[FelixPQ]

Though my CNC machine is a commercial one, an IronMan from Blurry Customs to be exact, I have serious orthogonality problems.

Attachment 301188

Unfortunately, this great looking model doesn't provide any means to adjust anything in a controlled fashion, no setscrews for micro adjustments, no pins of any kind, just bolts and slots which are useless in practice. The design as a table that is not directly connected to the X rails. If you prefer, this mean you could adjust one without affecting the other.

Attachment 301186

Using this tool above I have recorded a maximum deflection of 0.024" which is about 0.0576" per feet and it's not uniform across the table. Again, I have no practical mean to fix this. Maybe you dolt yourself guys have better ideas then mine.


You can assume I have the means to make each X axis rail leveled and planar and also I can make them parallel to each other.

Though I can check almost every thing else I have no mean to micro adjust anything in the current state.

Attachment 301190

Here is an idea of the method I`m considering now. It would allow me to raise or lower the gantry and even rotate it a bit, this on each side independently using the pink parts with the red setscrews at the bottom near the bearing. If I go all the way and add the green parts at top, I`ll be able to rotate the Y rails as a whole on each side independently. This allow me control orthogonality of the YZ the XZ and since I have 2 ballscrews driving the X axis it will be easy to achieve orthogonality of XY plane with YZ plane. I`m open ear to any other suggestion(s).

Unfortunately, I can`t use the tramming tool until I have all my axis orthogonal to each other. Thus ounce I`m done with the Y rails or axis I have to take care of the Z axis and here again I have basically nothing useful to make them orthogonal the other 2 axis. I haven`t given much thought to this axis yet. Suggestions please?

We still have make the table or the work plane if you prefer, parallel to the XY plane. That should be relatively easy now using a spoil board but at the expense of loosing access to the T slots.

Now, I hope you don`t expect the router mounting bracket to provide orthogonality on your first try, that would be like winning the lottery twice in a row. In case someone suggest using shims, I already know they work relatively well for rough alignment but for fine adjustment they`re useless. Even with setscrews it`s difficult, trust me I know.

I think I have an excellent understanding of what orthogonality mean and the kind of problems one will have if its machine axis are not orthogonal but I`m sure many of you have more experience in fixing this then I have.

Just in case, Alex Burt the guy from blurry customs, well he doesn`t seem to know what orthogonality mean to begin with.

Thanks for any help and or suggestion you could make,
Yves

[DonKes]

orthogonality

Damn, it is hard to satisfy engineers !!!!!!!!!!!!

Don

[FelixPQ]

Don,

I`m not an engineer at all but unfortunately for me I`m pretty good in applied physics, I`m just a retired teacher. I bought this machine because it present innovative features and the fact it uses ballscrews having a precision of 0.003 inch per foot and I have to admit it looks awesome, all this lured me into buying it. I was expecting, especially at close to $8k CND at my door, that I was getting a relatively fast and ultra precise CNC machine. Instead, I got half the speed I paid for and a bunch of ultra precise parts nicely bolted together. It`s like having a superb looking V8 in your car but they forgot to make the hole for 2 spark plugs. I`m sure you would complain to.

Yves

[FelixPQ]

Sorry Don, me again.

I most thanks you, I`m so use to do everything by myself that I didn`t thought of it until your message. I`ll go see an engineer and or a machinist hopefully we will find a solution.

Thanks,
Yves

[vmax549]

The ONLY thing the tramming device will be good for is tramming the spindle itself. It is also obvious you have little to NO background in mechanics or machining. You simply start with the frame and set it level(important) then work with the gantry and adjust it to level across Y Then adjust it Square to X . Then tram the spindle in square with XY. Now that you have the Gantry running correctly check the entire table surface for Z. Shim it up where needed. When you get it all to YOUR specs then drill and pin the joints to help prevent them from shifting again

The fact that it all bolts together makes it very easy to true it back up. DO to basic design routers are NOT known for their extreme accuracy to start with. Being a bolt together unit does not help.

Shipping sometimes knocks things out of true. Even brand new $500K machines sometime needs truing up when installed.

[FelixPQ]

I found this video on how to tram your router. I`ll try to be as positive as I can so I won`t comment on the video itself or on his method. I just want you to notice at around 7:30 into the video, that the router and it`s brackets are mounted on a separate plate that is bolted to the Z axis assembly from the front, where obviously it is much easier to make any required adjustment. Add to this the bolted side plates with screws to adjust say the rotation around the X axis. Ok! I would have added setscrews to adjust the plate back and forth but shimming works just as well.

video

All I`m saying here, is that this kind of feature that allows one to make adjustments like this should be the norm instead of the exception especially on commercial machine.

Yves

[vmax549]

Set screws are NOT the preferred method shiming/scraping is. In that video example it is a poor example of tramming a spindle to a table . It is only describing how to tram teh spindle to that piece of glass. That does not mean that the glass is true to teh table. The proper way to do it would be to put down a spoil board then surface teh spoil board true THEN set up teh spindle to tram it AND even then you are ONLY tramming in the actual spindle orintation it does NOT give you any other info of the table's trueness.

(;-) TP

[FelixPQ]

The ONLY thing the tramming device will be good for is tramming the spindle itself. It is also obvious you have little to NO background in mechanics or machining. You simply start with the frame and set it level(important) then work with the gantry and adjust it to level across Y Then adjust it Square to X . Then tram the spindle in square with XY. Now that you have the Gantry running correctly check the entire table surface for Z. Shim it up where needed. When you get it all to YOUR specs then drill and pin the joints to help prevent them from shifting again

The fact that it all bolts together makes it very easy to true it back up. DO to basic design routers are NOT known for their extreme accuracy to start with. Being a bolt together unit does not help.

Shipping sometimes knocks things out of true. Even brand new $500K machines sometime needs truing up when installed.

I agree with practically everything you say above and except for my background in mechanics, your right on. I did exactly as you say and in the same order. I`m sorry but English is not my primary language maybe I write words that don`t express what I mean correctly.

I know you responded to the tramming video link I posted just before, your way to fast for me, I began by saying:

I`ll try to be as positive as I can so I won`t comment on the video itself or on his method.

I gave the link to it not because of way he does a "nice" job at tramming his machine but because his machine is a CNC router, not a milling machine as most other videos I`ve seen and because it shows his Z axis design makes it fairly easy to tram.

My IronMan machine has no intermediate plate like his and bolts from the back and I can`t access the top bracket bolts when the spindle is in, which make tramming my spindle on my machine almost impossible in practice. IMHO having to add new parts like the intermediate plate and the temporary side plate to be able to tram my spindle on my machine is a sign of poor design.

It`s basically the same problem with the side plates of my gantry, it`s as if they use some kind of fixture or whatever to hold the Y axis assembly at the proper height and maintaining it square with the X rail on each side, then bolt these side plates to the X axis bearings. Like you said, bolted machines may easily become out of square in transport, so a good design should account for this and make it as easy as possible to realign.

Yves

[UA_Iron]

I have always found misumi leveling screws to work very well for these situations:
Leveling Bolts-Standard Type?MISUMI?MISUMI USA

Allows you to set height, then clamp down with standard fastener.

If alignment is way out of whack between axes, then using the leveling screws will induce a lot of preload in to the system which ultimately would not be good longevity of the machine. In that case, I have seen set screws with ball ends used along with other fastening features:
Clamping Screws - Ball Type?MISUMI?MISUMI USA

Another common method is a push/pull fastener setup, then once aligned, filling the gap with epoxy to lock into place. Epoxy shimming.

Standard shims are another good method.

[FelixPQ]

Here are a couple photos:

Attachment 301308

This is a side plate on my machine, it connects the X axis bearings to the Y axis assembly. As you can see if you enlarge the first photo, the slots are angle a little bit. If I understand what that means correctly, this limits the potential for adjustment quite a bit. I tried to adjust the Y axis assembly perpendicular to the X axis and at the same height on both side but I must do this the wrong way. Do you have any suggestion?

Attachment 301310

The second photo shows the router installed in its brackets. As you can see, the brackets are held from the back using bolts like the ones you can see on the front plate. I have to add that when the tramming tool is installed, I can`t lower the Z axis enough to access the top bracket bolts and don`t forget the bolts are V shaped and the hole is chamfered which leave very little room to adjust anything. Again, I must be to stupid to fix this, so I`d welcome any suggestion.

By suggestions I mean both how to do it and what tool I could use.


Thanks in advance,
Yves

[DonKes]

I was actually referring to the use of the word Orthogonality. I appreciate the use of the word in this case. It is spot on.

You are on the right track in fixing the problem.

Good skills and Luck

Don

[ger21]

I'd consider making some new side plate. I don't see how you can adjust it with those angled slots. I'm not sure what their purpose, is?

[pippin88]

I suspect the angled slots allow the use of different size linear bearings, eg could fit 15mm or 20mm.

[FelixPQ]

@pippin88: Your guess is as good as mine as what these angle slots are for. One thing I know though, they are not of much help when your trying to square the gantry and making the Y rails parallel to the XY plane.

@Ger21: Yes new side plate??? Your probably right but when I think of doing this I get sick in my gut that I loose all rational thought if you see what I mean...

@DonKes: Thanks again.

I`d like your opinion(s) on the following:

On my machine, the router is not perpendicular with the work table and by far (0.0048" per inch now). There is not much I can do to fix this using the tramming tool I have as I explain before. Since this is the cheapest solution I can think of, I thought of doing this first.

I`ll machine an intermediate plate such that I can make all adjustment from the front, something like this below maybe. The red pin allows the plate to rotate independently of any translation and it should be very easy to set the router axis properly from side to side this way. I`ll use shims to fix it back and forth angle. I know the new plate won`t be perfect, but it should help reduce the problem significantly, at least I hope.

I`m praying I won`t have to make new side plate, I just hoping there is enough play to be able fix any other problem or problems without having to new ones made.

I`d welcome any thought or suggestions on this.

Thanks,
Yves

[DonKes]

Be sure you check the spindle in the XZ plane at both ends of the Y axis.

If The spindle in the XZ plane is the same angle, you don't need to rotate the spindle, only shim it on the top or bottom to make it right.

[vmax549]

First remove the drives so you can slide the gantry around freely. Start on one side of the gantry and then get the gantry running level across the table by adjusting the other side up or down. Do whatever you have to do to the holes to make it adjustable. I would take a small dremal and straighten the slots so the sides can be moved. When you get that part done then go to square the gantry ( lets say it is X) with Y you would SHIM one side of the end of the starting side beam to square it to Y. THEN put in your tram guage and check tram of teh spindle. IF you need to adjust it in Y then slightly lossen the beam to the side plate and twist it to set it correct. IF you need to adjust in X then you slightly snug down the screws holding the spindle mount then tap it with a rubber mallet to bump it into alignment. Once you have it all adjusted then drill each point for a roll pin and install the pin.

Then you may want to put in the indicator and run the surface of the table looking for LOW spots and shim them up to height. after you get it all as close as you can then consider surfacing the surface dead flat to teh gantry. Even if you just knock off a few high spots it would help.

THat is it in a nutshell. It sounds a lot harder than it is.

[FelixPQ]

Thanks Don,

I just hope I`ll get lucky and I won`t have to use an intermediate plate. I`ll have to check this out.

Thanks,
Yves

[FelixPQ]

First remove the drives so you can slide the gantry around freely. Start on one side of the gantry and then get the gantry running level across the table by adjusting the other side up or down. Do whatever you have to do to the holes to make it adjustable.

This may be an excellent idea, I don't know but I'm worried that in practice, with vibration, the static and dynamic forces and the relatively tiny 8 bolts holding each gantry side would inevitably move in some way.

When you get that part done then go to square the gantry ( lets say it is X) with Y you would SHIM one side of the end of the starting side beam to square it to Y. THEN put in your tram guage and check tram of teh spindle. IF you need to adjust it in Y then slightly lossen the beam to the side plate and twist it to set it correct. IF you need to adjust in X then you slightly snug down the screws holding the spindle mount then tap it with a rubber mallet to bump it into alignment. Once you have it all adjusted then drill each point for a roll pin and install the pin.

Here, unfortunately, I`m not sure I understand.

Attachment 301700

Maybe this will help us speak of the same things. My machine doesn't have the trunnion as shown in this photo above, instead it as a table made of 1030 TSlots like the first image I've posted. My X axis is dual like shown here and it is set along the trunnion table as shown here. The Y axis is from side to side along the gantry and of course the Z axis is along the router. There are a couple of differences with this photo and my machine, first the top 2020 TSlot beam just under the drive motor for the trunnion is absent on my machine, this on both side. Instead of a 2020 as shown at the top of the gantry, I have 1020.

Maybe, there are other little thing that make our machine different. For example, I honestly don't see where I could use shims to adjust any axis perpendicularity except at the router mounting bracket level to adjust it back and forth if needed.

Sorry, I don't understand.
Yves

[vmax549]

Look where the gantry BEAMS mount to the side plates . IF the ends of teh beams are perfectly square and teh Y raceways are centered exactly on teh beam then it would move perpendicular to the X axis. To make it perfect you would shim one side or the other of the beams where it mounts to the side plate. You start with the master side and get Y moving perpendicular to X then anchor the far side to its side plate so that it does NOT pull it out of alignment.

You are NOT dealing with machined components they are simply extrusions(aluminum) You have to physically bring every component into proper alignment maunally. Then fasten it so it does not move out of alignment under normal conditions.

[FelixPQ]

Look where the gantry BEAMS mount to the side plates . IF the ends of teh beams are perfectly square and teh Y raceways are centered exactly on teh beam then it would move perpendicular to the X axis. To make it perfect you would shim one side or the other of the beams where it mounts to the side plate. You start with the master side and get Y moving perpendicular to X then anchor the far side to its side plate so that it does NOT pull it out of alignment.

You are NOT dealing with machined components they are simply extrusions(aluminum) You have to physically bring every component into proper alignment maunally. Then fasten it so it does not move out of alignment under normal conditions.

If I understand correctly, you suggest to put shims between the end of the Y axis beam(s) and the side plate on one side or the other. Here is my interpretation of what your saying as a simple drawing. I exaggerated quite a bit the effect of inserting shims where (I think) you suggest (the non vertical red lines).

Attachment 301808

I assume you know there are 2 machined parts between the 2 gantry beams, at each end, that serve to maintain the Y rails or raceways as you call them, parallel to each other. These parts also maintain the ballscrew parallel to rails, it's not shown here.

Lets say this drawing is the front view and for me it's the YZ plane because of the way I setup my machine coordinate system. As you can see, the only critical state that matters from this view point is that all 3 horizontal line, mark with a slash in the middle, must be parallel. Still referring to this drawing there is a second condition for this system to be orthogonal. As shown in the second 3D drawing, the horizontal red lines must also be square (green square) to the XY plane shown here in light gray.

Attachment 301810

There is a guy named Murphy that basically put into law the idea that if something can go wrong, inevitably it will. To me this means I'll inevitably have to realign my machine and probably much more often then I would like. Though your method would likely produce more then acceptable results. I'd prefer spending time and money to make this realignment process something that take minutes, not hours. This can only be achieve by having positive positioning control in a single plane at a time. For example, in my front drawing, this mean, be it with shims or whatever else, one must be able to set the 2 horizontal red lines parallel to the horizontal blue line and (ideally) not affect the squareness of the gantry with the XY plane. The same concept should be applied to the Z axis system and to the router as well. Just in case, I don't pretend my idea is better then anyone else it's just a question of personal preference nothing more.

Thanks for your time and effort and Merry Christmas to you all.

Thanks,
Yves