[OvaR8ed]

...so I've been a little inspired the last couple of months (more like possessed!) and I’m realizing that this thing has got a proper hold on me and I’m not stopping any time soon.

Guess all that’s left for me to do is start a build thread…

[CarveOne]

Been there, done that a few times.

[OvaR8ed]

I started on the electronic side of things, since it's small and easy to work with...

I'm planning on building a micro-controller board for each machine, but since I'm starting up it might be nice to have an off the shelf Arduino Uno for debugging purposes... check.

Next on the list was a Stepper motor driver. There was no way I was buying one and it had to be done from scratch. I decided that the L297-L298 combination would be a nice place to start so I got the datasheet and planned a layout on VERO Board.



I did not add the Current Sensing part of the circuit since this was only a prototyping exercise and I wasn't going to drive high current stepper with it...

[OvaR8ed]

By the way, I did a whole lot of testing to figure out the wiring of the motor and eventually converted it from a 5 lead Unipolar config to a Bipolar config by cutting a track on the external PCB.



I eventually got it working, but the vero layout was not very efficient when it comes to heat and research was underway for a proper PCB layout.

[OvaR8ed]

I found a layout marked P.Rouzeau 2/2003 which I believe to be very efficient so I stocked up on some pool acid and peroxide to start etching. By the end of that day the little stepper I salvaged from an old printer was purring like a kitten.

[OvaR8ed]

Here's a video of the test:



The chip handled a couple of minutes of testing no problem, but I realized I'll need a larger heat-sink if I want to run a motor for long hours...

[OvaR8ed]

With the electronics working I started researching on the mechanical side.

Among hundreds of designs out there I realised that the JGRO and Joe's 2006 designs are two good benchmarks to base design on. These two models are at two different levels completely, both in size and complexity.

I noticed a lot of people are basing their builds on these two models and most people who takes on the JGRO talks about using it to build parts for an upgrade to Joe's 2006 CNC.

I'm not taking on Joe's machine as a first project, so I was looking at something similar to the JGRO. I read a few build threads about the JGRO and decided I'm not going for the floating pipes to guide the axis due to quite a few people mentioning flex, but rather try to support the guides like in Joe's model.

I like how easy it is to calibrate alignment in the JGRO, but I feel that if the MDF chassis is cut precise enough I should be able to shim the alignment into calibration during assembly without using this concept. This will greatly reduce the amount of parts needed and design complexity.

I also decided that in the interest of cost, to rather use the same angle section that hold the bearings for the guides. Again, simplifying design and less leftover material.

[OvaR8ed]

I downloaded plans and sketchup models for the JGRO to get an idea of scale etc. and noticed that the MDF used is 19mm thick, but here in South Africa our nearest standard thickness is 18mm... What's up with that?

Anyways, I realized that in order to avoid little 1mm misalignments jumping my design from all sides, the entire thing would have to be re-engineered. This was the final push I needed to just start my design from scratch.

I've gone back and forth between a few design ideas, sometimes hitting "select all" and "delete" after several hours of work, but I finally got a partial design that I'm happy with and now I can start working outward from here.

The Z-Axis is similar to the JGRO (except for the guides), but I made the Y-Axis carrier wrap around the beam like a smaller version of Joe's design for rigidity.

I coloured the different MDF parts slightly different for illustration purposes.

[OvaR8ed]

[kuagchun]

Hopefully someone will get it pretty soon.

[jckstrthmghty]

I'm not a fan of the z rails being pressured outwards. It's the same issue the JGRO has with it's unsupported rails and deflection. The y rail in your drawing is the same way buildyourcnc does it and for short axis's I believe this is the better method. It takes advantage of the compression strength of MDF and allows for adjustment screws on the z-bearing. As long as you cut your rail support square your axis will be. The issue is where does the leadscrew go? Makes for a bulky carriage. In the end I think it's best to go with materials designed for the task. Solid hardened steel rod just like how joes2006 does it. For a bit more $$ microcarve has excellent complete bolt on z-axis assemblies. This is the choice I'm going with.

I have a jgro and I'm using it to build a joes2006. If you have any question on either I would be happy to share my experiences.

[Lee DeRaud]

I downloaded plans and sketchup models for the JGRO to get an idea of scale etc. and noticed that the MDF used is 19mm thick, but here in South Africa our nearest standard thickness is 18mm... What's up with that?

19mm is 3/4", which used to be the standard thickness here in the US. Last couple years, most of the '3/4" MDF' I've found here is actually 18mm anyway.

[OvaR8ed]

I finally decided on a router... I got some rough dimensions from the Bosch GMR 1 Laminate Trimmer to base the height of the gantry on and now design work is back in gear.

The clamps on the Z-Carriage may change slightly to suite the router, but I just added a provisional design in the mean time.

The Lead Screw for the Z-Carriage will run down the middle of the guide system with the nut between the two vertical pieces of MDF. The Lead Screw for the Y-Carriage will run above the rail support beam with the nut in that little housing that is visible inside the Y-Carriage.

The gantry will also run with bearings on steel angle. The guide system will be designed on the inside of the gantry frame, like Joe's design, not outside like the JGRO. (I hope this makes sense...?)