Kb58's granite CNC router build

[kb58]

Collecting parts for the 230V portion of the build: input disconnect box, fuses, breaker, contactor, on/off buttons, and VFD. The on-hand parts were moved around in a couple possible arrangements and while these two work from an electrical and cooling standpoint, both looks kinda clunky. Meh, I'll play musical chairs with it a bit longer, but it's probably going to be one of these for practical reasons.

 

[kb58]

Looks like the 220V parts will fit fine (though snug) in the box - one less thing to mess with.

Spent some time trying to find the best location for the Y-axis limit switches. I felt like a dope when it finally dawned on me that the "obvious" place is mounted to the stiffening rail - I won't say how long that took. Anyway, about the time I finished these awesome limit switch assemblies, Voice in the Head asked: "Why didn't you make them on the 3D printer?" Good question - and I have few excuses. I haven't used it in over a year and don't remember the workflow (my fault for not writing it down). Another excuse is being unsure if something produced via 3D printer will be robust enough, specifically, will the plastic "cold flow", resulting in a loose part that shifts, which could be very bad for limit switches. I'll go with that last one.

Picked a spot for one drag chain. While functional, the ugly aluminum block will be pretty'd-up later. This part, too, could be made on the 3D printer, but again, I'm unsure how the material acts under constant load, be it mounting screws for limits switches, or constant torque from a heavy wire harness.

 

[kb58]

All limit switches are now installed, though not yet wired. Yes, the X-axis switches should be oriented in "drive by" and not compression, I'll work on that. The switches were done now because they're an annoying little time-sink but critical to have functional right from Day 1.

 

[kb58]

Messy, but it actually works... that's the good news. The bump in the road is that the 240V power in the garage needs to be redone. When I first started treating the garage as a workshop, there was just a 240V air compressor, so one 240V outlet was fine. Over time, however, that single 240V outlet ended up powering a lathe, band saw, a mill, and soon, a router. In the future, there may be a belt grinder as well. It works because there's never more than one running at a time, though that may change if a power draw bar is added to the mill, requiring air pressure while the mill is being used. The same is true with the router if air-blow is needed.

Other reasons to update the wiring: the 240V outlets have no neutral wire, so it's not possible to run something on 240V if it also needs 120V. So, it's time to place outlets near each 240V machine, and a couple extra. Coincidentally, we're having our entry panel ("fuse box", "breaker panel", "service panel") replaced this coming week. Why? Several reasons. The manufacturer went out of business years ago, making replacement breakers expensive - they aren't form/fit/function compatible. The existing panel is too small; all slots are filled and a few more are needed (and 240V circuits consume two slots). More current isn't needed, just more breaker slots. Lastly, the manufacturer went out of business because - get this - "sometimes" when a short occurs, their breakers don't open! Nice, huh?

 

[kb58]

Update on the service panel upgrade: short story, it turned out... okay.

Long story:
I know that this is going to rile up Chris (

), but my/his expectation of disappointment was justified once again. Three electricians showed up, one fairly sharp, almost making up for the other two. One smoked - constantly. It just amazes me with everything we know about medicine these days, and no matter what you tell people, some still smoke. Anyway, one of them, as he was disconnecting everything, was telling me how many times he'd been shocked... Ugh. Watching them work (I let them know that I"m genuinely interested in how trades people do things and didn't pester them), I was reminded of something I saw when on a business trip to Japan. There, in a factory, I watched a technician removing an access panel. It slipped, touched the (very clean) floor, and he immediately stopped what he was doing and carefully wiped it down. I never forgot that, the sense of care in workmanship. Fast forward to this job, and I watched them - several times - drop brand new portions of the replacement panel in the dirt, and either leave them there, or pick them up and install with without wiping them off. That said a lot - a lack of care.

When nearly done, they asked me to check in the house to make sure all the lights and outlets worked. I did found that one outlet and one light didn't work. Instead of examining his work, he said "maybe the bulb is burnt out." That annoyed me because it suggested that he was looking for an out so that he could leave. Um, no, I said, it's got five bulbs in it, and the outlet is dead. He said, maybe it's a bad GFI... again, nice try, but the outlet is not GFI-protected, and yes, he eventually found that he'd missed swapping over one wire...

After they left, I examined the panel and saw what's shown below. The 2-pole breaker labeled "AC" at upper left is offset and could not be seated. Looking closer, they had installed a breaker of the wrong type! It physically couldn't be installed correctly. That lack of care continued because the cover panel could not be properly installed over this breaker, but they forced it anyway, bowing it outward. I supplied these pictures to the company asking that they correct the situation. Yes, I could have fixed it myself, but having paid a lot to have it done "right", sorry guys, you have to do your jobs. What gets me riles up is when I'm paying a company a lot of money and have to manage them. I feel like saying "Okay, since I'm managing this, my rate is $100 per hour until it's done correctly." I suppose I could say that, but then you end up having ******-off people working on your stuff - not an optimum situation.

Anyway, after that, I finished wiring the new 240V outlets, then started modifying both the router controller box and main AC control boxes to accept three wire primary power (120V/neutral/120).

Oh, and one last thing: I learned something new, I'm just not sure what. That is, in the US at least, if you buy, for example, "12/3" Romex house wiring, you get a cable that has a black wire, a white wire, a red wire, and a ground. This is of course four wires, but the industry assumes that there's always a ground wire, so it's not counted. Okay, so yesterday I walked in to buy power cord, to run from the outlet to the router, and asked for "12/3" cord. The guy cut it to length and I was halfway to my truck when I looked at the end closely, and saw three wires, and only three wires. I took it back and confirmed that he'd given me "12/3" cord (one black wire, one white wire, and one green wire). It turns out that - apparently - house wiring cable and power cord are referenced differently. I had to buy "12/4" cord to get what I wanted: black, white, red, and green.

 

[kb58]

I hate doing stuff twice, but have no one to blame...

I put many of these same pics on a CNC site and am getting various comments - some good, some "meh." I mentioned having a 240V spindle yet running the control box on 120V. Someone said "why aren't you running everything on 240V"... and I thought "well because not everything will run on it." Turns out he was right. The servo power supplies and the small MeanWell supplies both do, and even the wall adapter for the small PC that runs everything. I'm not going there with that, just because. Anyway, since I'd all but finished the AC side of both the control box and the second control box running the spindle, everything got disconnected and redone. Not hard, but unlike with 120V, I feel extra, mmm, careful, when wiring 240V, as any flash and smoke would no doubt be twice as exciting - and damaging. I did put several 120V outlets on the bottom of the box though, as there's always something extra that'll be needed. One outlet is being run through a relay to control a pump for the spindle cooling water.

 

[kb58]

After the drag chains were mounted, all the cable lengths were added up, then the wire ordered. One annoying thing is a company supplying connectorized cables for their product but make them too short. Only one of the four servos will be able to make use of the factory 10-ft cable. My router isn't that large, so I suspect most customers quietly fume about this but don't speak up. If the factory made them, say, 20 feet long, they're easy to shorten, but much more annoying to lengthen.

 

[kb58]

Mounted the controller box and AC-related components. The control box extends out about the same distance as the X-axis servo, so it doesn't grow the equipment envelope or "walk-around" space. it was mounted high for practical reasons: allows shorter wires, it's easier to reach the power switch, and there'll no doubt be many fun hours of sitting on a stool making wiring changes. Oh, and it provides a convenient place to set a coffee cup - where it may end up collecting chips...

Wires were then pulled through the drag chains. I have to smile any time a part originally bought for Kimini or Midlana gets a second life in the router. Today it was several small "P-clamps" used to route the limit switch cables. Silicon coolant hose was also run, and before it had been handled a bit, seemed to stick to everything...

The spindle power cable was found damaged as-received and unfortunately, it's right in the middle of the run. Ordinarily I'd fix it, but because it's going to be flexed about a million times, and because there's no rush, it'll be replaced.

The small white cables are to the travel switches. All the "Home" switches are wired in series, as are all the "Limit" switches. The controller figures out which axis is moving based on which servo is enabled. (Don't ask me how it figures out which is moving if two servo axis approach limit switches at the same time.) Because they're in series, wiring is simplified somewhat by having a terminal block up on the side of the router, where all the cables terminate, and then only two cables from there run to the controller.

 

[kb58]

Wired the limit/home switches in series and tested them. Also wired up most of the wiring for the AC panel - the controller will operate it but there'll no doubt be a bunch of VFD configurations to set first. The harness at the left end of the gantry remains open for now, awaiting the replacement spindle power cable. Left over mesh harness wrap (from Midlana) is used to clean up the appearance and better protect the wires. There's still plenty to do, including extending all but one of the servo's power and control cables - not hard but tedious. Notice that the fan outlet on the VFD is right where a bunch of chips will land. A divider panel will be placed above it, though it touches on a whole 'nother side project, figuring out how to protect the rails, ballscrews... and the whole garage for that matter, from flying chips.

After everything is wired comes the scary step of switching on main power for the first time

Then there's calibrating the dual Y-axis servos, which isn't straightforward because they're share the load and can't be calibrated simultaneously. The manufacturer offered to help virtually and I may take them up on that since I don't know how to do it myself*. Then there's about a million things to set or scale before running any G-code on it.

* YT personality Clough42 used a process to calibration a very similar setup, but his X and Y axis used the same ballscrew pitch. I can't use his method because the Y axis ballscrew pitch is 5 mm, while the X axis is 10 mm. I ordered the ball screw shaft before realizing this limitation. Oh well...

 

[kb58]

In this case, the red lights are a good sign. The AC-side was checked out and then switched on - fire extinguisher in hand, though with everything inside running on 240V unplugged. After confirming that the voltage was only where it was supposed to be, everything was connected. The VFD and spindle are wired enough that it could be powered on as well, confirming that it spins - the wrong way it turns out, hah.

Now starts the next phase, configuring the servos to match the setup, followed by configuring the VFD to be controlled from the Acorn board. Somewhere in there will be calibrating the servos and as mentioned, having the factory assist with the two Y-axis drives.