[GME]

Yes sir, I worked in plastics - injection, rotational, and blow molding - for 12-13 years so I'm well familiar with actual release agents. That's why I put "release agent" in quotes when referring to parchment paper. I know it's not a real release agent but what I figured is that you could use two pieces of parchment paper between the anodized aluminum bearing block mount and the welded steel. Then make sort of an epoxy sandwich. The parchment paper should keep the epoxy from actually sticking to either metal piece, assuming it would at all, and thus become a "release agent".

David

Didn't know your background. I meant no slight. I've used wax paper, paraffin and even polyurethane finish for decades to handle dripping glue. It never occurred to me to try parchment paper. Looking forward to trying it.

[wmgeorge]

I am sure the stepper motors and a lot of the rest is China import as is 99% of the CNC hardware is it seems now. Imported low cost steel kept the prices on everything produced from steel lower. Washers, dryers, cars and the like, anything steel. When the tariffs went on, we pay the tax. Domestic steel prices also increased to match. Until our US production of steel gets back online we will see high prices for most of those goods.

I think the gantry as shown will be fine as shown in those pictures. If my machine was out of square 1 mm in 4 feet I would not bother loosing any bolts. What you do want is no binding and trammed correctly left to right and front to back. This is a router, not a large milling machine making parts for the space shuttle.

[ger21]

It didn't occur to me that the 4x4 had $800 in imported parts

Linear rails alone would be at least that much. I have over $1500 in linear rails for the 4x8 machine I'm building, and those were closeout models at 1/2 price.

[ger21]

I don't believe you should drop the R&P to make adjustments.

Correct me if I'm wrong, but isn't the R&P attached to the aluminum plate over the linear bearings? If so, then adjusting the gantry beam should have no effect on the R&P.

I'd make sure the R&P is properly adjusted before attempting to square the gantry. The linear rails should always maintain that alignment, even when moving the gantry beam around.
If you don't have enough adjustment, just drill the holes out a little.

[PLJack]

So its the blue plate that is the issue? Not in full contract with the base plate? So are you shimming to get the same height as the other side? Otherwise there should not be that much stress or load on those bolts to make any differance. You can prove that by locking the gantry in place and try to see if you get any movement. Otherwise I would contact FLA and see about a solution.

I don't think I made my point. Not news.
I'm saying that if you have the RnP down all machines like this allow for one side to move forward or backward some amount without the other side moving. I think CNCRP even takes advantage of that with some Gcode to auto square on some models.

My statement was that I'm curious how to know what position to start each RnP before raising.

Thanks.
Jack.

[PLJack]

Jack has a problem in need of a solution. If I can be of assistance, I want to help out.

Thanks, that is why we are all here.
Although my "problem" is that I'm learning how to square a gantry from scratch. And you all are watching me bumble through it.
I really don't think I have a defective machine here.

Jack.

[PLJack]

Jack,
I believe I see the problem. For openers, I don't have a good solution for getting rid of the R&P bind other than loosening the gantry bolts and trying to move the bearing block around until everything meshes well.

I don't believe you should drop the R&P to make adjustments. Rather, home the gantry to the sensors and continue from there. As a next step, use blocks and clamps to immobilize the bearing blocks. You immobilize each block both front and back. I believe they need to be in a fixed position, when making squaring adjustments. This makes sense, when you think about homing later on. The sensor blocks have practically no adjustability. Were you to move the bearing blocks around, then you could never be able to use both front sensors for homing. If they are fixed in the homing location, you are moving the gantry relative both sides of Y=0. I would leave the steppers turned on to add additional holding power.

You are left with rotating the gantry on 2 fixed platforms. That can entail having one bolt on one side semi tight and moving the other side, or moving one side, semi tightening a bolt and moving the other side in the opposite direction. It all depends upon how much adjustability you have.

I sent an email to Nate looking for advice on squaring. What I received was a nearly useless perfunctory reply - loosen the bolts and take a mallet to the gantry. However, whether by oversight or intention, there was no mention of loosening R&P.

Also, when moving the gantry, whether to the front or rear, you might try using an f-clamp on the gantry base (or back of the gantry, depending on the direction of movement) at one end, and on the bearing block at the other and try screwing down on the clamp to cause the gantry to move. Any movement would be in a much more controlled manner than hitting it with a mallet.

I hope this makes sense. Let me know if it doesn't.

Gary

Hi Gary. Good ideas. I did have the RnP's engaged when using the mallet. I dropped them for a different reason. I guess my point is that once you are square dropping a RnP will put you out of square.
As for the sensors I'm don't see how they come in to play with squaring.Once I know it's square I plan to drill out the two mounting holes on the sensor blocks to make a slot.
That should allow for plenty of adjustment. At the very least it should be a good indicator that I am out of square if only one side fires during a homing.
I'm sure I will sort out a good process for squaring. Just not there yet.
Hopefully it will be useful to someone in the future.

Jack.

[PLJack]

This is a router, not a large milling machine making parts for the space shuttle.

George keeping it real.
It's easy to get lost in the numbers sometimes on these projects. I'm sure I'll get it all sorted. I've tackled larger problems.
Coming up with a solid squaring procedure is really the only issue I have. Like I said this RnP stuff is new to me. Been all ball and screw in the past.

Jack.

[ger21]

My statement was that I'm curious how to know what position to start each RnP before raising.

You should be squaring the machine with the power off, so the pinions can move freely. This way, you get the gantry square with no tension. So it doesn't really matter where the R&P are.

[wmgeorge]

George keeping it real.
It's easy to get lost in the numbers sometimes on these projects. I'm sure I'll get it all sorted. I've tackled larger problems.
Coming up with a solid squaring procedure is really the only issue I have. Like I said this RnP stuff is new to me. Been all ball and screw in the past.

Jack.

If you want high accuracy you really want quality ball screws all the way around. With R&P there is always a little loss of accuracy, but see what David is doing with his machine. I did some work yesterday that was very very good. Of course with Ball screws there are speed limitations and its far below what our FLA can do. So you trade one for another, and there is also a price point to factor in.

[GME]

You should be squaring the machine with the power off, so the pinions can move freely. This way, you get the gantry square with no tension. So it doesn't really matter where the R&P are.

Gerry,

I've had mixed feelings about the homing setup for the Saturn 2. What you describe is what I consider the "traditional" method, where the homing sensors, which face toward the rear, can be adjusted forward or back to account for the mounts being out parallel. With the "traditional" method, the relative position of the R&Ps, as you say, really doesn't matter.

It appears to me that on the Saturn 2, as designed, homing both sides of the Y axis can only be accomplished if the bearing mounts are first homed, and locked in that position. That's because the sensor trigger mounts are in essentially a fixed and were done that way intentionally. That means moving the gantry around on fixed mounts until square. If the mounts are not parallel, when you call up the homing routine, the stepper would pull one side out of square while it tries to reach the fixed stepper mount. The only way I see to avoid the problem Is do as mentioned, or home to one sensor only and consider it good.

Do you consider the sensor block system to be a design flaw, at least as pertains to Y axis homing? Seems to me that the "traditional" method does not contemplate a sensor block system like on the Saturn 2, and the only way the traditional system can be employed is by using only one Y axis sensor - which isn't necessarily "traditional".

I'm probably missing something, but this is my take so far. In light of the above, it would be great if you would weigh in further in the context of the Saturn 2.

Thank you,

Gary

[difalkner]

Didn't know your background. I meant no slight.

And please understand none was taken, none at all.

You should be squaring the machine with the power off, so the pinions can move freely. This way, you get the gantry square with no tension. So it doesn't really matter where the R&P are.

That's how I did mine, Gerry. No power and things loose enough to set by hand with no binding or issues. The gantry moves effortlessly with the pinions disengaged. And with the pinions engaged and power off I can move the gantry fairly easily though I wouldn't want to do a lot of it unless I want a workout.

David

[ger21]

What you describe is what I consider the "traditional" method, where the homing sensors, which face toward the rear, can be adjusted forward or back to account for the mounts being out parallel.

That's not what I said.
I'm saying that the gantry should be squared with the machine off, so that it doesn't need to auto square. But I would keep the pinions engaged, so the squaring process does not push the pinions out of alignment.

[wmgeorge]

GME Garry when I first saw the non adjustable blocks for homing I was and still am wondering IF that was a construction or build error? They could be remounted to be somewhat adjustable. I never went any further because my machine when pulled by hand with out the R&P drives engaged, was a smooth as glass and even when pulled slightly out of center both sides (of the bearing block) hit the bumper and measured the same as the other. And thats all you really want. PS My blocks have a slot, but the screws are mounted at each end of it, so it effectually can not be moved. You could drill and tap another hole.

Two issues I can see here one is bearing alinement, the other is having the gantry set square so it cuts square without screwing up that critical bearing alinement. Is there a way to lock in the bearing set point with a jig or clamp arrangement?

[wmgeorge]

So what could be considered is after the bearings are alined, move the gantry forward until your near the bumper stops and lay something like 1-2-3 block on that rail, and up against the metal part of that stop. IF it needs to be shimmed so the 1-2-3 block is parallel with the bearing mount do so. Maybe the bumpers need to be removed. Then take clamps and clamp both the left and right sides to the stops and bearing blocks. Do your bolt loosing and adjustment as needed.

Not sure if this is a good idea or not?

[GME]

[QUOTE=wmgeorge;2220900Two issues I can see here one is bearing alinement, the other is having the gantry set square so it cuts square without screwing up that critical bearing alinement. Is there a way to lock in the bearing set point with a jig or clamp arrangement?[/QUOTE]


I've been thinking about it for awhile now. I've thought about taking 4 wooden blocks at least as wide as, the steel tubing the linear rail is attached to. Wider would do no harm, and might be better. Wider would contact more of the gantry mount's edge. Route out a dado a little wider and a little higher than the linear rail. This prevents any stress on the rail. Butt one block up to the front and one up to the rear of each gantry mounting plate and use clamps to clamp the wooden blocks in place. This would effectively trap and immobilize the mounting plates. I would be reluctant to clamping something directly on top of the linear rails. Maybe it wouldn't hurt anything, but when I'm not sure, I tend to err on the side of caution.

There may not be enough room on the bumper side to clamp blocks, so I would probably home, with the gantry bolts loose, then use the steppers to move the gantry in the Y+ direction to create room. It shouldn't mess anything up to clamp the blocks with the mount held by the steppers, and once set, turn the power off.

Anyway, the methodology I've discussed is an attempt to accommodate the lack of sensor adjustment.

Whether there is any utility in using 2 Y axis homing sensors is an open question for me. In other posts, Gerry had said that one should get the gantry as square as possible when doing a squaring operation. I doubt that anyone would disagree. I agree that resorting to auto squaring to solve an out-of-square condition is probably not a best practice, and is something that has the potential to introduce binding and unnecessary stress. However, whether using two sensors on Y might nevertheless be a good idea may depend on how steppers act when power is applied. Do both motors turn on in the same position when power is applied? If not, what is the range of possible variances? It seems to me that if there is some variation, having 2 sensors would be a good way to bring both motors into sync each and every time. I don't know the answers to the questions I've posed, so I've assumed worst cases and tried to work out a method that takes worst cases into account.

Gary

[GME]

So what could be considered is after the bearings are alined, move the gantry forward until your near the bumper stops and lay something like 1-2-3 block on that rail, and up against the metal part of that stop. IF it needs to be shimmed so the 1-2-3 block is parallel with the bearing mount do so. Maybe the bumpers need to be removed. Then take clamps and clamp both the left and right sides to the stops and bearing blocks. Do your bolt loosing and adjustment as needed.

Not sure if this is a good idea or not?

Your above comments posted while I was replying to your earlier post.

I think what you are describing is a variation of what I described. It sounds like you are discussing referencing off the bumpers/bumper plates. I see no reason why it shouldn't work, but referencing off the end cuts the sensor blocks out the the process. What I been discussing embraces the sensor systems and tries to accommodate it. That's why I talk about homing with sensors, immobilizing the gantry and then making adjustments. To help avoid problems with the R&P alignment, I believe the gantry bolts need to be loose when homing.

Basically, the mounting place are set to a fixed and repeatable position and immobilized. Once immobilized, adjustments then proceed.

Gary

[PLJack]

[Note:I'm typing this message as the job goes along]

When I see myself chasing my tail my favorite technique is to start from the top.
First, off with it's head!



The bearings feel great by the way. Quite and no play.
Next is to create that reverence point I've been *****ing about.

.

There it is. The cut on the rail is superficial. Rubbed right off.
I now have two points on both sides that are square off the rails that I can measure from if need be.
This will help with having the bearing shoes in the same location respective to each other.

.


Now lets look at those plates. I'm assuming that stout back end is the foot, as it were.

Left Side:




Right side:




Definitely a warp but not as bad as it looked when they were on the bearings. It seems exaggerated in some of the images for some reason.
I don't see myself beating out those warps. Not with the right angle supports. So depending on how it sits when the X rails are straight up and down I will be shimming the front or back.

The tap holes were a little above the surface so I cleaned those up with a dremel.




Now sitting under it's own weight again.



With both shoes aligned to the new reference marks I snugged up the back boles to see what is what.
Here it is with the bearings aligned and the back bolts tight.




At this point the bearings push little harder but very smooth. Nice sound to them.
I'm feeling better about my second gantry mounting. I know more now than I did the first time.
Looks like I am shimming the front.
I'm going to put the RnP's back on and decide what to do next.

Suggestions?

[wmgeorge]

Your above comments posted while I was replying to your earlier post.

I think what you are describing is a variation of what I described. It sounds like you are discussing referencing off the bumpers/bumper plates. I see no reason why it shouldn't work, but referencing off the end cuts the sensor blocks out the the process. What I been discussing embraces the sensor systems and tries to accommodate it. That's why I talk about homing with sensors, immobilizing the gantry and then making adjustments. To help avoid problems with the R&P alignment, I believe the gantry bolts need to be loose when homing.

Basically, the mounting place are set to a fixed and repeatable position and immobilized. Once immobilized, adjustments then proceed.

Gary

The clamping to the front blocks is temporary just to keep the linear bearings in alinement or indexed. The object is to get the bearing blocks alined and then holding that setting while you loosen those bolts and move the gantry square, if needed. The sensors are not moved with the gantry, they are threaded into the bearing blocks. They should never change, only the steel gantry now is being moved.

[PLJack]

Ran a squaring test with just the back bolts tight.



Right around three mm. At least now I trust what I am seeing.
I'll see if I can close that gap.