Z-Axis upgrades Consolidation

I ended up making a temporary mount with $30 of 1/2" thick delrin. Delrin cuts with standard woodworking tools, so I was able to make something usable, though rough.

If it is too rough, I’ll likely cut it again using either the maslow, or a shapeoko that I have yet to build (it’s been sitting in its box for about 7 years now)

150 mm c-beam 8mm lead, original z-motor
25 teeth on the z-motor and 20 teeth on the c-beam side
This seems to have given me the extra 2mm to end up with exact 10mm travel per rotation of the z-motor.
Changing the $18 to 20 instantly gave me the expected feed-rate results. With a possible travel of 72mm I’ve chosen 60mm travel for testing to stay away from the ends. F60 takes 1 minute for 60mm, F120 30 seconds. It seems not quite linear because to get to 20 seconds I need to use F188 instead of F180 (21.5 sec).
There seems to be plenty of scope for more speed from the desktop tests i’ve done. Your setup looks like to explore where the upper limits are. A thin and sharp bit will show at what speed the sled is lifted away from the sheet.

Thanks @Gero
sounds like good results. at ~7 seconds for a half inch move, that seems pretty reasonable (did I do my math right there?)

I’ll have to time mine to see what 60mm takes, but I am pretty sure it’s less than a minute without changing any code ($18) and just using the z-axis adjustment, which makes me confident that I won’t need a faster motor.

Does anyone remember what the default feed rate is for the z-axis popup screen movement?

I believe i was maxed on F188 ->20 sec for 60mm = 3mm/sec
1/2" ~12.7mm = ~ 4.233333333 seconds

even better! Though I think I was keying off your F120 settings

SO, I managed to find some time today to get the z-axis mounted on the sled and it worked so well that I decided to go ahead and try my first cuts! I posted about that in Projects.

I am very pleased with the speed of the z-axis. I was counting to about 20 seconds for a 60 mm of movement, and during the actual cutting, the z-axis performed very well, and the tabs barely slowed down the cut. For the 1/8 inch distance that the axis had to move for them it was just a second or two.

The setup is the stock motor running to a 60 tooth belt gear to a 30 tooth gear on the lead screw. so i get a 2x speed bump there. The pitch on the screw is 2mm, with 4 starts, so that is another ~8x there. In testing it turned out that 15 mm pitch in the settings got me where I wanted to be with z-axis accuracy. It probably isn’t quite right, but I can work it out in more detail later

Here are some pics and video:
plan view of z-axis: https://goo.gl/9ZFv9e
angled view of z-axis: https://photos.app.goo.gl/c9D4RoszWBzJzvKG8

side view:

video of raising the axis 60 mm. It takes about 18 seconds to make the move, which comes out to a little over an 1/8" per second or 3.33 mm/second (this is with no changes to the defaults):

video of going over a tab (the small amount of time for the z-axis adjustment makes me smile :slight_smile: )

Video of peck drilling a circular pocket (the CAM descends into the workpiece, then circles out the hole, then lifts to safe, at .25" over surface, and goes again, in .1" increments with the last being .05" for a total of .25". Not the best way to do it, but I didn’t feel like fighting with the CAM on it). The hole (pocket) is meant to be 0.17" diameter, drilled with a .125" endmill. The end result was closer to 0.21" diameter. Hopefully I’ll be able to fix this with calibration:


Hi @Keith ,
About the last point: 0.17" hole diameter more in the 0.21" range…
It is possible that the router bit does not cut the exact promised diameter. I have seen that in the Zipper Tree Chalenge. I then cheater bit diameter in the CAM software to change the path distance from the edges.

1 Like

That’s a good thing to know @c0depr1sm, thanks!

I’ll have to dig into my CAM again and see if I just made a mistake. It’s been many months since I actually produced that g-code.

Looks great @Keith! Could you link to the belt and wheels you’re using to connect your motor and cbeam? I have most of the parts to do this project but I’m just certain if I have the motor sticking out the top of the cbeam I’m going to bump into it in my small space and bend something. Having the motor mounted to the back of the cbeam seems like a much better design

Here are the parts I ordered from Aliexpress:

in 30 tooth, 6mm width, and 8mm bore (for the c-beam)
(you could probably also do a 20 tooth for a faster 3:1 ratio, which I bought but didn’t use)

in 60 tooth, 6mm width, 6.35mm bore (you might want a 6mm bore, but the 6.35 worked for me, and allowed a little play to aid in getting the belt on. Also, I had to replace the set screw on this one since the one that came with it was too short to let me tighten it up enough to not slip. I will try to remember to update with the screw size when I get home, but I think that it is an m4)

200mm length

200mm length of travel, 8mm lead

I used one of the original L brackets from the z-axis kit and a piece of aluminum sheet to provide connection to the c-channel. I can get more pictures if you like once I am home.


If you haven’t already, you may also want to look at @gero’s design that puts the belt at the bottom of the c-channel. The nice thing about that setup is that you can easily utilize the whole travel of the c-channel axis. With mine, I lose nearly an inch at the bottom with my current router mount. I am thinking I will redesign the mount to be able to utilize that travel… unless I switch to a spindle which may make that issue moot. Ideally, I want my collet to bottom out at z-zero to allow use of the whole bit.

That is another bonus of @gero’s design in that at the top of travel, the collet should be above the ring to make bit changes easier. As mine is set up, I am able to change the bit without removing the router from the sled, or the sled from the work piece, but it does take a little finesse.

I also like that the vacuum system could potentially run underneath, and that it is less likely to accumulate chips in the c-channel. This last may not be as much of an issue as I fear, though. And a vacuum system would also serve to mitigate the chip issue. I am planning to pipe in some 3/4" loc-line from the vacuum hose.

The only drawback I see is that it moved the CG of the z-axis away from the work surface.

Edit finally found the post that I was thinking of showing @gero’s setup:

1 Like

Still under heavy development. The weekends are to short :frowning:
Vac will be 2-sided plastic pipe ending in a hood from a plastic bottle from the company that was never sued for turning a green Santa red.


I like the upside down motor c beam will help prevent dusts from getting clogged inside. hopefully dust does not make it way to the gt2 rubber belt.


Thanks for the links!! That first link is 20 tooth right?

What was the belt length you ordered? 200mm seems like it would be way too long, so I assume you meant 200mm for the linear rail right?

The links are all to the generic pages, I believe. You then have to choose the options you want. So choose the 30 tooth with the appropriate bore size.

so, 200mm for the belt is right. I think it is the full length (circumference) of the loop, which means it’s about 3 inches between sprockets.

and I also used the 200 mm z-axis

1 Like

Cool thanks for the clarification!

Maybe worth correcting to 20 tooth in this post so it doesn’t confuse the next person

Except that I bought the 30 tooth one. I like having the 2:1 ratio, but 3:1 would be even faster.

But I will change the link to the 30 tooth one. I hadn’t realized I put the wrong link in.

Also, not sure why all the links are showing in Italian now, but they are the same, one just needs to change the language on the page to english (or whichever you prefer).

1 Like

Ah got it, well I ordered the 20 tooth one from the link so hopefully it works well enough!

You probably got this one in 30T W6 B8 size right?

most likely. I am sure you can make it work with the 20 tooth one. Plus it’ll move that much faster. I found that I didn’t need to have a belt tensioner, but if you find you need one, I am certain you can figure it out. drop me a line if you want to commiserate on design.

1 Like

Cool yeah as long as we think torque is enough to move it fine I can’t imagine other issues with the 20 tooth. I was thinking I’d 3d print a mount with mounting slots so I could adjust the motor to tension the belt. This mount is also one fewer part that I would have to strip from my original sled meaning I can keep it in service while building the new sled, right until the moment I need to steal the z axis motor.