[FelixPQ]

I hope I'm in the right forum for this.

I have an IronMan from blurry customs. To make things as simple as possible, my situation is that every aspect of the machine orthogonality is simply bad to practically useless, I could list all of them here if necessary.

My problem though is more about fixing it and unfortunately, the only adjustment the machine provide, is the play between the bolts and the hole they go through. In addition to that and or because of that it is practically impossible to adjust anything on the CNC a single direction at a time or in other words, a single degree of freedom at a time. The design doesn't seems to allow much shimming either.

On the positive side, I have ballscrews on all axis, dual on X, I have solid rails something like this.

Attachment 312508

I also have a few other things I wouldn't throw away, Geko 540, Smoothstepper, etc. I don't want to throw good money after bad like they say but I'm already in for close to $7K and can't even cut a stupid little veneered box 7" x 7" x 11" using the miter fold technique.

My work table is not planar and not parallel to the true YX plane, I've faced my spoilboard at minimum 0.080" and I'm not there yet, I'm doing this with a small 1/4 inch bit and it is clearly visible by the marks that the Z axis is not perpendicular to the board just machined. One last little thing, I check using a precision square if my gantry (Y) was square with the X axis and on a length of only 10 inch, I'm off close to 0.085". I double checked just in case...

I don't even know where to start. I'm sure you'll need more then what I said here. Lets say it's just a first step.

Thanks in advance,
Yves

[RCaffin]

Blurry Customs seems to have had a LOT of bad press here recently. It seems they promise, but don't.

You may have to look at a total rebuild. Really total.

Cheers
Roger

[ger21]

You've already started a lot of threads on this, but this one appears to be a new one?

My work table is not planar and not parallel to the true YX plane, I've faced my spoilboard at minimum 0.080" and I'm not there yet, I'm doing this with a small 1/4 inch bit and it is clearly visible by the marks that the Z axis is not perpendicular to the board just machined

As Roger said, you need to look at a total rebuild.

Here's where I would start.
Take the gantry off of the machine, until you are left with just the table, or XY frame.. Also remove the spoilboard.

Is the machine just sitting on a bench or table, or does it have a stand?
Step one, is to make sure that whatever the machine frame is mounted to is as close to perfectly flat as you can get it.

Step 2, is to get the frame square and flat. You may need to take the entire thing apart, and make modifications to it to accomplish this. It may require re-cutting the ends to make them square, or enlarging holes to provide adjustment.
If it's as bad as you say, then it sounds like you may need to be re-cutting parts. If you have to, take them to a machine shop and have them do it.

Once you have the frame square and flat, then you add the gantry, and do the same thing.
And continue to the Z axis.
This is the only way to end up with a good machine. Complete disassembly, and rebuild from the ground up, making sure everything is right each step of the way.

[FelixPQ]

Gerry, Roger,

maybe, I don't understand what you guys mean by a total rebuild. I've spent countless hours measuring and trying to level, to straiten, to square and to tram this machine, I even enlarge a few holes almost against my will. I bought a bunch of tool to help me do all that and yes I improved the result a little bit.

Just to give you an excellent example of a solution the guys from Blurry Customs came up with. I assume at some point they had trouble keeping the gantry beam and the Y rails parallel to each other, so they added a part at each end of the gantry to assemble the beam precisely parallel, every time, no more guessing and waste of time. Unfortunately, they where not as ingenious for other aspect of the assemblies.

This is the kind of solution I'm seeking, an ingenious way that will save time and increase precision. My experience with other type of machine is that they don't stay nice and shiny, you have to take care of them. You want precision, you have to pay the price sort of speak.

It could be as simple as adding one piece on each side, say a 1 inch flat bar maybe 1/4 inch tick, machined to the proper length and bolt them to the side plate such that gantry beam and Y rail are "perfectly" parallel to the XY plane or work table. Say every one thinks it would be a good idea then I would need another idea of the sort to control the angle the Y rail and support beam make with the XZ plane and of course I'll need to bring this angle to 90 degree with the work table.


Attachment 313060

The above image shows this idea of adding a piece to maintain the parallelism of Y rails with the work table.

Tomorrow I'll continue with, sorry I have to go.

Thanks,
Yves

[RCaffin]

countless hours measuring and trying to level, to straiten, to square and to tram this machine,

Precisely.
Ingenious solutions rarely if ever exist.They are a myth.
Strip into component parts and REBBUILD from scratch.

Cheers
Roger

[ger21]

Strip into component parts and REBBUILD from scratch.

Yes, take the whole thing apart, and as you put it back together, fix each flaw as you encounter them.

[FelixPQ]

Gerry, Roger,

yesterday, I forgot a detail, ie. the famous spacer, so here it is.

Attachment 313092

You can see the spacer now, it assembles together the beam and the linear bearing into a single rigid assembly, which makes it easy to maintain the parallelism between the Y rails.

The cyan profile is represent the side plate. As you can see the side plate assemble both the gantry beam-rail assembly and the X bearings from the side. IMHO this is not a great idea, but it's the way it's build. I show a part (green) that would insure the Y axis being parallel with the X rails and thus with the XY plane, just an idea for now.

To be honest, I still don't understand your insistence and or the need to take the machine apart. It remind me of a practice in the old day of MS-DOS when folk reformatted the hard disk and reinstalled everything almost every week. It works for sure but it didn't solved the underlying problem and IMHO, it's the same here.

Thanks,
Yves

[RCaffin]

the old day of MS-DOS when folk reformatted the hard disk and reinstalled everything almost every week.

Well, that was one proven way of keeping the system running, wasn't it?

Why are we 'insisting'? We aren't: it's your machine, and you do what you want. We are just voicing some long hard (painful? expensive?) experience, that's all.

Cheers
Roger

[ger21]

To be honest, I still don't understand your insistence and or the need to take the machine apart.

I don't have your machine in front of me, and can only go by what I've read.
You've been posting about your machine for a few months now.
My understanding was that nothing is square, to the point that the machine is unusable.
And the machine has no means of adjustment to make it square.

So if that's the case, then what I'd do, is take it all apart and start over.

[FelixPQ]

Gerry, Roger,

maybe it's the language barrier or something, I'm a french Canadian or I'm just plain dumb. What you say still doesn't make sense to me because, if I just take the machine apart bolt by bolt, screw by screw and I just bolt it back screw by screw and bolt by bolt, I don't see how this would fix anything,

I would understand much better something like, yes take it apart, keep the good ones (if any) and completely redesign the machine to correct the problems you find along the way. If it's something like that you guys mean, then I'm sorry, I really didn't understand it like that. For me the word rebuilding means something else, more like above.

This weekend I check a few more thing and I even double check because I didn't believe my first observations. To make a long story short, the gantry need a complete redesign or rebuild as you guys would say, the beams are to flimsy, they flex way to much, about 1/16 inch and of course the rails follow that nice curve toward the table. Imagine, I wanted to do 90 degree cuts in Y to do V folds.

Thanks for your time, I think I'll need some time to digest this one.

Yves

[ger21]

I would understand much better something like, yes take it apart, keep the good ones (if any) and completely redesign the machine to correct the problems you find along the way.

Yes, that's what we meant.
However, you may be able to fix individual parts, rather than a complete redesign.

If I tried to answer in French, I'd do a lot worse than you.
Fortunately, when I was in Montreal, everyone I needed to speak with spoke English.

[FelixPQ]

Gerry, Roger,

it's unfortunate it took me so long to figure out we where talking about the same thing but using different words. I'll try to be more vigilant from now on.

I also think I don't how to ask the right questions to get the kind of answer I'm looking for. Essentially, I want my machine to be able do these cuts here.

Attachment 313316

such that the end result look like that.

Attachment 313318


Nice and square little boxes and to do this right every time, my machine needs to cut accurately and repeatably straight, parallel and square in all 3 planes.

How accurate and repeatable it need to be, well I don't know. I don't know either what to do to get my machine "there". Yes, now I understand what rebuilding it means to you guys, but I've spent a lot of money on this machine already and I don't want to spent a single penny more before I'm convinced it's necessary. So I'll try to do it on paper before anything else or at least I'll try to model it first.

Another thing I'll do, I'll model this same box purposely out of square by a known amount, large to begin and I'll reduce it until I can't see the difference in the rendered image. This will be my goal for the accuracy of the machine.

Thanks.
Yves

[awerby]

If you want to use the same parts, then as well as taking it apart, you might want to machine slots where some of the holes are. That way you can slide it into position first, then tighten it up when it's square.

I can't say I understand why you'd bother modeling a box out of square - is the idea to compensate for the screwiness of your machine by drawing everything cock-eyed? It seems like it would be better to make your machine adjustable enough so that the axes can be orthogonal to one another. Then you can give it a model that's drawn square and it will come out square.

[RCaffin]

if I just take the machine apart bolt by bolt, screw by screw and I just bolt it back screw by screw and bolt by bolt, I don't see how this would fix anything

You are right of course. That would do nothing.

What Gerry and I meant was that as you put it back together, you FIX the places which are not good enough. That might mean adding steel to increase the strength or it might mean replacing major parts which are just too weak - such as the gantry.

Yeah, it's a big disappointment when you find that there are major faults. But then you CAN fix them. The rails and leadscrews and motors are more expensive than plain steel. On the other hand, spending a bit of time planning the fixes is worth every cent of your time. And of course, once you take the plunge, you can always keep improving it.

Cheers
Roger

[FelixPQ]

Andrew,

modeling a box out of square is just a simulation thing to see how much out of square I can live with, nothing else.


Roger,

What Gerry and I meant was that as you put it back together, you FIX the places which are not good enough. That might mean adding steel to increase the strength or it might mean replacing major parts which are just too weak - such as the gantry.

Yeah, it's a big disappointment when you find that there are major faults. But then you CAN fix them. The rails and leadscrews and motors are more expensive than plain steel. On the other hand, spending a bit of time planning the fixes is worth every cent of your time. And of course, once you take the plunge, you can always keep improving it.

I bought this machine in 2013 and I left it there since then as soon as I realized the big mistake I made. So this time I perfectly understand the meaning of "disappointment". Recently, I convinced myself to figure out what I could do to make it useful or just sell the parts so that I don't have it in my face all the time. Every time I check something on the machine, I find a new problem.


Gerry suggested I physically decouple the X axis and its slave on another thread, which I did and now I can square the Y axis (gantry) with the X axis. I've ordered some switch and I'll use Mach3 auto squaring routine while homing, this should be acceptable. Add to this a surfaced spoilboard to this equation and now I could use the machine to draw "square" rectangle on a piece of paper, great. But when you take into account Z, things aren't so cool anymore.

Here is an idea I had a long time ago.

Attachment 313436

I would have something like these made such that I would be able to install the side plate of the gantry and be sure the Y rails are parallel and square with the XY plane. Already I see, that it would useful to secure these parts to both the X rails and the Y rails, that way I would have both hands free.

There are 2 points I can see that could still cause problem. I know my Z axis isn't square with the XY plane as it is but is it only because of the Y rails not being square or something else?

The flexing of the gantry beams I discovered this weekend, that's a serious problem that mean the tip of the cutter always follow a downward curve and not a straight line as it should, it also mean I can say goodbye to my V folds. I've look at this and it maybe possible to add a 16 gauge U shape plate and bolt it on the T Slots from above and below. I know this part wouldn't cost to much and if my calculation are correct, even an aluminum plate by itself would do, so no more problem with flexing if I can fit this plate in. I'll go right away in the shop to check this.

Any idea, suggestion or other things I possibly forgot? What do you think of these modification?

Thanks,
Yves

[FelixPQ]

I just verified the space I have to insert a U or a C shape sheet of aluminum or steel and it's a bit less then 0.050" which is actually 18 gauge. But it may be possible to increase that a little. I've double check my calculations for a 6 inch height piece with 1 inch flap (top/bottom) and 0.050" thick and even a sheet of aluminum bent like that is sufficient by itself, now we have the 2 beam and the 2 rails that add plenty enough rigidity that I won't be able to measure the deflection. If someone want to verify my calculation, I have a moment of inertia of 20.88995 and with a 50 pound weight in the center of a 32 inch wide piece, I have a deflection of only 0.000163396" for aluminum and only 5.63434E-05 for steel.

Essentially, this would fix the deflection problem beyond my target of 3 to 5 mils per foot. I don't see any cheaper way to fix this better then that.

Nobody yet commented on the idea of having some kind of jig to maintain the gantry beams at the proper height above and at the proper angle (square) relative to the X rails. If this works as expected, I'd be able to repeatably install the side plate as perfectly as the jig allows.

As it is I can't use this tool below to square the router axis with the XY plane. I've made this intermediate plate model shown here from the back. This plate would allow me to rotate the router from side to side in a controlled fashioned via the center pin. As for the rotation from front to back, I would use shims and both of these rotations are independent from each other and can be done from the front.

Attachment 313482 Attachment 313484

Anyone as something better to propose, I'm all ear.

I should be left with making the Z axis square with the XY plane. I don't how much of the current out of square is caused by the gantry beam and or rails not being square. We see about that pretty soon I hope.

Yves

[RCaffin]

Hi Yves

Yeah, that's getting to be what we call a 'rebuild'.

Written BEFORE I read your latest posting. 4 mm may not be possible?

Diagonal braces for the verticals are the right idea. You still have the problem that the gantry could sway in the Y direction.
Change the plates you are adding to 4 - 6 mm aluminium, and add another 4 - 6 mm plate across the back of the Y axis: height of plate matching the vertical distance between rails. Join them with some solid angle strips. Consider adding some 32 mm Al angle down the side plates right to the bottom to prevent those plates from flexing. Ideas for you.

Imho, Al would be as good as steel if you are mainly cutting wood etc, and it is much lighter. Lots of hex-head bolts done up tightly though.

Along the way you need to check the alignment and get that right, but it is possible! You need a good large right angle square and a good steel ruler 1 m long.

Cheers
Roger

[FelixPQ]

Hi Yves

Yeah, that's getting to be what we call a 'rebuild'.

Written BEFORE I read your latest posting. 4 mm may not be possible?

Diagonal braces for the verticals are the right idea. You still have the problem that the gantry could sway in the Y direction.
Change the plates you are adding to 4 - 6 mm aluminum, and add another 4 - 6 mm plate across the back of the Y axis: height of plate matching the vertical distance between rails. Join them with some solid angle strips. Consider adding some 32 mm Al angle down the side plates right to the bottom to prevent those plates from flexing. Ideas for you.

Imho, Al would be as good as steel if you are mainly cutting wood etc, and it is much lighter. Lots of hex-head bolts done up tightly though.

Along the way you need to check the alignment and get that right, but it is possible! You need a good large right angle square and a good steel ruler 1 m long.

Cheers
Roger

I agree with you, there is a side to side weakness (Y) in those side plates that I'll take care of, this one is easy. I'm not sure I understand the following "Diagonal braces for the verticals are the right idea."

I'll try to take a photo of the back of the gantry to show how these axis are driven and that it is not possible to add a plate in the back. It's the nut that rotates, the screw is fix and the stepper motor follows the along its travel on the Y and X axis. The Z axis is different because it would waste to much space having to move the stepper.

I'll verify again tomorrow but trying to insert a ticker sheet then the one below is not necessary because a part like this can support a load of 25kg (50 pound) in the middle and not flex more then 0.004mm (0.00016"). I'm afraid installing something thicker would require a more complete "rebuilt" of the Z axis. Maybe, I'm wrong with my calculation, I used the formulas of my Machinery's Hand book 23 edition, I know, it's very old.

Attachment 313508

This is basically what it would look like with this new plate bolted on. And yes, I have a pretty decent 450mm (18 inch) square and a meter long straight edge and a few other gauge block, indicators, etc.

Thanks for your input Roger.
Yves

[RCaffin]

"Diagonal braces for the verticals are the right idea."

The plates you showed on the sides of the gantry are effectively 'diagonal braces'. All is good there.

I used the formulas of my Machinery's Hand book 23 edition, I know, it's very old.

Ooh - pre-Internet. Obviously completely wrong. :bs: :bs:
The physics and maths hasn't changed.
Anyhow, 23rd edition? Recent. Mine is 10th Edition.

Yes, there will be constraints due to the existing hardware. Don't expect to get a complete 100% fix in the first go. Do what you can, and then try it out to see where any remaining weaknesses are.

From an engineering PoV, even thin steel plates (eg 18#) can work wonders if they are designed to operate in tension. Give it a try.

Cheers
Roger
(PS: I'll be off the air over Easter.)

[ger21]

Essentially, I want my machine to be able do these cuts here.

such that the end result look like that.

Nice and square little boxes and to do this right every time, my machine needs to cut accurately and repeatably straight, parallel and square in all 3 planes.

How accurate and repeatable it need to be, well I don't know.

We do a lot of miter folding on our big router at work.
I can tell you that it needs to be very precise, or your parts will not line up. They need to be perfectly square, as any amount out of square will cause a lot of misalignment.