Belt Driven Z Axis

Hello All!

Its been a while and now that the weather is finally warming up I have been able to spend longer hours in the garage in more comfort. From this I have now upgraded my momus to FULL belt drive. Sourcing ACME rod is somewhat of a burden in my location so I wanted to experiment and see how well a belt drive would hold. I've decided to break this into a new thread as its easier to find via a google search...not many belt driven Z axis machines out there...I've looked.

The first thing I had to do was stiffen my x axis carriage as it had a large HDPE block in the middle under compression...I've since replaced that with solid steel square. There are 2 thread rods that run thru the centre of the square and allow me to compress the entire assembly together very very tightly.

https://farm4.staticflickr.com/3679/...b5381303_c.jpg

After this I added a 3/16" alum plate to the front of the assembly to hold the new belt drive. Here you can see the general overview of how it comes together. The bottom bearing hole is oval and allows for tension to be placed on the belt after assembly. The black socket cap screws are threaded into the steel bars and hold very firmly.

https://farm4.staticflickr.com/3795/...a28876be_b.jpg

Due to my very limited space between the carriage and the z axis I had to mount the stepper slightly behind the z axis plate (5/16" nuts as standoffs). This makes it a bit of a pain to assembly, but not something I need access to all the time.

https://farm3.staticflickr.com/2902/...52f09883_b.jpg

I've installed my original oak z axis beam and plywood router clamps. I still plan on replacing these but after all my testing they are still not the weak point in my re-design. The bearings supporting the z axis are still the location of greatest flex (notice the double nut standoffs) but its very very small at that...not something I notice unless I rapid into aluminum...which I try and avoid.

https://farm3.staticflickr.com/2919/...00a527d2_b.jpg

With everything rebuilt I was eager to see the results. I first 3D milled some foam and once I was satisfied that nothing would run away on me I moved into some aggressive cutting into some exotic hardwoods. The machine was behaving just as a screw driven Z would and I had no lost steps. For the curious I was cutting Paduk at 100-200 IMP with a 1/4" DOC with a 1/4" 0-Flute endmill.

The dial indicator was showing that the machine could jog (z axis) accurately down to 0.001" with backlash sitting at about 2-3 thou over its max travel...not prefect but would it be precise enough for my most challenging work? Could it mill PCB's while using its maximum rapid speed?

https://farm3.staticflickr.com/2917/...b96f18b6_b.jpg

https://farm8.staticflickr.com/7012/...133c2f2b_b.jpg

Here I have to say I was totally impressed. Normally I cut my PCB's on my micro desktop machine but I think I will now let this machine take over. Also note the diagonal line cut across the PCB...that would sadly be operator error.
You don't gain much in the way of speed as you can only push the little v cutters so hard but its nice to plop down full sized PCB stock and run multiple boards at once. I made sure to include some fine text to see how the machine would cope, so heres a couple close ups...you can see the stepper motor driver's step breakdown on the diagonal passes (currently running 1/64th micro stepping).

https://farm4.staticflickr.com/3714/...27c5530a_b.jpg

If you look close enough you can see that the machine took 5 passes per trace to fully isolate to the settings I require (milled from eagle CAD's pcb-gcode plugin). The board was then soldered and conformal coated hence the sheen in the pictures. Text engraving was done with a single pass. All cutting done with a precise bits 60 degree v cutter...delicate but AMAZING cutters. Ruler in the next photo reads in mm...the centre of the 'o' is only a fraction of a millimetre large!

https://farm8.staticflickr.com/7085/...acfbfc4b_b.jpg

All the above cuts were taken at 0.005" DOC (except the drilling, that was 0.0625") at 20 IMP cutting and 400 IMP rapids. The only thing you need to remember when running a belt driven z axis is that the head is going to drop when you disconnect power. On this machine it comes down slow enough that its not going to snap any bits larger then 1/8" but it will easily snap my PCB bits so I'm sure to remove then from the collet before shutting down.

I've also included a short video showing the machine cutting the above pictured circuit. Next step for me is to now finish the top of the enclosure as my z axis is short enough to fit :) yay!

http://youtu.be/lsfxo5WzFb8

Thanks all!

Hi Freerider,
very impressive work, I can only guess that you pretty much intimidated most of the folks around here with the results you got with the new belt driven z-axis you came up with.
Perhaps that's why everyone is speechless and nobody has posted a reply yet, lol.

To be able to cut a PCB of that quality (all hats off, seriously) with a stick of wood as the core of your z-axis blows my mind too.
As you know, I've messed with the carriage design too, but what you've come up with and the simplicity of the mod is very cool, awesome job... :-)

The part I don't quite get is why you went through all this trouble with the belt drive.
I mean, you could have just asked one of us to get you the fairly low cost screw (most of us have a large cut-off piece of the screw sitting somewhere in the garage) along with the backlash nuts instead of this.
Don't take this the wrong way, you must have had your reasons and it came out pretty darn awesome, but the original design is pretty darn solid too and works very well.

For myself to attest, the one sided screw with the thrust bearing design has not caused any trouble so far and has been repeatably accurate to the thou without any issues in my machine.
I've done 100's of complicated carvings by now, some taxing the z-axis for several hours, I can't get it to fail or backlash in any way.
I've also cut PCB's too and know what happens if there's just a minute amount of backlash in the Z, so I get your point.
The original design has none, so I still don't get it - unless there was no way to get the parts for you.

In the end, it's just what you need it to do I guess. :-)
Awesome work once again and impressive video editing by the way ;-) I wish I could do that properly yet, lol.
--
Mac

Mac,
As always thanks for your kind words. I figure people are out enjoying the sunshine and not huddled around the glowing screen of cnczone posting away:).

As to the question of "why the belt drive", honestly its just because I wanted to. I'd like to say first off that it would be very hard to improve on the original Momus design so this was more of an exercise in could it be done, and less of one in practicality. I agree with you 100% that the side screw design is great and works very well (as you have attested to) but I wanted something different. Another issue I had was that with my redesigned machine a side mounted screw would eat up quite a bit of my x travel due to my fat router clamps and wider carriage. Nothing I couldn't work around I'm sure but the belt was the path of least resistance in this case. Clearance issues were also a factor in my belt design, a side mounted screw needs to be either out further on my machine, or off to the side more to clear my larger router. Both of these kill off cutting volume and bring my z higher then the top lid will allow for...a sane person would just build a stock machine and avoid all this mess yet I'd do it all the same if I were to start from scratch...possibly get rid of my dead tree z stick ;).

The final point in this modification (perhaps carrying the most weight)...it cost me nothing. I know for a fact I can get trap/acme rod and nuts sent to my door but I had the belt and an extra pulley sitting on my workbench looking for a purpose...so that naturally won out.

Thanks again for the feedback!

Why is your motor centered on Z axis? (I saw your gallery on flickr, and you had it offcentered from the start). Or should I ask, why Bob decided to put i on side?

That is very interesting FreeRider ... I know what you mean about having the weather control your CNC/garage projects!
If the Z motion when off is an issue, you could use the stepper itself as a brake. When you power down, connect the stepper coils in a short circuit (A-A', B-B') and the resistance to rotation goes WAY up, perhaps enough to lock it in place. A pair of small relays controlled by a system on signal would be enough to make this automatic.
Is there any evidence of Z bounce when changing direction or stopping rapidly?

Hello!! I have only one question, what avoids the Z axe (spindle, cables, etc) from falling down when motors are no longer electrified?? What holds the Z axe???
Please i'm tring to build my own and need to understad this problem.
Thanks!!

[QUOTE=FreeRider;1474572]Hello All!

Its been a while and now that the weather is finally warming up I have been able to spend longer hours in the garage in more comfort. From this I have now upgraded my momus to FULL belt drive. Sourcing ACME rod is somewhat of a burden in my location so I wanted to experiment and see how well a belt drive would hold. I've decided to break this into a new thread as its easier to find via a google search...not many belt driven Z axis machines out there...I've looked.

The first thing I had to do was stiffen my x axis carriage as it had a large HDPE block in the middle under compression...I've since replaced that with solid steel square. There are 2 thread rods that run thru the centre of the square and allow me to compress the entire assembly together very very tightly.



After this I added a 3/16" alum plate to the front of the assembly to hold the new belt drive. Here you can see the general overview of how it comes together. The bottom bearing hole is oval and allows for tension to be placed on the belt after assembly. The black socket cap screws are threaded into the steel bars and hold very firmly.

https://farm4.staticflickr.com/3795/...a28876be_b.jpg