional slop, like the clip does.
Do not attach the bushing to the lower (bit side) of the clip. If you do, it will hit the lower worm mount and not allow the router to extend its bit to the worm’s fullest extent. This will likely prevent the bit from reaching more than 1/2-inch into the material. Since most material used is 3/4-inches thick, this wouldn’t work.
I carefully glued the bushing to the top (z-axis motor side) of the clip with Gorilla Glue, ensured it was aligned properly, and that an excess amount of glue hadn’t seeped into the inside of the hole. Clamp the parts together and allow to dry overnight - Gorilla glue expands and there’s no room for the bushing to be even slightly off-axis with the clip.
Don’t leave the orange button and spring in like this - see below.
There’s nothing good that can happen to it while the assembly is being glued and finished, wait until it’s time to put the assembly back on the worm.
Gorilla glue will expand into the hole and harden. I waited for it to do so and drilled it out. I didn’t have a 1/2-inch drill bit, so my 3/8-inch bit did fine. Make sure not to twist the bit in the hole assembly or you’ll flange the hole out and make it looser on the worm shaft.
Grease the shaft before assembly. (I’m concerned about sawdust contamination in the grease, so perhaps a stronger spring is needed, but…) The clip has a small spring that presses its tab into the router’s tab cutout. The spring is not designed to be strong, it’s meant to be easily overwhelmed and allow the tab to pop out if the router is trying to jam into something too forcefully. (This is called a “mechanical fuse”) The added drag of the longer hole on the shaft is something the spring struggles to overcome, and the clip can stay stationary instead of rotating into the router body’s cutout.
The router’s sleeve has a clamp that’s meant to hold the router body after it’s been adjusted. The z-axis motor isn’t strong enough to fight this, so I used a ratchet and manually tightened the nut on this clamp so the router sleeve is a closer fit to the router body. I leave it unlatched, and just tightened the nut on its bolt to my desired tightness.
(clamp is seen top-right below)
I’m waiting for some tough, slippery tape to put strips on both the sleeve and router body so it can slide better, but I suppose I could grease it as well.
The clip’s tab is slightly smaller than the cutout in the router body. This leaves a tiny amount of play (<1mm for me) That’s enough where I adjust the router for proper depth, and just set the traverse height to 5mm and let it be slightly off. Not sure how to remedy without compromising the function of the mechanical fuse.
Excellent and detailed write up with great pictures!
Also a much more scientific write up about what is really happening with the stack up of “lost motion”.
you were supposed to loosen the nut so that when the clamp is closed it’s a snug
how does the rest of your work compare to just putting a bungee cord over the
router to take up the backlash?
I found it easier to make minute adjustments to the clamp’s tension with it open, instead of snapping it shut, opening, adjusting, closing, checking, opening, etc.
It may have been my imagination, but I could have sworn it didn’t shut exactly the same way every time.
I just tightened to my preference and run it open.
My router didn’t work with the bungee method. Maybe a bad bungee, but I wanted to try and make the link stay perpendicular so the performance would be repeatable. I didn’t know if the degree the link was angled would consistently stay angled that way, etc.
Now I don’t have to mess with it, and it cost $6 to fix.
I have used the Bungee method, and, along with a lot of work on smoothing the sleeve, filing out the slot in the sleeve to eliminate a casting error, and trying various lubricants, have come up with pretty reliable Z axis operation. Nonetheless, what you have done seems to add some needed precision to the router’s movement, and I will consider this a needed upgrade when I get back to my machine in a few weeks.
Thanks for the effort!
I’m going to have to give this a try! I want to build the @meticulous z-axis” which is a great design. But for now Tim his could be a quick trip to Lowe’s to take care of some of the issue. Thanks so much for the awesome walkthrough.
You may be able to use a small c-clamp on the flange lip. Not as permanent as gorilla glue.
Exactly the “fix” employed at The Maker Station in Marietta, GA. A simple, inexpensive, easy and effective solution.
… and don’t forget that when you use the bungee cord, if you don’t get the tension right it puts too much pressure on the router. I have had 2 things happen: When the worm gear turns to raise the router, the arm that locks to router in place pivots out just enough to disengage from the router and when the worm turns back the other way to feed the router into the work, the router does not move because it has disengaged itself. The second thing that has happened to me is that with the bungee cord, the pressure (and it does not take a lot) wears out that PLASTIC orange clip (you know, that little tongue on the inside of the clip that rides in the groove of the worm) and soon gives way which causes the router to ram into the work piece at full depth. If your not standing there, Maslow with still continue to mill burning up your bit and create another zombie apocalypse!
I’m converting to Maticulus Z. Eliminates all these issues! … hopefully!
In addition to adding the washer to the bottom, we glued your recommended bronze bushing to the top of the z axis clip to increase surface area. Instead of clamping - we used a spare 1/2 bolt and nut that we preemptively oiled just in case any glue seeped through - and then tightened a bit to dry over night.
That’s a great idea @ChuckC!