Rack & pinion vs ball screw drive

[CanisLupus]

I've been looking at the CAMaster Stinger II (SR34) and the ShopSabre 23. The CAMaster uses rack & pinion on the X and Y with a ball screw drive (BSD) on the Z. The ShopSabre has BSD on X, Y, and Z. As you can image, both claim their design is the best. Both claim their approach is used on high-end machines and that demonstrates the superiority of their design.

Based on the size of the machine in question, how much does R&P vs BSD really matter. Is one really better than the other?

[rscrawford]

I doubt you could tell any difference between the two. The rack and pinion is a little cheaper to set up, but probably more maintenance over the long run -replace the pinion every few years under heavy use, maybe change the rack every 10 years or so. But for hobby use, probably never need to change either. Ball screws are typically a little slower rapid movements. You will maybe get a slightly better edge finish on harder materials like aluminum with ballscrews. With most cuts, not a big deal.

I think you'd be happy with either machine. Both use WinCNC, which is (in my opinion) the best you can get at that price level. I wouldn't be bothered much about either R&P or ballscrews on that size of a machine. You'd be happy with either.

For what its worth, my Cobra has a single motor on the y-axis and rack and pinion. I've been told that both of those things are 'bad' (all the newer CAMaster machines use two Y motors now). I've never had a problem with it in the 7 years I've been using it, and I pump out some pretty exacting furniture parts.

[TReischl]

As you can image, both claim their design is the best. Both claim their approach is used on high-end machines and that demonstrates the superiority of their design.

Uh huh.

Talk about obfuscation! WOW!

"high end" has nothing to do with it. An example is in order (all from the same company, Mitsubishi):

Their EDM machines are extremely accurate and use ball screws BECAUSE they do not have a 'need for speed' on those machines.

Their laser cutters are rack and pinion. They are also very accurate. IIRC their positioning accuracy is less than .002 inches over 10 feet of travel.

I just betchya their sales folks do not bloviate about one being superior over the other.

Accuracy has zippity doo dah to do with the type of motion system used, ball screw or rack and pinion. It has to do with things like "open loop" vs "closed loop". If a machine is running a closed loop system the motor is not controlling the positioning of the machine, either an encoder or a flat scale is doing that job. When it comes to open loop systems then things change. If a high accuracy ball screw is used then higher accuracy is obtained. But speed suffers unless big honking motors are used to drive the ball screw. A rack and pinion open loop system can be very accurate depending on the gear ratios used and the precision of the rack and pinion.

A lot of this stuff is just talking points for sales folks that is used on the unwary. The bottom line is this: What will they guarantee the positioning accuracy of the machine?

Anyone selling machines in the price ranges of these two companies trying to ride on the coat tails of the big boys by stating they use the same technology is just plain employing sales tactics.

If a sales person said that to me I would reply "Oh, so you are running a closed system on your machine then?"

Simply put, statements like that are similar to a Chevy salesperson saying "We use the same method as Rolls Royce in our cars, an engine, transmission, drive train, wheels, etc".

[Ken Rychlik]

I have had both and my latest machine is Rack and Pinion for x and y. I can't tell any difference in cut quality from one to the other. Racks require more maintenance, but at least maintenance is inexpensive. Keep em lubed and replace pinion gears every few years. I do think ball screws for the Z is important though.

[Leo]

I work with HIGH END machines. They are in the price range in the many hundreds of thousands of dollars and have accuracies in the .0001 range. One of the machines I work with is a REALLY HIGH END grinder capable to millionths of an inch. There are also machines with linear motors that are in the higher end range. I snicker a bit with these conversations.

Trust me - I work with those machines in my line of work but I do not own one at home.

Teds machine has rack and pinion X_Y drives. When I saw his machine I was convinced that it was superior to my small NOT high end hobby machine with ?Ball Screws? I question this because so many machines out there in our price range are not even ball screws, many are actually acme thread drive screws. Yes there are some decent ball screws in the lower end machines that we play with.

I upgraded my 24 x 24 machine that was commercially manufactured to a much larger machine 1300 x 1300 x 254, which is 51" x 51" x 10".

My smaller 24 x 24 x 4 machine was all ball screw drives with NEMA 23 steppers. It was under powered and I needed to upgrade the steppers. I also needed to upgrade the electronics and a bit of other work. I got the machine used, but it was super super low hours, and I got a super super deal on the machine. That machine served me very well and I made a ton of money with it, then I sold it for almost 10 times what I paid for it.

When I upgraded to a larger machine I seriously considered the CNC ROUTER PARTS professional machine with rack and pinion drives. I also looked seriously at the CAMMASTER but it was not in a price range that I could afford.

I knew a Chinese seller and was able to communicate with several people in Canada, Mexico and in the USA that bought machines from my contact and they all gave good reviews. So I bought my 1300x1300x254 machine from China. Since then I bought a 48x96x12 machine (rack and pinion), installed and trained for a client. He is happy with it. I also bought a 6090 100 watt Co2 laser from the same seller.

Like I said in your other thread I thought the Cammaster with the Hiwin rails vs the round rails and with the pack and pinion vs the ball screw is a fantastic machine. I have never hear anyone complain about a Cammaster. I have never heard anyone trade in the Cammaster for another brand.

I do not own a Cammaster and likely never will.

I am quite happy with my chinese machine. I was cutting Poplar with a 1/8 end mill at 375 IPM this past weekend, no problem. I tested it out cutting Walnut at 500 IPM with a 1/2 end mill, Threw big chips across my shop, what a mess, no problem. The Cammaster CAN do this, not so sure about the others.

[TReischl]

I have had both and my latest machine is Rack and Pinion for x and y. I can't tell any difference in cut quality from one to the other. Racks require more maintenance, but at least maintenance is inexpensive. Keep em lubed and replace pinion gears every few years. I do think ball screws for the Z is important though.

Interesting. I agree with you about ballscrews on the Z axis. Z axis are short axis so all the the issues that arise with using ballscrews on a much longer axis do not exist. Since the screw is short it has a low moment of inertia. As a machine gets larger it requires larger ballscrews to move the larger load. Ballscrews have a moment of inertia that the motors need to overcome to turn them. The bigger the ballscrew, the greater the inertia. So the solution is to put a bigger motor on the screw. That adds significant cost. My first machine would stall at about 135 IPM, it was only 24 X 36 and driven by ball screws. I use the same motors on my second machine with a much heavier gantry and it does not stall until about 900 IPM. No, I do not cut at that speed nor do I rapid at that speed. Achieving high feedrates with ballscrews is problematic at best. As Ger21 mentioned in another post, pretty soon the screw has to be tensioned to reduce whipping, a support needs to be added, tensioning the screw means the machine frame has to be some kind of massive to handle the load of tensioning the screw.

I am curious though Ken, I have been running my rack and pinions for about 7 years now and have not had any maintenance issues with them. The pinion is spring loaded into the rack and does not collect any debris. I do not lube them at all. When I started with them I figured the pinions were made of aluminum and very inexpensive to replace, about $10 IIRC so I thought that leaving lube off might be a good idea as many lubes collect dust and become grinding pastes. With an aluminum pinion it is obvious the rack (steel) is not going to be the item that is wearing.

One thing that Leo did not mention in his post was that the machines he works with will all have either an encoder or glass scales on each axis to provide the accuracy he mentions. There is no such thing as a high accuracy machine like he describes running on an open loop depending on the drive mechanism for accuracy. Those machines will also have mapping data to compensate for any slight errors that may be in the feed back loop.

Most of the machines sold these days to guys like us are downright crude when compared to what is sold industrially. But then, most of the tolerances required in woodworking are not even in the same league as what is required in metal working. If one of these "salespersons" tried to sell an open loop cnc milling machine to a machine shop these days the owner would gather up all his employees to help laugh the guy out the door.

If I were buying a machine I would ask two questions first: What is it's positioning accuracy and what is it's maximum feedrate? After that I would have them run the machine at various feedrates while cutting and keep my hand on the machine. (BTW, Leo did exactly that when he visited with my machine and yup, it was vibrating a bit which has since been resolved).

[rscrawford]

The higher end CAMaster machines all have closed loop servos. The Cobra elites have rapid movements of 2500 ipm. Not really recommended to have an ATC machine without some sort of closed loop. Never had a problem with mine, as far as accuracy goes, and I have one of the older, more simple Cobras with only one Y axis motor. When I push on the gantry, I can hear the servo motors constantly making tiny adjustments to ensure the spindle stays exactly where the WinCNC controller 'wants' it to be. That is where the accuracy comes from, the closed loop system.

ShopSabre uses Mitsubishi servo motors with glass encoders on their higher end machines. (I was told the IS series uses 1kW servos with glass encoders on the Y axis). To overcome the high inertia and the whipping problems, their ballscrews stay stationary and the ball nut assembly spins instead. This allows them to have rapid speeds of over 2000ipm.

Both of these machines get their accuracy more from the closed loop system than the rack and pinion or ballscrew setups.

I think what is more important, is how accurately the rails are placed - is the frame milled after its welded to have highly accurate guide rail placement? This will have far more impact on your accuracy than R&P or Ballscrew.

[Ken Rychlik]

TReischl, if you still have original pinions after 7 years, you might need to check em.

[Adrian]

TReischl, if you still have original pinions after 7 years, you might need to check em.

I got nearly 8 years out of my first set of pinions. Depends on how often the machine is used. I used it more in the last two years of the pinions life than I did in the first six. I expect it would have gone on for years more if my usage level had stayed there.

[Leo]

The industrial machines I work with have glass scales.

I posted a youtube video showing the accuracy of my Chinese machine with rack and pinion.

It is well within .0005 on all three axis's, as measured with indicators.

See the video



Leo

[TReischl]

........
I think what is more important, is how accurately the rails are placed - is the frame milled after its welded to have highly accurate guide rail placement? This will have far more impact on your accuracy than R&P or Ballscrew.

Takes a big mill to do that with a 5 X 10 machine.

On a smaller machine we would have the frame stressed relieved and then milled.

What we used to do is use a laser interferometer to get the rails straight and parallel. Time consuming process. Several passes have to be made because as each adjustment along the rail is made the previous one is effected. Sort of like adjusting the nodes on a bezier curve without the handles.

Once the rails were straight we moved on to mapping the screw. Even precision ground high accuracy balls screws have a tolerance that needs to be accounted for.

Like Leo, Russ and a few others have stated, high accuracy comes from using a closed loop system combined with straight and parallel ways.

It is amazing how those glass scales or similar electronic scales make a machine very, very accurate. I purchased a mini mill a year or so ago. Downright cheap lemme tell ya. Also incredibly inaccurate, lots of backlash. I added electronic scales to all three axis. Solved the entire issue. My scales are only accurate to within .001 which is more than I need for the things I make on it.

They would be nice to have on a cnc router, but I am not convinced that they are necessary for doing wood for guys like me.

Going to go see Leo's video!

[ger21]

Takes a big mill to do that with a 5 X 10 machine.

The big Italian routers have machined rail mounting surfaces.
Afaik, they don't use glass scales.
Our Morbidelli uses absolute encoders on the servos, but that's it.

[rscrawford]

[quote="TReischl
Takes a big mill to do that with a 5 X 10 machine.

On a smaller machine we would have the frame stressed relieved and then milled.

[/quote]

I saw this video of ShopSabre milling one of their IS series frames. I was impressed! They stress relieve and then mill frames for their 6x12 machines as well. I'm pretty sure they mill the frames on their lower end Pro series as well, which I believe you can buy a 4x8 for less than $30k. I haven't seen how CAMaster mounts their rails now. I know on the older Cobras like mine, the rail mounts don't appear to be milled. For the wood products I cut, this doesn't affect anything. But I could imagine edge problems if you are cutting aluminum or composite materials.

https://www.youtube.com/watch?v=fMZgR6ojOf8

[TReischl]

Russ,

That is a heckuva nice mill those guys have. Being able to machine all the surfaces with no need to change the setup is a huge advantage on the road to accuracy.

When we were doing that type of work we could only machine the top surfaces, so everything was keyed from that surface. Of course every time something is added to mount something else not on that plane a bit of accuracy is lost.

[Leo]

Back in the mid 90's I visited a couple of machine manufacturers in the Carolinas and they had some Ginormous machines doing that sort of machining. I have never worked with anything that big.

We were buying an Okuma Cadet lathe at the time and they were milled in a similar process.

I also visited Huffman as we were investigating purchasing a cutter grinder at Winchester.

Manufacturing is a fascinating world.