Sidewinder CNC 2

[atarijeff]

After finishing a CNC build that someone else started but gave up on, I decided to build one from scratch from a much better design. I liked the Sidewinder CNC design from Dave G. to use as a base model. Right away though I could see there were some tweaks that needed to be made. Here is my version:



Changes Made:

• (Not really a change) Fully modeled in Fusion360.
• New motor mounts with thrust bearings, shaft collar, and 3D printed bearing retaining plate to isolate the motor from the lead screw.
• Commonized the motor mount plate and bearing plate. Now there is just a left and right part.
• 3D printed bearing retaining plate on free end of lead screws.
• Open the top and bottom gantry plates to shave some weight and give access to dropped fasteners, etc.
• Extended bed to give full 24" x 48" cutting area
• All fasteners countersunk where possible. I think this give a better looking finished product plus accommodates the use of flat and locking washers in the assembly.
• Changed some fasteners to use T-Nuts.
• All fasteners are button head cap screws with locking and flat washers.

Still TBD:

• I think I may also change the lead screws to a 1/2-8 4 start sold by Roton.com plus their anti-backlash nuts (since McMaster REFUSES to ship to Canada, grrrr)
• Aluminum vs steel angle to use as the rails. Thoughts?
• The side force on the plates that hold the anti-backlash nuts and drive the moveable sections is iffy at best. I may need to add gussets or something similar.

I'd like your feedback on the design, especially if you've already built a Sidewinder and have ideas for improvements.

Similar Threads:

• Sidewinder CNC
• New Machine Build- Sidewinder CNC
• Sidewinder CNC Blues... Dave, where are you?
• New Machine Build- Sidewinder CNC

[wendtmk]

You might want to check with McMaster-Carr again. I've heard they've resumed shipping to Canada.

Mark

[atarijeff]

You might want to check with McMaster-Carr again. I've heard they've resumed shipping to Canada.

Mark

Well it was worth a shot. This really burns my bacon

Due to the complexity of U.S. export regulations, McMaster-Carr accepts international orders only from businesses. This decision also applies to orders shipping within the United States, because it is based on the final destination of the items. We are unable to provide a quotation or accept your orders.

[ger21]

After finishing a CNC build that someone else started but gave up on, I decided to build one from scratch from a much better design.

Wow, I'd hate to see the first one.

The Sidewinder is an easy to build, entry level machine. It's not fast, and not very rigid.

What is your intended use for this machine? With low end machines like this, the users expectations are the biggest factor on how good the builder thinks the machine is.


As for your proposed upgrades.

If you go with the roton screws, you'll likely need larger motors and better drives.

Removing weight will often weaken the machine. Weight is your friend.

Even steel angle is not the best option for V rollers. It constantly wears, creating high and low spots and inconsistent forces.

[atarijeff]

For drives it will be a Gecko G540. The motors are TBD but run off of 48V. I intend to use it for 3/4" ply/MDF, lexan, drag knife foam board or balsa, thin CF ~3mm, and maybe light gauge aluminum under 1/8".

[wizard]

If you already have an entry level machine why would you make another one if you aren't satisfied with the first?

[ger21]

I think you'll be disappointed in the results you'll get from that machine.

[atarijeff]

The one I've cobbled together right now is barely an entry level I'd say. Pretty flimsy but probably good enough to build a new one.
So, what would be a recommended design to use?

[wizard]

The one I've cobbled together right now is barely an entry level I'd say. Pretty flimsy but probably good enough to build a new one.

It would help to step back and come up with a list of things that leaves you dissatisfied with the current machine. At best people would be guessing as to what your issues are but a quick guess is that most of them revolve around machine rigidity.

So, what would be a recommended design to use?

At this point it would be guessing as we don't know what your expectations are. However I might point out a few things, avoid any design with unsupported rails for one.

Second make sure the design properly supports those linear rails.

Third do zero in on the materials you expect to machine, if it involves metals or even some composites I would strongly suggest an all metal design. The reason for an all metal design is pretty simple coolants required to do a reasonable amount of metal machining negatively impact wood machines. If your focus is primarily wood then a wood framed machine can be fine.

In a nut shell you can't pick a machine design until you know what you will be doing with it.

Comments on your design changes:
1. Don't use button cap screws unless you have too. There a strong tendency for the sockets to strip out.
2. This is a machine not a piece of furniture. Don't go countersinking stuff just for the hell of it.
3. I'm not down with 3D printing some of the parts, at least not plastic prints. Bearing retaining plates would be better done in aluminum for example. I may be biased here but I'd think carefully about each plastic part, I can see it working for some things but not others.
4. You would not want to open the top and bottom ganty plates. This highlights to me that you don't understand the importance of rigidity in improving a machine. It doesn't matter if your gantry beam is wood, aluminum, steel or unobtanium you really want a box structure for strength and simple design.

[ger21]

So, what would be a recommended design to use?

I would look through all the build logs here, and pick out all the best parts of all of them. Most of them are far superior to the Sidewinder.

If your looking for plans, I don't know any that I'd really recommend these days. Most plans are really aimed at low quality, entry level machines. Joe's 4x4 is a popular machine, but most users end up making a lot of modifications after it's done, to improve it's shortcomings.

Are you trying to do this really cheap? You're planning on spending quite a bit on drivetrain and electronics upgrades. These will make the machine faster, but won't do anything for it's biggest problems, which is a lack of rigidity.
A good machine only really needs a few basic items.

A solid frame, good linear bearings, and a rigid gantry.
Buy the best linear bearings you can afford, as they are the #1 most important factor in building a quality machine.
Then build a really strong box shaped gantry, and don't worry about how much it weighs.

[atarijeff]

This will be a budget conscience build, especially with the Canadian dollar in the tank right now effectively doubling the price of most parts to my door after taxes and delivery.

Design Goals:

• Primarily constructed of 3/4" Maple ply. Bolted and/or glued. Boxed construction.
• Cutting area of 24"x48".
• Aimed at milling soft materials (wood, plastic, thin gauge aluminum). Composites such as CF would be a bonus but I don't know enough about cutting/cooling requirements at this time.
• Air cooled. The router I am using (Rigid kit R29303 2HP 10K-23K RPM) vents through the bottom but I could add more airflow if needed.
• Lead screws will be ACME precision or ball screw. Direct drive with proper anti-backlash mechanisms.
• Control system will be Gecko G540. Stepper specs TBD but driven at 48V.
• Parts available through local vendors, eBay, international friendly.

Linear Drive Decisions (Economy or DIY):

V-Groove Bearings and Aluminum/Steel Angle
Pros

• Easy to acquire parts.
• Easy to assemble.
• Likely to give good performance initially.

Cons

• Bearings are pricey.
• Performance is likely to degrade quickly.
• Even with proper hardened and ground rails I've come across some complaints of too much play side to side in the bearings. Adding compensation adds complexity.

Roller Bearings on Pipe/EMT Conduit (Solsylva style)
Pros

• Easy to acquire parts.
• More complex to assemble but still not too hard.
• Captures bearing/block to rail so should give very minimal play.

Cons

• Accuracy and straightness of pipe/EMT is questionable.
• Seems likely to wear quickly.
• Using threaded rod to support rail seems like it would introduce flex and/or alignment issues.

Linear Rail and Ballscrew Package (from China)

Sample eBay Link -> 3 Set SBR16 Linear Rail ballscrew RM1605 350 900 1200mm BK BF12 End Bearing CNC | eBay



Pros

• Quality and precision.
• Includes most of what you would need for drive mechanism.
• Not in the stratosphere for cost.

Cons

• Quality? and precision? although 100% feedback rating for a Chinese supplier is pretty rare.
• Most expensive of the DIY options.

At this point I'm favoring the eBay linear rail and ballscrew package but it's not final as of yet.

[ger21]

If I had to build something decent for a low price, I'd go the skate bearings on conduit for the X and Y, and pick up some used THK's or Hiwins for the Z axis.

Not Solsylva style, though. I think my X and Y axis are still the best example of bearings on pipe that I've seen, and I designed it 13 years ago. http://www.cnczone.com/forums/cnc-wo...st-router.html

Joe's 2006 copied it a little, but I think that the larger conduit works better.

Mine has been running for almost 7 years now, and after a little initial wear, it really doesn't wear much after there are "flats" worn into the pipe.
There are 3 important things.

All components must be torsion boxes for stiffness and rigidity.
The bearings must be mounted firmly with as little flex as possible. I used 1/4" aluminum angle, which is far stiffer than 1/8", and only a few $ more.
The bearings must be held tightly to the conduit. I used threaded rod to pull them tightly.

The Z axis on my machine is the weakest link. You can probably pick up some used profile rails for $100-$150, which would make a far superior Z axis. I'd probably try to find some aluminum plates, to make an easy and rigid Z.

[cyclestart]

Yeah, it's tough being a Canuck. in this hobby.
China is generally your best bet for fair shipping policies, often better than what's available within our own country.
Also many (most ?) Chinese vendors will provide those rails and screws to your drawings for petty change. Try getting that done locally
while keeping all your limbs.

The Kronos Robotics machines look pretty sweet though not super-cheap to build.
Kronos Robotics

And if your shop is poorly equipped all of this becomes much harder.

[plcamp]

D. Gatton's design is a simple and straight forward machine suitable for the hobbyist and light use. I'm not sure if I would attempt to cut aluminum on it but for wood it is adequate. I was in your shoes a while back trying to decide on a build direction. Long story short I designed my own machine based upon Gatton's and CNC Router Parts machine. Using extrusions was simply to expensive for use as a hobby machine, $1,498 to be exact so I opted for 3/4" Baltic birch. The cost of my wooden machine complete less software was less than what it would have cost me just for the 8020 extrusion.

My wooden machine has proved very reliable, it is fairly fast yielding excellent results at 100ipm and after some initial breakin on the angle iron rails the machine is tight and rigid.

http://www.cnczone.com/forums/diy-cn...ghlight=plcamp

[wizard]

This will be a budget conscience build, especially with the Canadian dollar in the tank right now effectively doubling the price of most parts to my door after taxes and delivery.

Design Goals:

• Primarily constructed of 3/4" Maple ply. Bolted and/or glued. Boxed construction.
• Cutting area of 24"x48".
• Aimed at milling soft materials (wood, plastic, thin gauge aluminum). Composites such as CF would be a bonus but I don't know enough about cutting/cooling requirements at this time.
• Air cooled. The router I am using (Rigid kit R29303 2HP 10K-23K RPM) vents through the bottom but I could add more airflow if needed.
• Lead screws will be ACME precision or ball screw. Direct drive with proper anti-backlash mechanisms.
• Control system will be Gecko G540. Stepper specs TBD but driven at 48V.
• Parts available through local vendors, eBay, international friendly.

First there is nothing wrong with wood construction for a machine that is intended to machine wood or other materials commonly done dry. It is when you talk about aluminum that I start to worry because if you want to do more than a trivial amount you really want to be using a coolant and lube. Fluids and wood don't mix well.

So what I'm saying is that if you go the wood route don't expect to be doing a lot of aluminum. Especially aluminum so that have a tendency to weld to the cutters.

24x48 is very doable in an all wood design just make sure you create the components out of box structures. Your goal is strong twist resistant components. One approach here is torsion boxes. However you might want to consider engineered lumber, for some components, which is an idea that has been floating through my head lately. I'm thinking PSL lumber here,

Don't assume a voltage for your steppers. In an ideal world power supplies would be selected based on the steppers inductance. The Gecko web sight has a document that can help you understand power supply selection.

Linear Drive Decisions (Economy or DIY):

V-Groove Bearings and Aluminum/Steel Angle
Pros

• Easy to acquire parts.
• Easy to assemble.
• Likely to give good performance initially.

Cons

• Bearings are pricey.
• Performance is likely to degrade quickly.
• Even with proper hardened and ground rails I've come across some complaints of too much play side to side in the bearings. Adding compensation adds complexity.

Roller Bearings on Pipe/EMT Conduit (Solsylva style)
Pros

• Easy to acquire parts.
• More complex to assemble but still not too hard.
• Captures bearing/block to rail so should give very minimal play.

Cons

• Accuracy and straightness of pipe/EMT is questionable.
• Seems likely to wear quickly.
• Using threaded rod to support rail seems like it would introduce flex and/or alignment issues.

I'd certainly look into the cost of supported round rails.

Linear Rail and Ballscrew Package (from China)

Sample eBay Link -> 3 Set SBR16 Linear Rail ballscrew RM1605 350 900 1200mm BK BF12 End Bearing CNC | eBay



Pros

• Quality and precision.
• Includes most of what you would need for drive mechanism.
• Not in the stratosphere for cost.

Cons

• Quality? and precision? although 100% feedback rating for a Chinese supplier is pretty rare.
• Most expensive of the DIY options.

At this point I'm favoring the eBay linear rail and ballscrew package but it's not final as of yet.

Ball screws are an interesting discussion, sometimes you are better off with acme in dusty environments.

[poorman56]

Do you have the plans? i would love to have them.