Rack and Pinion vs Acme Precision Multi-Start

[roy_okc]

Months-long lurker/student of the build logs, first time poster, seeking discussion on above subject. Also have visited with and received feedback from another member here.

I'm in the design phase of making a Solsylva-inspired 3-axis router (similar to the 25x25 and 24x48). I will likely build a strap-on 4th axis rotary for occasional use. I'll be liberally using 8020 in place of wood for the Y gantry and X upper rails, aluminum plates for the Z assembly; pretty much the only wood will be the bed assembly and that might even change to 8020 as well. I'll be using Rollon TLV-43 and maybe ULV-43 rails for my X and Y bearings and 15mm x ~11.75" TDK-like linear rails for my Z; no skate bearings. My planned cutting size is about 35" X (limited by single length of Rollon), between 16-24" Y (overall machine width narrow enough to fit easily through a typical outside door), and 4-6" Z. I'm planning on using one of the ~380oz-in stepper/G540/power supply kits from CNCrouterparts or Keiling. I'd like to be able to reuse as many mechanical parts as practical if I decide to build a larger machine later. I plan to use my current Hitachi router and/or purchase a Colt trim router, either one using precision collets if/when needed.

My planned projects are entirely wood/MDF and likely plastics, to build kitchen cabinets that will include 2.5D reliefs in the panels as well as possibly cutting some of the cabinet door and drawer parts. Other/later work likely will be toys for grandkids, woodworking jigs, maybe some signs/plaques/lithophanes to make a little side income, etc. I might try cutting aluminum and/or pot metal for my son-in-law's motorsports hobbies if the machine is beefy enough.

I'm still trying to figure out my best option for X and Y mechanical movement parts, R&P or precision Acme multistart. I believe that R&P would be cost about the same as Acme given that I'd probably need two 6' and one 3' sticks of Acme plus DumpsterCNC nuts (although the price for either is not a significant factor for me). R&P would also have more parts moving to a larger machine. I plan on using Acme multistart for my Z. However, I'm concerned about backlash in R&P (due to my lack of knowledge of R&P and unable to find the right info on CNCZone).

I don't care that much about high movement speed, as long as the machine is fast enough to not burn my materials. Of course I'd like to optimize my time as long as it doesn't negatively affect the spindle, cut quality, etc.

I'm close to starting a wooden rolling mockup to help me determine final sizes, designs, etc. Probably a month or two away from starting the final machine. Hope to start running and fine tuning by the end of the year.

So, my questions are:

1. How much backlash is in CNCrouterparts R&P compared to say Acme 1/2-10 5-start w/DumpsterCNC?

2. What is a suggested maximum acceptable amount of backlash before 2.5D carvings start looking bad? I understand and am fine with final cuts using very fine overlaps and small diameter bits, requiring long cut times. However, at what point does backlash negate overlap?

3. Am I correct in assuming that R&P should be on both sides of X to prevent racking?

4. What are good machine resolution and backlash goals for 2.5D wood/plastic?

Thanks in advance,
Roy

[Dorsal]

I don't have any answers for you, but you should read the Joe's-2006 CNC thread and the one about the Joe's Hybrid. Maybe someone there has posted some comparative numbers.
The new design started with screw drives, but many of the guys have changed over to R&P. Part of that decision is based on the size of the screws - they need to be about 5 feet long (plus couplers and flange bearings) and the possibility / probability / reality of a L O N G shaft wobbling and whipping at any higher rpm's is one a person wants to avoid. In fact, Joe has incorporated the R&P option as a part of the design drawing now.

[arizonavideo]

What was the price on the Rollon TLV-43 rails? They look like a nice product but I have not seen anyone use them before?

[jsheerin]

I don't know the answer to your first question from first hand experience (no R&P), but the cncrouter parts r&P (like many others) uses a spring to load the pinion against the rack. This reduces backlash. It should get it a very small level, but when a force is applied to it (like when you're cutting something), the backlash will probably increase slightly. I would imagine that is all low enough that it will not be an issue for woodworking though. I have 1/2-10 5 start screws with dumpster nuts and get a few thousandths of backlash, like 1-3. Similarly, I'd expect that to increase a bit under load as the screw and plastic nut flexes.

Once again, not exactly sure on the amount of backlash that would cause problems, but with the backlash I have I have not seen any problems.

There's no reason you couldn't put a R&P in the center of an axis the same way that people put a single screw under the center of an x axis. I'm pretty sure I remember someone here building a large moving table that way. I'd generally avoid center drive on larger machines, but on a 16-24" wide y axis, a single x drive would probably be fine.

I'd shoot for resolution around 0.005"/ full step or smaller and backlash in the same range or lower. That's because you can achieve that without too many problems or excess expense. If you can get lower, it just means you have to worry about it even less, but I wouldn't go to great lengths to get below that.

[roy_okc]

Originally Posted by arizonavideo What was the price on the Rollon TLV-43 rails? They look like a nice product but I have not seen anyone use them before?

ArizonaVideo: $25 ea + S+H for used rails at http://cgi.ebay.com/ws/eBayISAPI.dll...=STRK:MEWAX:IT.

I've bought a total of 6 ULV/TLV rails/trucks from this seller. Although advertised as stiff, they seem to be in good shape (one loose bearing out of 30). However, a truck or two is adjusted too tight. Each bearing is on an eccentric axle that is adjustable.

I'll likely machine new trucks from wider aluminum stock, at least for my Y trucks.

Roy

[roy_okc]

Originally Posted by jsheerin I'd shoot for resolution around 0.005"/ full step or smaller and backlash in the same range or lower. That's because you can achieve that without too many problems or excess expense. If you can get lower, it just means you have to worry about it even less, but I wouldn't go to great lengths to get below that.

Jsheerin,

Thanks for that information. That's part of what I was looking for. I see people trying to get to 1 to 2 thousandths, but I wasn't sure if that tightness is really necessary for my needs.

Roy

[roy_okc]

Originally Posted by Dorsal I don't have any answers for you, but you should read the Joe's-2006 CNC thread and the one about the Joe's Hybrid. Maybe someone there has posted some comparative numbers. The new design started with screw drives, but many of the guys have changed over to R&P. Part of that decision is based on the size of the screws - they need to be about 5 feet long (plus couplers and flange bearings) and the possibility / probability / reality of a L O N G shaft wobbling and whipping at any higher rpm's is one a person wants to avoid. In fact, Joe has incorporated the R&P option as a part of the design drawing now.

Dorsal,

Thanks for the suggestion. My earlier desire was to make a router that could cut 24x48; however, the overall size would be a bit much for my workshop, possibly transporting it occasionally, etc. I may still build one that size later on, thus trying to future proof as many components as I possibly can (and having seen many comments about whip, is what is leading me to investigate R&P now).

Thanks,
Roy

[jsheerin]

I don't know if .001"-.002" is necessary, but it's nice. However I don't really think you'll have trouble getting there using either one of these approaches.

[ahren]

I have many customers using our R&P system, and most routinely get below 0.002" of backlash. The 5 start ACME systems routinely get down to 0.001", so it really is about the size of your axis. Above 4', I definitely recommend R&P to avoid whip and get better speeds. Less than 3', ACME is typically more economical.

Ahren
www.cncrouterparts.com

[Matt Clara]

Originally Posted by ahren I have many customers using our R&P system, and most routinely get below 0.002" of backlash. The 5 start ACME systems routinely get down to 0.001", so it really is about the size of your axis. Above 4', I definitely recommend R&P to avoid whip and get better speeds. Less than 3', ACME is typically more economical. Ahren www.cncrouterparts.com

I notice the new Camaster Stinger 25" X 36" X 5" has R&P on all three axis. What's your take on that, as in, why would they do it?

[ahren]

Matt,

I'm personally not a big fan of R&P on the Z, as typically this sort of drive can be easily backdriven with the power off, so your z will fall. In addition, the short travel doesn't really justify all of the components needed for an R&P system in my mind.

I haven't personally examined their machine, and it's a little hard to tell what's going on from the website, but maybe they have a different implementation with an electronic brake or some other means of overcoming this problem.

Ahren
www.cncrouterparts.com

[roy_okc]

Originally Posted by ahren I have many customers using our R&P system, and most routinely get below 0.002" of backlash. The 5 start ACME systems routinely get down to 0.001", so it really is about the size of your axis. Above 4', I definitely recommend R&P to avoid whip and get better speeds. Less than 3', ACME is typically more economical. Ahren www.cncrouterparts.com

Thanks Ahren,

That's exactly the information that I was looking for. I'll almost assuredly put R&P on my X axis, probably dual to prevent racking. Will re-think Y; no plans (room) to go ever go much beyond 24" cut width.

Thanks,
Roy