Starting to design out my build, looking for critique

[Chrono]

You're planning to beef up the side rails, add the bed whilst you're at it.

Much advice has already been given to you and there is sooo much more on this site from some VERY talented, knowledgeable and experienced people :

http://www.mycncuk.com/forums/253-DIY-Router-Build-Logs

Don't get too hung up on the design of CMM's.

Time you put a new blade in your hacksaw!

I have to completely agree, without all the advice from the people here I would not even have remotely come this far!

Currently I am reading through some of those build logs to hopefully find a good one about the PSU section of the build. All the electronics are pretty clear to me, but the interface between machine voltage and net voltage. I see that a lot of people build their own transformers, which is definitely an option, but I am still looking into that. Most of what I am still missing on the machine side mostly related to the spindles and which I want to have in hand, before commiting there.

If all goes to plan, I thus want to sort out the electronics over the next two weeks and then move on to buying parts. That is when I want to make a final pass and move on to the next step, if there is not a flaw that needs sorting out. Once I do move on to buying, I will probably make a proper build-log post as well.

I am also scouring the posts for something related to cutting forces and speeds, to be able to dimension the steppers. As far as I could gather from various posts here, I would pretty much always like to use the steppers in bipolar parallel connection, and using that, getting all the info and all the steppers I can find, I have not found a single one, so far, capable of driving two 2510 ballscrews at 7.5mm/s rapids... who would have thought? :D So I am thinking about moving to a one-stepper-per-screw setup. With that, however, I am still looking at one of AndyUK's first pieces of advice to me and hesitating. So I am thinking about putting a timing belt between the two motors to synchronize them as well, but I don't know if that is a good idea or if it even works. Looking at the motor calc sheet, however, there is a motor that would permit me to get 6.1m/s rapids with only a single motor, as well as letting me cut at 2.1m/s with 50N of cutting forces. But then again, I am still searching for posts on speeds to look out for.

[Chrono]

Long time no post, was busy these last couple weeks, but I am finalizing the build. Pretty much everything is done, except for the timing belts and the electronics layout.

I solved my motor problem by simply shrinking the machine to a tiny 1.8m on the X-axis. This allows the use of the smaller 2010 ballscrews, and, in turn, a single motor.

Other than that, I have been verifying dimensions and hole placements and been putting in manufacturer CAD files for all the rails and screws. I have been putting together some angles for the inside (the yellow one in the left side of the machine in the picture) that I can remove if I need the width and otherwise keep in. Since the CAD file already takes ages to change anything in, I only put in one, to keep the computation time down a bit, but the plan is to have two per column.

I guess I only have 4 questions left:
1. I have seen that people usually use transformers for their steppers, and I understand that that is because regulated PSUs may have problems with the current spikes (at least according to the app notes). I could not find a proper post on how to do it on here though, so If anyone has a good source on that, that would be perfect.
2. How do people usually tension their belts? I am still trying to figure out where to place my tensioners.
3. Where do you even get belt long enough for something like this, and what type of Belt? Using a manufacturerer's tools, I have been able to calculate that my belt would be approximately 4.8m long, for the X-Axis. I have not found closed-loop belt in that size yet.
4. Is there anything you would still change in the design? When I say finalizing, I mean that I am compiling prices and sources for my parts, and ready to pull the trigger whenever I am done and move this over to the build-log section, so this is the "final chance" to change anything, so to speak.


As always, here is a web browser accessible version of the model if you want a more detailed look.

[AndyUK]

1. I have seen that people usually use transformers for their steppers, and I understand that that is because regulated PSUs may have problems with the current spikes (at least according to the app notes). I could not find a proper post on how to do it on here though, so If anyone has a good source on that, that would be perfect.

My build thread, post number 30 has an outline of the process I followed, but I emplore you to find multiple independent sources of information and fully understand and verify before playing with a mains level PSU - this is the scary bit of the project to me, because one slip can potentially kill. Always ensure the caps discharge and never work near the PSU whilst alone in the room. I've have given the wife very clear instructions on how to isolate the supplies if it goes wrong. ALWAYS have a bleed resistor - it's not something I mentioned in my thread, but it's a must.

Scary bit over, it's not that hard or expensive. You just need to stay safe.

As for information, I didn't find a definite thread here, probably for good reasons. You're going to have to go research the theory of unregulated PSUs. Joe's videos are also quite good at explaining, and so are ThisOldTony's CNC build videos.



Sent from my SM-G950F using Tapatalk

[Kitwn]

There's a nice guide to designing linear power suplies here http://www.skillbank.co.uk/psu/

Whilst it's true a linear supply of a given rating can deliver significantly higher current peaks than a switch-mode supply of the same rating, you should look at the relative cost of a linear design and a suitably rated switching supply or possibly two separate supplies, one for, say, two motors on X and one for the single motors on each of Y and Z.

I can't see the point of spending a lot of time and money on building an unregulated power supply which can, even briefly, deliver far more current than is required to saturate the motors. The available current in case of a fault would also be much higher.

Kit

[Clive S]

Have you seen Joe Harris build log here is a link to his youtube channel on power supplies https://www.youtube.com/watch?v=4OoQ...Y&index=2&t=0s

[Chrono]

Hey guys! Sorry for the radio silence, I had hoped to be able to post a finished schematic, but mine and the person helping me with that part's schedules just did not meet up yet, so that has not come to fruition yet.

So first up, thanks to all the tips for the PSUs. I think I have a firm grasp on the matter now.

Instead of a schematic, I worked on finalizing the budget and getting everything priced out, and so far, the only thing missing from the budget are timing belts and the control box parts (like contactors, etc.), but so far, everything is savely in my 3.5k€ budget, even with a couple hundred of wiggle room. So, since I have not posted this before, I "present" the semi-finalized list of electronics to be used:

• A smoothstepper ESS. From all the controllers I have looked at, this one seems to have the highest price/performance ratio. Compared to the UC400eth, for example, I get one more port for a similar price point. And the CSlabs are just too expensive for me, in comparison.
• A 3kW watercooled chinese Spindle with ER20 collets, able to reach up to 24 000 RPM with a packaged submersible pump
• A YL-620 VFD for that spindle
• AM882 Stepper drivers for all three steppers, supplied with around 60V
• X-Axis Stepper: 86 HS155-5508-01
• Y-Axis Stepper: ACT 34HS1456
• Z-Axis Stepper: 60BY G401-03
• 6x Inductive Proximity switches to be used as Limit/Homing switches

Other than that, not much has changed on the mechanical front: I have had a two parts re-drawn in a different CAD program by someone well-versed in designing sheet metal parts, but other than that nothing has been altered. I hope to be able to get the schematic done in the short term rather than the long term, so I can finally move this to the build log section ;) Anyways, until I have more news!