Yet another welding/machining/misalignment problem on Saturn 2. Sigh.

[GME]

We still have some bearing noise on the slave side and will replace those bearing cars soon.

Is is safe to assume that you attribute the bearing car problem to the rack & rail misalignment? It's sometimes hard to accurately diagnose root causes of a failure, but in this case, there appears to be a direct link. However, I suppose it's possible that the bearing cars we defective. I wouldn't put my money on it being a manufacturing defect.

In other threads, there have been been many comments about things like excess wear from misalignment. Your machine appears to provide a useful practical example.

Are you concerned that the rails on the slave side may also be damaged? I don't know whether the rails are hardened. Some advertise that they are. Others don't say one way or the other, which probably means they aren't. If the rails on the Saturn 2 are not hardened, I'd guess there may also be an issue with the rails.

Dan,

If you haven't already done so, it may be worth going through the precise measurement exercise. IMHO, better to address potential problems now, rather than incurring extra expense having to address them when things start failing. Just my 2 cents.

Gary

[he1957]

Are these variations caused by the box section steel frame or something else?

Thanks,

[micknm]

Are these variations caused by the box section steel frame or something else?

Thanks,

It's a combination of errors, IMO. I believe the errors in the rack and rail mount alignment are due to the lack of a machined datum in the steel frame. As I said before, that practice is an industry standard anywhere there's a rack or rail or ballscrew or other precision component mounted. It is near impossible to get alignment tolerances in a welded frame anywhere near what FLA quotes on their website without machined datums. No doubt the cost of doing so is a factor. Again, the FLA machines are SO close to being great that it should at least be offered as an option, and at a minimum be mentioned on the website. A steel frame should be welded, stress relieved (normalized) then machined. Not doing so will require shimming, not just when the machine is initially assembled, but down the road once time has done the stress relieving. I've begun to think that time is the factor here. The frame is gradually stress relieving itself and as a result, changing shape.
Another factor is how the bolt holes for the rack and rail are machined once the frame has been painted. Too many comments on this forum indicate to me that it's not a precision controlled process. The table holes certainly aren't. There are some really big (visibly off) variations in those, which make it a pain to mount, but in the end don't contribute to the tolerances of the machine. They do indicate to me a somewhat careless approach to the build in general, though.
In a post about a year ago I went through a detailed approach to squaring the gantry on the Saturn 2 when we initially set up our machine. It was out by almost 1/4". Maybe it got knocked around in shipping, but more likely the build methods used did little to ensure it was close. Some of the mounting holes for our gantry plates are double drilled side by side because the first hole was off by 3/16" or so. The plate welded to the gantry itself is visually out of square to the connector plate that the bearing cars are bolted to.
The whole process reminds me of my earliest days as a woodworker. I look at some of the pieces I made that I was so proud of at the time (they look nice) and cringe when I see the now obvious errors I made. I'm a much better woodworker now not because I have better, more accurate equipment, but because I started paying attention to how things should be done to guarantee better alignment, precision and better fit. It does not take me any more time to make something right - just the opposite - it just takes better practices. I said earlier that I'm at fault for assuming that good practices were followed in the construction of these machines when I decided that this was the machine for us. For example, I was impressed that Nate used 25mm rails on a machine in this price range. But putting $40 Blumotion slides on a drawer that's out of square does not make it any more square. It simply makes the out of square drawer open and close a little more smoothly.
My whole purpose for bringing this to the forum is not so much a rant for ranting sake. We knew and discussed at length the pros and cons of this purchase beforehand. It took over a month just to get a quote from Nate. I posted because this could be a great bang for the buck with just a few improvements. If I were buying a Chinese machine sight unseen I would know going in that making these adjustments, corrections and modifications might be likely. I didn't expect it to be the case here.

[wmgeorge]

I don't know about you but we have had some extreme temperature changes over the past few weeks. Do you keep your shop at a constant temperature or like me its only 70 when I am working. If you have a slab on grade perhaps its changed, I would check your machine for level. Seems odd that your machine was fine after you set it up a few months ago and now its off?

CNC RP does not do any machining on their frames, but it looks to be all readily adjustable.

[micknm]

I don't know about you but we have had some extreme temperature changes over the past few weeks. Do you keep your shop at a constant temperature or like me its only 70 when I am working. If you have a slab on grade perhaps its changed, I would check your machine for level. Seems odd that your machine was fine after you set it up a few months ago and now its off?

CNC RP does not do any machining on their frames, but it looks to be all readily adjustable.

Yes, it's climate controlled. Machine is level. We set it up over a year ago and this is the first time we've had these issues. Stress relieving (or more precisely, failure to stress relieve) could cause distortion in the frame. When steel is welded it distorts because you're heating one area and not the one next to it. Over time it "relaxes". I believe that may be contributing.

[wmgeorge]

Mick as I see it, IF there were some adjustment built into the design it could all be corrected. But this linear rail mounting depends on the top tubing being flat, square and on the same plane. Whereas other designs mount the linear rail on the side of the square tube. I was surprised you are able to get enough adjustment out of yours to make it right.

Do me a favor and look at the Camaster or other welded steel frames you will see what I mean. No milling required, and as Gary found out its not a practical solution.

My flat plate with the slotted adjustment and linear rail mounted on top is one way to make it work on the ones already out there.


Lot of detail here>

[micknm]

Mick as I see it, IF there were some adjustment built into the design it could all be corrected. But this linear rail mounting depends on the top tubing being flat, square and on the same plane. Whereas other designs mount the linear rail on the side of the square tube. I was surprised you are able to get enough adjustment out of yours to make it right.

Do me a favor and look at the Camaster or other welded steel frames you will see what I mean. No milling required, and as Gary found out its not a practical solution.

My flat plate with the slotted adjustment and linear rail mounted on top is one way to make it work on the ones already out there.

I'm not sure what difference, if any, it makes whether the rail is top or side mounted. Both work fine if the tube (datum) is straight.

If I bought a CAMaster and it was out of quoted tolerances as much as our machine was I would post the same complaints on their forum. Nate quotes .001" repeatability - CAMaster quotes .001". They've obviously figured out a workaround if their frames are not machined. Maybe some of the additional cost goes to shimming and QC in addition to the obvious turnkey items. For what it's worth, I'll almost certainly recommend a CAMaster when the school where I teach pops for a router. I've heard very few complaints.

On the upside, our efforts appear to have paid off. The machine is running much, much more smoothly and after running a few calibration routines is cutting closer to quoted specs than ever. I'll put a detailed description of our process together and post it as soon as soon as I get a chance.

[wmgeorge]

I’m glad things are improved.

The reason I posted the video was so you could see the adjustablity built into the Camaster. Notice the linear rail is under the table. It can be raised or lowered, as required No machining needed. Both the rail and gear rack can be adjusted.

The Camaster Forum is pretty tightly controlled. Make a comment, a not nice one and see how long it stays up. Unlike CNC Zone where non FLA owners can drop in.

[micknm]

Bill, you must have linked to a different video. The one posted is a guy sitting on his router table talking about how much he likes his router for 10 minutes.

The problems with our machine were twofold. The rack was curved along the Y axis on both sides.The rails were also curved along the Y axis. Surprisingly, neither was measurably out vertically, so raising or lowering the rails would not have helped.

[wmgeorge]

The point I am trying to make, If you lookedat the construction of the Camaster it allows for adjustments to be made Whereas the FLA does not. So if the CM was not welded prefect, it could be corrected. 20 years from now if it’s moved, the CM can be adjusted,

[micknm]

Yes, I get it. I'm not trying to argue with you, Bill. The point I was trying to make was that Nate should build his machines to the tolerances he claims. He doesn't.
Before we bought this machine I talked to Nate at length. After learning of my decades of experience in industrial CNC routers he indicated to me that he would welcome my input on what I thought he could do differently. That's what I'm trying to do.

[wmgeorge]

Yes, I get it. I'm not trying to argue with you, Bill. The point I was trying to make was that Nate should build his machines to the tolerances he claims. He doesn't.
Before we bought this machine I talked to Nate at length. After learning of my decades of experience in industrial CNC routers he indicated to me that he would welcome my input on what I thought he could do differently. That's what I'm trying to do.

Great now maybe something will change, my understanding they are now milling the steel on the gantry where it mounts to the bearing plate. How much adjustment were you able to get on the rails? Gary GME said his calculations showed there was not enough on the rails mounting holes to make it feasible.

[ger21]

If you looked at the construction of the Camaster it allows for adjustments to be made Whereas the FLA does not

Do you have any pictures or other info on how they can be adjusted? If they are bolted to the steel rails then they would be no more adjustable than the FLA machines?

I've never heard of anyone having to adjust linear rails or gear racks on a Camaster, and for $20K-$50K, you better not ever need to.

[ger21]

There's a machine build here that's local to me where a welded machine frame was stress relieved and machined, for I think under $400.
If Nate is serious about building a quality machine, he should be able to do this for maybe ~$500 or so.
As long as he is bolting rails and rack to an unmachined frame, the quality will always be suspect.

[micknm]

The amount of rail adjustment is limited to the difference in the rail bolt hole diameter and the bolt diameter without redrilling and tapping the bolt holes. We never put an indicator to it. We simply used a spreader bar and clamps to make them as parallel as we could and it worked (for us).

The Saturn 2 frame is rigid enough that IF it were cut, welded, stress relieved, machined and finally, painted it should hold tolerances (assuming it's leveled properly) withought ever having to make rack or rail adjustments. Those two missing italicized steps are critical to reaching the .001" repeatability and .005" accuracy that Nate claims.

[micknm]

100% in agreement. I worked at the manufacturing facility for several years when I was with MultiCam. Early on they subcontracted to a local place to have their frames and gantries stress received and machined. Then they bought their own mill and were able to do it in house for anything up to around a 80" x 144" frame. Anything over, they sent out. Then they bought another mill with a 10' x 30' capacity. Anything longer than 30' they manufacture in sections.

The only machine I remember ever having to make any adjustment to the racks or rails on was a machine we used while running tests for an aircraft window manufacturer. They cast their own acrylic and needed to flatten it prior to polishing on a separate machine. Our machine flattened it to well within specs, but there were periodic parallel ripples visible across the width, so slight that you could only see them in the right light. We determined that the torque on the rack bolts needed to be adjusted so that there was less flex. Problem went away.

[wmgeorge]

The MultiCam is just like the Camaster as to what I can see, Rails are NOT mounted on top but under or on the side where they can be adjusted IF needed. Rails on top are subject to the accuracy of the base they are attached to and need to be shimmed up if needed, and holes elongated for parallel adjustment.

Gerry what machines do you have at work that are not adjustable? When you mount the rails on the side, bolts and captive nuts, tap up or down as required. No reason why the whole rail and gear rack assy can not me mounted as one piece and bolted to the frame.
It is my understanding the Camaster is delivered and setup, meaning adjusted if needed for the customer.... not the table top machines of course!

How can you weld a frame to hold milling tolerances without some means of correction? The square tube metal on this machine at 10 gauge or so does not have much material left for milling.

I am going to guess that FLA frames may be Jig welded but that's it.

A little googling shows Camaster is not without needed adjustment or correction.
https://www.camheads.org/showthread.php?t=177

[ger21]

Rails on top are subject to the accuracy of the base they are attached to and need to be shimmed up if needed, and holes elongated for parallel adjustment.

1) Rails should never need to be shimmed. If they do, the surface is not suitable for mounting them.

2) If the holes are drilled absolutely perfect, you only have 0.5mm of adjustment. Some of these Saturns are out of parallel more than that.



We have a $150,000 Morbidelli, and all rail mounting surfaces are machined with a straight ledge to mount the rails against. There is no adjustment, and none is ever needed.

How can you weld a frame to hold milling tolerances without some means of correction? The square tube metal on this machine at 10 gauge or so does not have much material left for milling.

You can't, which is why the frames should be stress relieved and machined. Thicker tubing should be used, or more common, a plate is welded to the frame and the plates are machined.

It is my understanding the Camaster is delivered and setup,

My understanding is that they are setup at the factory, and only minor adjustments are necessary. I've never heard of anyone adjusting linear rails or racks on a Camaster.

A little googling shows Camaster is not without needed adjustment or correction.
https://www.camheads.org/showthread.php?t=177

That was 9 years ago, and in the end, it sounds like he just trammed his spindle, which IS a common adjustment.

[micknm]

The MultiCam is just like the Camaster as to what I can see, Rails are NOT mounted on top but under or on the side where they can be adjusted IF needed.

How can you weld a frame to hold milling tolerances without some means of correction? The square tube metal on this machine at 10 gauge or so does not have much material left for milling.

A little googling shows Camaster is not without needed adjustment or correction.
https://www.camheads.org/showthread.php?t=177

Camaster and MultiCam machines are not the same. MultiCam steel frame machines (as well as SCMI, KOMO, Anderson, Biesse, Morbidelli, CMS, Shoda, etc., etc.) are all welded, stress relieved and precision CNC machined anywhere a precision mating surface is required.
"How can you weld a frame to hold milling tolerances without some means of correction? The square tube metal on this machine at 10 gauge or so does not have much material left for milling."
The way MultiCam does it is to weld a flat bar to the frame and then machine the bar. You don't weld a frame to milling tolerances. The correction is in the stress relieving and precision CNC machining done to the frame after it's been welded. If it's done properly, no adjustment is needed. The rails only go in one place. The racks can only go in one place.
And there are plenty of beautifully performing CNC machines that have the rails mounted to the top of the frame without shims.

[ger21]

Camaster and MultiCam machines are not the same

+1

Camaster is not a top tier CNC router.

It's near the top of the "second tier".

High end routers tend to start at $80K-$100K and go up from there.