Saturn 2 4x4 - Building 2nd Machine

[GME]

Hi, as long as it is possible to get the rails parallel it should be pretty easy.

1. Remove the rails
2. Measure the distance from each thread to the outer edge and draw a sketch in cad.
3. Measure the angle of the threads line to the frame
4. take the CAD sketch an create two rails and set the angle to the frame like measured
5. figure out the needed distance between the rails to mount them parallel

With a dial cauge you can align one rail to the frame edge. Now take a large tube and mount it on the linear block with the dial cauge at the other end and drive along the second rail. You can use a clamp to create a little force on the rails side. Start at one end and work along the screws until it is parallel.

Hello Millhouse, and thank you for your input. Are you overlooking something? What you describe may work with an rail that hadn't been drilled and threaded, but how would you accommodate this situation, where the rails are already mounted and the the holes in the rails aren't oversized enough to allow for the amount of adjustment that's needed. What you describe, and what I need, is just going to be a small fractional (in the several thousands) move. So, once you realign the rails that fraction, you would necessarily have overlapping holes, which won't work. As I've mentioned in prior posts, I believe the holes would need to be weld filled, ground off, and new holes reestablished. Also, I don't trust the frame rail edges to be consistent enough to use as a reference surface. Using a taught piano wire seems to hold promise, I think. Very old school, but I don't have the equipment to employ the latest techniques.

As far as aligning the rails for parallel, I believe you are describing a technique where the first rail installed is called the master rail, and the other rail is set to parallel off of it. I seem to recall that this isn't the preferred method, but maybe a common one. As non-professional DIY'er, it may be the only one available to me. I'm stilling looking for info on using piano wire, and there may be information about how to use it to set rails parallel.

Am I misreading you, or was your first sentence intended to mean, "as long as it is possible to get the rails 'in the same plane' . . . ."? IMHO, nothing about trying to rehab this beast is going to be "pretty easy." I agree that there are techniques to get things the way they should be, but some, like milling and/or grinding, are almost certainly be budget busters. Others would benefit from specialized equipment I don't have. (I have the basics, dial gauges, precision squares and straight edges, but that's about it). Since I already had the electronics and motors, I budgeted for the frame and stand only. I did not anticipate having to make, and pay for, major modifications to get the frame to within the specs Fineline advertises for the Saturn 2. Fineline advertised, and continues to advertise:

"We take quality seriously, which is why every frame is:

• Built with an accuracy of 0.1mm."

.01mm is .003937" So far, it appears that pretty much nothing is actually built to that specification.
Edit: I meant 0.1mm is .003937". I intended to type the 0 before the decimal point, but my fingers got in the way. GME
Thank you, again. If I've missed something, or misinterpreted you comments, please be sure to let me know.

Gary

[difalkner]

I keep reading about filling holes and re-tapping but are the rails hardened? They weren't on the early machines but somewhere along the way I thought I read the new rails may be hardened. Anyway, inaccuracies and not being built to spec aside, if they aren't hardened can't you just enlarge the holes in the rails to allow enough movement to align them? To me that's easier than filling and tapping the frame.

David

[ger21]

0.1, or 0.01?

[GME]

0.1, or 0.01?

0.1mm. Good catch, Gerry. I appreciate it. I did an edit in my post, so I don't confuse folks. The quote from the FLA site (different type face) was a cut and paste and is accurate.

Thanks again,

Gary

[GME]

I keep reading about filling holes and re-tapping but are the rails hardened? They weren't on the early machines but somewhere along the way I thought I read the new rails may be hardened. Anyway, inaccuracies and not being built to spec aside, if they aren't hardened can't you just enlarge the holes in the rails to allow enough movement to align them? To me that's easier than filling and tapping the frame.

David

Thank you, David. I thought about what you suggest, but decided it wasn't workable for me.

Without going into excruciating detail, suffice it to say that the amount of error to correct would necessitate drilling out both rails - to divide the error in half. The hole size required to drill just one rail would be more than 12mm, which would mean that the head of cap screw would pass through the adjustment hole and wouldn't tighten against anything but the frame.

A cap screw has about 2mm bearing surface (from the threads to the edge of the head). If I enlarged the holes in both rails, I would cut the bearing surface just about in half. That's too small a bearing surface for my taste. A button head cap screw (BHCS) is out, because the screw head is the same diameter as the allowance hole, and would prevent any adjustment. To accommodate a BHCS would mean enlarging the access hole, and I'm not prepared to do that. Too much metal to remove, and too many opportunities to mess it up.

Oh, and one other thing. I've found screw holes in the gantry that were tapped pretty crooked; the bolts were very cockeyed. Obviously, the holes weren't drilled, and the threads weren't cut, with any sort of guide. If any of the rail bolt holes were done the same way, any discussion of enlarging holes would most likely be moot.

Gary

[mactec54]

I keep reading about filling holes and re-tapping but are the rails hardened? They weren't on the early machines but somewhere along the way I thought I read the new rails may be hardened. Anyway, inaccuracies and not being built to spec aside, if they aren't hardened can't you just enlarge the holes in the rails to allow enough movement to align them? To me that's easier than filling and tapping the frame.

David

All Linear Rails are case hardened so it would not be an easy task to open the rail holes, plus you would most likely get added distortion in the rails from doing this

If he can't correct the rails by the existing mounting screws / holes, then they need to send him a new machine, there is just to much wrong with this machine to expect someone to fix it

When I install rails I use the next size up SHCS so that there is more head for better clamping of the rail, the holes in the rails are large now I can't imagine having to make them larger

[GME]

I hope this is not a bad sign. I tried to go to the FLA website to log into my account to check on a part of my order that did not ship, and I asked to be held for now. Here is a screen print of what I got when I tried:



I have tried several times, but the result has been consistently the same. I don't recall ever seeing a message like this on the net, but others may have. I'll keep trying to reach the site and report back.

You will note that the address I tried drills into the site. It's a default address that comes up in Chrome. I also tried to go the home page, Finelineautomation.com. Same result. The whole site is down.

I called and got the usual voice mail, so the phone is still operational. It wouldn't necessarily shut down right away, though. Hopefully, it's just some technical issue with the website.

Gary

[difalkner]

I can still ping the domain so it's probably nothing more than it appears to be, albeit bad timing given the communication issues. When I clicked on the Technical Details I got a little more info -



David

[DDgitfiddle]

I can still ping the domain so it's probably nothing more than it appears to be, albeit bad timing given the communication issues. When I clicked on the Technical Details I got a little more info -



David

The home page came up for me this evening. Might have been a temporary problem with the web site earlier today. Hopefully just unfortunate timing.

[GME]

1. Fineline is still around and the website is back up and running.


2. I've been back in contact with Nate. He is going to make things right. One of the options I have is try to rehab the machine.


I am not too excited about trying to rehab this machine, but I have a potential plan and would y'all to critique it. Here goes.


Step 1. Remove the y-axis linear rails and remove the paint from the top of the steel tubes.


Step 2. Form around the steel tubing and create bridges between the 2 Y axis frame members.


Step 3. Use Precision Epoxy's SC-15P self leveling epoxy to create a flat surface for the linear rails. This will involve boxing in the top of the frame rail, creating 3 bridges across the width to ensure both rails are coplaner. Epoxies Page

Step 4. Have an aluminum machined to act as an interface between epoxy surface and the linear rails. What follows is a picture of what I had in mind:



The plate is 3/4" thick and 3" wide. The slot milled down the center is to give me reference edge for the edge of the linear rail to mount against. It is 1/8" deep. The slotted holes on the edge are from mounting on the frame. The are slotted to allow for lateral adjustment, which is to ensure I can get the linear rails parallel.

5. Drill and tap 16 holes on each frame rail, using the interface plate as a guide. Install the interface.

6. Mount the linear rail and adjust for parallel.


The feet of the gantry are not flat. There is some irregularity and a bit of twist to the feet. Fortunately, the deviation is just a few thousands. My plan is to use an epoxy putty to level things out. It would involve figuring out what shims would be needed to achieve perfectly level, super gluing the shims to the underside of gantry mounting surface, applying a thick epoxy putty to the gantry interface plates, and lowering the gantry onto the interface plates and loosely tighten down the gantry. I may use clamps, because of epoxy squeeze out. I will cover the threaded holes and use an angle grinder to remove epoxy from the mounting slots. I assume I will need a small burr to remove the epoxy from the area over the threaded holes. In the alternative, I can liberally apply a release agent to the mounting screws and hold the gantry in place with them. I will want to do some experimentation, before trying it with screws. I plan to use a release agent on the interface plate, and let the epoxy stick to the underside of the gantry mounts. That should leave me with good, flat surface, and promote adjustability for squaring the gantry.





The rails on the gantry are also out of parallel a bit. If things break right with the mounting screws, it may be able to adjust the linear rails to parallel. I they don't, I will be looking as something like the above, only I use only one interface plate and will have two slots milled in parallel for mounting the linear rails. When inspecting the gantry, I noticed something that surprised me. Under each linear rail, there is a strip that looks machined. The edges aren't crisp, like would occur with milling. I though maybe it was the powder coating that made it seem the edges were rounded. I have no experience with powder coat, other buying some things with that finish. I've seen it thin, and I've seen it thick. The Saturn 2 is appears to have been done on the thicker side. Attached is a photo of the gantry depicting what I'm speculating is a milled slot. Attached are a couple of photos depicting the part of the slot that's visible:




Okay, am I on the right track, or is even starting down this road a huge mistake?

I have the option of sending the machine back and walking away. The problem with that option is it leaves me back in the market for a CNC. The only thing I've found for this size machine in the price range is the CNC Router Parts 4 x 4 pro. It looks nice for an 80/20 extrusion machine, but has the comparative lack of rigidity and adjusting the gantry for square is problematic, given the way it is designed. The folks at CNCRP recommend adjusting the gantry via the limit switches, which is just another way of saying that they expect you to force the gantry. I don't like that approach. I have some workarounds in mind, but it would involve cutting slots in the gantry interface plate and using shims on what they call a joining plate, which fastens to a piece of 80/20 mounted to the riser plate.

I appreciate any and all input, including questions on what I've covered in this thread. All comments welcome.

Thank you,

Gary

[difalkner]

I'm glad Nate is going to make it good for you. My assumption is that he'll cover the expense if you're having the pieces made or will make them and ship to you. And I'm glad you got in touch with him. I'm not getting any replies to text or email but then I don't have a problem that he needs to tend to other than shipping some parts he said he'd ship.

From what I read of your solution it sounds like it should work. What will be the deciding factor(s) before proceeding?

David

[ger21]

The only thing I've found for this size machine in the price range is the CNC Router Parts 4 x 4 pro. It looks nice for an 80/20 extrusion machine, but has the comparative lack of rigidity and adjusting the gantry for square is problematic, given the way it is designed.

I would think adjusting the gantry for square on a CNCRP machine would be much easier than adjusting a Saturn, since the CNCRP machine has a ton of adjustability.
Unless he's giving you a lot of money back to cover your trouble, I'd send it back and go with the CNCRP machine.

[he1957]

Why not send it back and get a new one that works as designed and intended?

The maker can then examine what went wrong with the bad unit you have and use that to fix the errors in manufacture. This is what should happen with _all_ lemon machines (cars included) so they can produce oranges which do not leave bitter tastes.

Why would anyone suggest squaring a gantry with limit or home switch sensing? All this will do is introduce twisting forces on the structure. If it's out of square then its out of square and using such techniques will introduce curls in cut patterns. Remember, the cutter needs to be square in and to all 3 axes to cut flat and perpendicular in/and to all 3 planes.

A$0.02.

[ger21]

Why not send it back and get a new one that works as designed and intended?

Based on the experiences of several recent buyers, It would appear that the chances of getting a "good" machine are about 50/50.

[he1957]

Oh, that sounds disappointing. I read Nates Return policy here: https://www.finelineautomation.com/returns

For a returned machine - in whole, it would be in everyones best interests (future Customers included) to ensure the replacement machine is spot-on before shipping :-)

[DDgitfiddle]

Gary,

I'll throw my $0.02 in. The basis of your workaround is using a leveling epoxy to flatten things out. To get this right, you will have to do the following things:

1) Set up "dams" to pour the epoxy, making sure they are well sealed to the frame to avoid significant leakage.
2) You need to make sure that the epoxy is poured so that the left and right sides are level with each other. You will have to have the frame very level to do this.
3) The aluminum mounting block you sketched up will need enough adjustment to straighten out the rails.
4) You will have to drill and tap alot of holes pretty precisely so that the aluminum mounting block will mount without using up all of the adjustment that you have designed in.
5) You have to flatten out the feet of the gantry as you described without getting epoxy putty all over bearings and other precision surfaces. This while creating a fairly thin layer of putty.
6) You have to deal with the rails on the gantry being out of parallel (maybe not a major issue if you have enough adjustment).

When you are done, you will have added alot of additional component interfaces to the structure of the system. This will have the net effect of reducing the stiffness of the system. One thing I learned pretty early on in machine design is that it is best to avoid having paint in the middle of structural bolted joints. Well-designed, properly tightened bolted joints will not loosen over time and can hold tremendous loads. Tensioning the bolts slightly compresses the metal at the interfaces, and these materials act as VERY stiff springs (essentially equivalent to welding the materials together). Plastics are not nearly as stiff as steel by a factor of close to 10:1. So putting paint in the joints reduces the stiffness of the joint a little. In addition, over time the paint creeps and reduces the tension on the bolts. This reduces the preload on the joint allowing it to slip. I fear the same phenomenon with the bedding epoxy, at least at the gantry to bearing plate interface. If you could get a metal shim at each bolt hole, this would not be an issue. (I REALLY am not trying to restart the epoxy leveling discussion, I'm just trying to present the 20,000 ft view of the machine you would have if you went that route).

My feeling is that you are going to put a TON of effort into doing this and you will have a result that, on the whole, is not what you set out to achieve and you will realize it every time you use it. I don't think the CNC Router parts machine is as non-rigid as you might think. When I was doing my research initially, I calculated that a 3"x6" 80/20 aluminum extrusion is really not that much less stiff than an equivalent sized steel box beam. The steel is stiffer to be sure, but the difference was measured in 0.0001" increments or less. Relative to the errors in positioning, I really don't think this will be an issue unless you really want to cut alot of metal. Better machines exist, but they are all much more expensive. When money is no object, you can approach perfect compromises. The Saturn machine represents one option for making those compromises (even when it is built as intended). The CNCRP system represents another approach. One benefit to the CNCRP machine is that you can carry the parts into your shop without a gantry crane.

At the end of the day what will bring you more enjoyment, fixing this machine (and the satisfaction/disappointment that could come with it) or building the projects that you have planned once the machine is running? Building my machine was alot of fun, but I didn't have to deal with correcting these sorts of problems.

-Robert

[GME]

My assumption is that he'll cover the expense if you're having the pieces made or will make them and ship to you.

From what I read of your solution it sounds like it should work. What will be the deciding factor(s) before proceeding?

Your assumption is close. One of the options involves a discount, which amounts to a repair allowance. I believe it will be up to me to do the rehab, including milling work I will have to hire out.

What will be the deciding factor(s) before proceeding?

David

Cost of rehab vs cost of a different machine
The length of downtime before the machine is finally running and I'm back to cutting
My confidence level re: whether I can pull off a rehab and whether it will turn out as planned

I would think adjusting the gantry for square on a CNCRP machine would be much easier than adjusting a Saturn, since the CNCRP machine has a ton of adjustability.
Unless he's giving you a lot of money back to cover your trouble, I'd send it back and go with the CNCRP machine.

I'm surprised to hear the CNCRP machine "has a ton of adjustability." I'm not doubting you. It's just after studying their assembly instructions and seeing how the gantry and related parts are put together, it looks like there would be little or no adjustability. This is good news indeed, because forcing the gantry square with limit switches isn't something I'm prepared to do.

FWIW, CNCRP is located in North Bend, WA. That's about 60 miles from me.

Why not send it back and get a new one that works as designed and intended?

The maker can then examine what went wrong with the bad unit you have and use that to fix the errors in manufacture. This is what should happen with _all_ lemon machines (cars included) so they can produce oranges which do not leave bitter tastes.

Why would anyone suggest squaring a gantry with limit or home switch sensing? All this will do is introduce twisting forces on the structure. If it's out of square then its out of square and using such techniques will introduce curls in cut patterns. Remember, the cutter needs to be square in and to all 3 axes to cut flat and perpendicular in/and to all 3 planes.

A$0.02.

As Gerry observed below, even odds of getting a "good" machine, which would include a replacement. The lack of "out the door" quality control is disturbing. Nate offered to video his making the measurements I want before sending a replacement, but I would want to do it myself. The fact that this machine made it to my door doesn't exactly instill confidence. As much to the point, the poor quality control starts at the manufacturing facility, and there are obviously serious problems at that end. A short while back, someone observed that overseas manufactures sometimes make prototypes and early models to or near the expected spec, and then later let quality go to the devil. I'm inclined to believe that is probably the case here. FLA is a relatively small fish in the realm of global manufacturers, so I suspect it has little leverage. This probably translates into more of the same for the foreseeable future.

The only see one reason for suggesting using homing switches to force a gantry into square. There is no other practical way to do it. To paraphrase your comments, it is just moving the error around. It doesn't correct it. And yes, I understand the need to tram the spindle. I bought and use a commercial tramming gauge to ensure I get it set correctly.

Based on the experiences of several recent buyers, It would appear that the chances of getting a "good" machine are about 50/50.

The only quarrel I have with your probability estimate is that it may be a little too optimistic. Caveat emptor.

Gary

[GME]

Gary,

I'll throw my $0.02 in. The basis of your workaround is using a leveling epoxy to flatten things out. To get this right, you will have to do the following things:

1) Set up "dams" to pour the epoxy, making sure they are well sealed to the frame to avoid significant leakage.
2) You need to make sure that the epoxy is poured so that the left and right sides are level with each other. You will have to have the frame very level to do this.
3) The aluminum mounting block you sketched up will need enough adjustment to straighten out the rails.
4) You will have to drill and tap alot of holes pretty precisely so that the aluminum mounting block will mount without using up all of the adjustment that you have designed in.
5) You have to flatten out the feet of the gantry as you described without getting epoxy putty all over bearings and other precision surfaces. This while creating a fairly thin layer of putty.
6) You have to deal with the rails on the gantry being out of parallel (maybe not a major issue if you have enough adjustment).

When you are done, you will have added alot of additional component interfaces to the structure of the system. This will have the net effect of reducing the stiffness of the system. One thing I learned pretty early on in machine design is that it is best to avoid having paint in the middle of structural bolted joints. Well-designed, properly tightened bolted joints will not loosen over time and can hold tremendous loads. Tensioning the bolts slightly compresses the metal at the interfaces, and these materials act as VERY stiff springs (essentially equivalent to welding the materials together). Plastics are not nearly as stiff as steel by a factor of close to 10:1. So putting paint in the joints reduces the stiffness of the joint a little. In addition, over time the paint creeps and reduces the tension on the bolts. This reduces the preload on the joint allowing it to slip. I fear the same phenomenon with the bedding epoxy, at least at the gantry to bearing plate interface. If you could get a metal shim at each bolt hole, this would not be an issue. (I REALLY am not trying to restart the epoxy leveling discussion, I'm just trying to present the 20,000 ft view of the machine you would have if you went that route).

My feeling is that you are going to put a TON of effort into doing this and you will have a result that, on the whole, is not what you set out to achieve and you will realize it every time you use it. I don't think the CNC Router parts machine is as non-rigid as you might think. When I was doing my research initially, I calculated that a 3"x6" 80/20 aluminum extrusion is really not that much less stiff than an equivalent sized steel box beam. The steel is stiffer to be sure, but the difference was measured in 0.0001" increments or less. Relative to the errors in positioning, I really don't think this will be an issue unless you really want to cut alot of metal. Better machines exist, but they are all much more expensive. When money is no object, you can approach perfect compromises. The Saturn machine represents one option for making those compromises (even when it is built as intended). The CNCRP system represents another approach. One benefit to the CNCRP machine is that you can carry the parts into your shop without a gantry crane.

At the end of the day what will bring you more enjoyment, fixing this machine (and the satisfaction/disappointment that could come with it) or building the projects that you have planned once the machine is running? Building my machine was alot of fun, but I didn't have to deal with correcting these sorts of problems.

-Robert

Good points, Robert. My three main worries when evaluating a rehab have been getting the epoxy done right, and drilling an tapping holes accurately and having to deal with a tap broken off in a steel beam and not being able to get it out without buying an outrageously expensive tool.

From what I've read about epoxy, it should hold up just fine. That is, if it's mixed and installed correctly. I found one firm that uses epoxy for rebuilding ways on monstrously large machines (tons). The internet provides mixed results regarding self leveling epoxy, and when folks seem satisfied, you never really know about their tolerances for error. Maybe it didn't turn out so well, but it was so much better than their starting point that they just sucked it up and moved on. The internet has so much information, but so much of it is bad information. For me at least, it's often to separate the good from the bad. I suspect too much is taken at face value.

Yeah, I get it. I really wanted the Saturn to be what the specs promised, and it has to look like I'm trying to get the machine I wanted at all costs. However, when all is said and done, I will end up not using the machine, if I can't get it right. I won't be able to bring myself to turn it on, if, in my heart of hearts, I feel like it's junk. Time to adjust my strategy.

Thank you for your perspective. Much appreciated.

Gary

[ger21]

I'm surprised to hear the CNCRP machine "has a ton of adjustability." I'm not doubting you. It's just after studying their assembly instructions and seeing how the gantry and related parts are put together, it looks like there would be little or no adjustability.

Don't take my word for it, as I don't have any experience with one. I'm assuming that since it's extrusion based, that everything has some adjustment? Based on all the post about adjusting the gantry on the Saturn, I have to believe that the CNCRP is much more adjustable??

I just looked at the manual, and the gantry extrusion is attached with t-nuts in the slots, which should give you a little bit of wiggle room.
I'm pretty sure that their parts are machined pretty accurately, and getting it square shouldn't be an issue.

I LOT of people here are using CNCRP machines, and I've never heard of an issue assembling it square.

[wmgeorge]

I am with Gerry, send it back. There are lots and lots of CNCRP machines being used everyday. It does NOT have to be pulled square with both motors if its level and squared when put together. Download the assy manual for the machine and read it. Everything I have purchased from CNCRP I have been impressed with, they are a quality company with a track record.

IF Nate could 100% guarantee you a new machine with Zero defects thats another thing. I love mine and just finishing up the spindle install. I do not plan to be a frequent poster here, because I do not want to cause problems.

Added: One other .... consideration a repaired with epoxy machine would have zero resale value.

The Camaster Stinger series looks to have a aluminum gantry, this one was $14k 6 months ago when I priced before I got the FLA.