[cootcraig]

I am researching a DIY CNC router build. The two designs that
have caught my eye are the OpenBuilds C-Beam and the Solsylva
10x9. Some thoughts on what I'm interested in:

A small work area and machine size
Lower cost
Machine fabrication that I can accomplish
Produce pieces and design engraving for small wooden boxes.
Aluminum plate milling
Resin cast mold making

C-Beam Machine Specs:
X Axis Travel 350mm - (13.5")
Y Axis Travel 280mm - (11")
Z Axis Travel 160mm - (6") Actual working depth is 1.250"

Solsylva 10x9 has a cutting area of 10 x 9 inches with 4 inches of Z travel,

I would like to use the Smoothieboard controller:

SmoothieBoard

My current choice is the Solsylva 10x9 with Acme leadscrews.

I would like to hear from others to help guide my research and
eventual project choice.

[wizard]

I'm not a big fan of T-Slot extrusions, especially on bigger machines. This mainly due to a couple of things. 1. extruded aluminum is expensive. 2. You have to use an incredible number of gussets and other reinforcements to keep the machine from loosening up over time. You have a smaller machine so it isn't a big deal, but if you are serious about machining metals I'd reconsider.

Also you really want a larger Z axis working area. 1.25" working depth is a little thin if you add a waste board and or a little fixturing.

[cootcraig]

I'm not a big fan of T-Slot extrusions, especially on bigger machines. This mainly due to a couple of things. 1. extruded aluminum is expensive. 2. You have to use an incredible number of gussets and other reinforcements to keep the machine from loosening up over time. You have a smaller machine so it isn't a big deal, but if you are serious about machining metals I'd reconsider.

That makes sense, I was already leaning toward the Solsylva 10x9 plans.

Also you really want a larger Z axis working area. 1.25" working depth is a little thin if you add a waste board and or a little fixturing.

I can see where a larger Z axis travel would be useful.

[jueston]

to put 1.25 inch working depth in reference, i used 1.5 inches just for my spoil board, and then i put 3/4 fixtures on that of that and then the work piece on top of that. i have lots of z-room to waste, so it doesn't matter. but at 1.25 you will be very limited in any fixtures you use to hold your material, unless your sure 100% of what you want to cut is plywood, i would not build anything with such a small z-working area.

you mentioned you want to mill aluminum, you should be prepared that while tools made for cutting wood will almost always work in aluminum, it will be very slow going, and you will go through bits quickly if you don't use coolant, which will then pour down over your machine and make a general mess of everything and if any of that aluminum that your cutting works into your cabling it can vibrate its way right through the insulation.

what type or router/spindle are you thinking about using?

[wizard]

That makes sense, I was already leaning toward the Solsylva 10x9 plans.






I can see where a larger Z axis travel would be useful.

I would most certainly want more Z clearance. Even if your focus is PCB and electronics enclosures, I'd recommend more clearance. Obviously sheet metal and PCBs are thin but you still need clamping and other support. It just gets worst if you expect to machine thicker materials, especially if those materials sit on a spoil board. There is always a flips side though excessive clearance just puts more torque on the Y and Z axises of example.

This issue of clearance is perhaps the most important one you have in my mind. Now we need to keep in mind that this is a small bench top machine so don't over do clearance of the Z saddle. Do realize too that going higher does impact rigidity, that may require refactoring how the Y axis is supported

[cootcraig]

Even if your focus is PCB and electronics enclosures,

I build DIY audio equipment for myself. This is one of my expected uses.

This issue of clearance is perhaps the most important one you have in my mind. Now we need to keep in mind that this is a small bench top machine so don't over do clearance of the Z saddle. Do realize too that going higher does impact rigidity, that may require refactoring how the Y axis is supported

I'm going to order the Solsylva 10x9 plans now. I expect to
have questions about this mod as I study and plan.

[DSpeck]

Have you looked at the microcarve BBox builds? You get about 12"X x 12"Y x 4" Z travel (Z clearance to suit your needs). The build is pretty straightforward, and results are very good with it.

[cootcraig]

Have you looked at the microcarve BBox builds? You get about 12"X x 12"Y x 4" Z travel (Z clearance to suit your needs). The build is pretty straightforward, and results are very good with it.

Yes I have. I'm collecting builds to study here and microcarve is on the list
and I've been looking.

I have the Solsylva 10x9 plans now and will learn a lot from looking them over.

[cootcraig]

what type or router/spindle are you thinking about using?

My current choice is Porter-Cable 690.

[LS120]

i'm building a 2x4 solsylva at the moment. one thing i would do is ditch the solid wood for a good ply of 1" x 6" and cut it down to 4" wide. you will find the solid wood will dish/curve but ply will stay more true. that is the one big thing i did as all my 2 x 6 did this.

[Devastator]

I built a Solsylva. My machine is pictured in their photo gallery at the Solsylva website. I make rocket parts with it. Been real happy with what it can do for the price.

[cootcraig]

The folks on this site are a great resource. My current thought
is to build the David Steele 10x9 machine from the plans with
the hardware store All Thread leadscrews. I plan to use the
smoothieboard controller. With no experience in CNC, a low
cost DIY starter machine will get me going just fine.

[Devastator]

If you spend money anywhere on the machine, I would not skimp on the lead screws. Multistart acme screws will be an affordable worthwhile investment. The lengths you'll need are short enough to not be very expensive at all. I'd even go cheaper on the driver board and steppers over skimping on the lead screws.

[cootcraig]

If you spend money anywhere on the machine, I would not skimp on the lead screws. Multistart acme screws will be an affordable worthwhile investment. The lengths you'll need are short enough to not be very expensive at all. I'd even go cheaper on the driver board and steppers over skimping on the lead screws.

Good point. When I'm closer to a build and buying parts, I will price ACME components from DumpsterCNC and McMaster-Carr.

I'm also thinking that the first frame and hardware store leadscrews might constitute throwaway components
to build experience with cad/cam and gcode generation.

Also, my mechanical skills are such that just following the Solsyva plans as written should help me to
get it done.

[Devastator]

I'm also thinking that the first frame and hardware store leadscrews might constitute throwaway components
to build experience with cad/cam and gcode generation.

Here's my thoughts on that. Everybody starts the build with the idea of upgrading later. In reality it is a PITA. You have to tear a working machine apart add the new components and re-align everything. For some upgrades, it's not too bad, like adding limit switches or better rigidity components, but for leadscrews it is a pain. And the fact that threaded rod is an exercise in frustration.from the start. Sure you can learn on it but I've also seen a lot of people get really disappointed after they make their first parts with threaded rod lead screws. The cost difference of the screws is not that much. The AB nuts will run about the same for each. My Solsylva plans came with a well written section on lead screw selection that made the choice pretty clear.

My first CNC I ever built used threaded rod. Since I've used both, my opinion now is to cut corners anywhere else but the leadscrews, and to be sure to get multistart screws. I really dig the 5-start screws.

[wizard]

Good point. When I'm closer to a build and buying parts, I will price ACME components from DumpsterCNC and McMaster-Carr.

For a machine that you want to do actual machining on your only economical,choices are ACME or ball screws.

I'm also thinking that the first frame and hardware store leadscrews might constitute throwaway components
to build experience with cad/cam and gcode generation.

If you want to focus on learning G-code you don't even need a machine as things like Linux-CNC can run in simulator mode. If you really need to see physical movement belt driven slides can move pens around.

Also, my mechanical skills are such that just following the Solsyva plans as written should help me to
get it done.

Depending upon which plans you settle for your may find yourself in upgrade hell. The concern is the robustness of the machines. While useful for learning they may leave a lot to be desired production wise.

You don't need hardware right away to start to learn CAD/CAM, G-Code and the like. I'd step back a bit and consider just how big of a machine do you really want or need for your intended usage. Build that machine instead. If all you need is a small machine build that robustly.

Unless you have tons of money to throw away it is better to stay off the upgrade path. To an extent anyways, there isn't a machine built that can't be improved. However it makes little sense to improve a machine by throwing out 90% of it during upgrades.

[Devastator]

Depending upon which plans you settle for your may find yourself in upgrade hell. The concern is the robustness of the machines.

Agreed. That is the biggest obstacle to overcome with the Solsylva plans. If you follow the plans exactly, the machine you end up with may not meet your expectations.

[cootcraig]

My first CNC I ever built used threaded rod. Since I've used both, my opinion now is to cut corners anywhere else but the leadscrews, and to be sure to get multistart screws. I really dig the 5-start screws.

This makes sense to me. I will ask specifics about leadscrews when I'm ready to build.

[cootcraig]

If you want to focus on learning G-code you don't even need a machine as things like Linux-CNC can run in simulator mode. If you really need to see physical movement belt driven slides can move pens around.

Good advice, I've started researching G-code generation and picked a simple project of interest.

I will start working with LinuxCNC simulation mode.