Starting to design out my build, looking for critique

[Chrono]

Hey everyone! More progress on the redesign. I sat down and actually worked on it everyday this week, and got some real progress done, finally. I am much more satisfied with how that one turned out, compared to the old one. Here is a side-by-side:


It is probably obvious, but the old one to the left, the new one to the right. This will be the convention for pictures for the rest of the post as well. (On a side note: The blue extrusions are going to be changed from 45x90 ones to 45x45 ones, to save money. I just have not gotten around to fixing them in CAD, since they were pretty low on the list in terms of priority)
While being at it, I also changed a whole couple of things: Like the Z-Axis

New:

The Z-Axis got completely changed. Where it was a complex part before, it is now a rather simple bent sheet steel piece. The motor is now also affixed to the Z-Axis itself, and is moving with it. This is to keep the Gantry-side Y-Axis sled small. It also theoretically means, that I don't need to change anything but the Z-Plate, the rails and the Gantry sideplates to add more Z-space later on, if I ever decide on it. All in all, I like the design change of putting the bearing blocks on the Gantry sides, and the rails on the Z. Allows the parts to shrink a bit. That is one design change I picked up from AndyUK's build. The Ballscrew is also mounted on the moving Z-Axis, with the Ballnut being fixed on the Gantry Sled. One question that I still have, however, is how the cutting spindles actually get mounted to the Z-Axis. I understand that the spindle itself is mounted in the red aluminium ring, but how is that ring mounted? None of the pictures I could see in the offers showed any mounting holes. Are you meant to drill those yourself?

The Gantry itself got a revision as well!

New:

The obvious thing is probably the size and shape change. I went from an L-Assembly to a C-Assembly. This is so that I could move the rails up front, which should not only give me more rigidity overall, but also allowed the massive simplification in assembly for the Z-Axis and Y-Axis Sled. This is one change I really took a liking to.

Something that was not quite obvious from the last post, but was already in it, is that one of the side plates fixing the gantry beam to the X-Axis is much thinner than the other one, in this case the left one is thinner (3mm of steel) than the right one(10mm of steel). This is to give the machine a way to have a bit of play, should something not be fully aligned, and off by a tiny bit. Another measure to the same effect has been brought up by my colleagues multiple times now: To only put linear rails and bearing blocks on one side, and have the other only constrained against translation/force, not torque, and we finally came up with a workable example, I want to discuss here:



Let me explain. The plan would be to put the second linear rail on the lower extrusion on the right (looking from the front of the machine), and have both of them on the same side. Then, instead of a rail, have a supported smooth steel rod, and a cylindrical type round linear bearing on the left side, to replace the rail. That way, the linear bearing still constrains the movement on the translational axes and takes on the forces, but it would permit torque, which would fix the problem with it being statically overdefined. What do you guys think of that? None of the Logs and posts I have read here, so far, have reported any real problems with any of this, but one of my colleagues who has had contact with similarly built machines, with a different purpose (cannot exactly remember) is bringing it up as the thing they and their competitors did, to avoid problems in the same vein. And looking at the budget, one rail with linear bearings should not be the point that bankrupts me here :D

Another question I am having is regarding the long ballscrews on the X-Axis and the motor calc sheet. As discussed way earlier in this thread, I want to stay with a single stepper for the X axis, even though I have two ballscrews, linking them with timing belt. Now my question is, in the Calc sheet, don't I need to input a "times two" into the calc of the torque needed to spin the ballscrews, since the stepper needs to drive both of them at the same time? Since I want to go to 2010 or even 2510 ballscrews, that is one hefty variable to think about. From my initial messing around, I can either go less fast with a 2010 ballscrew (Around 6.3m/s rapids) with a weaker stepper, or find a much stronger one to drive the 2510 ballscrew at the full 7.5m/s rapids. Of course, cutting is another thing entirely, and I am still not quite sure what speeds to aim for here, but for reference, I have managed to get around 2.5m/s with 50N of cutting force, in my testing. Another way would be to stay with 1610 ballscrews and scale the X down from 2m to 1.7m, but that is something that I only want to be doing as a last ditch effort, so to speak.

And another on the ballscrews: Would I just mount normal timing belt pulleys to them, to connect them to the motors, or would I need special machining or parts here?



Regarding the Electronics: other than the Steppers, I have also been looking at all the other electronics. And I think I have it down to the following:

• X-Axis Stepper: Still looking
• Y-Axis Stepper: ST8918S4508-A
• Z-Axis Stepper: ST5909M2008-A
• Stepper Controllers: Either Leadshine AM882s or EM806s times three, after AndyUK's recommendation, still trying to figure out the exact differences there.
• Controllers: Either the UC400ETH or the Smoothstepper ESS, although I am still dabbling about the electronics section of the forum. I don't think that I will be getting breakout boards, since I can easily deal with electronics, and just soldering the proper cableing sound much more reasonable to be, instead of buying a 130 pound breakout
• The VFD will likely be a YL620-A packaged with a chinese 2.2kW watercooled spindle, alongside a water pump
• And finally: six microswitches for end stops. A thought here: would there be any profit had from using proximity switches?

Regarding the choice of controllers: I had a quick dabble with the Mark 4 demo, and as I understand, it is pretty much indifferent to the exact board I choose, and the board itself does also not really need to be wired up in any special way, since I set all the necessary port/pin assignments in Mach 4 itself (granted that I match input pin to input, output to output, etc.)?

When it comes to schematics: I just installed AutoCAD today to get something done, and thankfully I have got some people with a background in high-power electrical installations to help me with the cabinet and all the safety measures needed. After I have everything in there, I can think about all the PSUs needed.


All in all, I am really satisfied with the progress had in a week's time, and I am constantly working to finally get to the point where I am comfortable with the end result. If anyone spots anything that seems off with the revision, however, I would be very glad for any criticism on it. Ultimately, the more eyes and voices are involved, the less likely that mistakes fall through the cracks. And to that effect, here is the link to the web interface of Fusion, where you can take a look around the newest, up-to-date version of the 3D CAD Model yourself, and take a closer look.

On the note of getting a lot of feedback, however, I am noticing that other post here get a lot more traction, when it comes to user response. Don't get me wrong, I am by no means meaning do downplay the contributions made my AndyUK, CliveS and Washout(Who literally brought me here), you guys are the reason I have comes this far, and I cannot thank you guys enough , but at the same time, I am wondering if my post broke etiquette, or if I did something else, that made people not want to engage here? Because if that is the case, I would like to work on fixing that.

Anyways, I am already staying up way too late to finish up this post, so I am just gonna end it here and head to sleep.
Thanks,
Chrono