That's exactly right. It's a flat, four-conductor cable, wound a few times around the motor, with a flat metal spring coil wound with it, and the whole thing is covered in grease. The hip is designed with two Go2 motors attached together, and this allows them to rotate relative to each other (it's unfortunate the motor doesn't have a hollow core to pass cables through).

@QuestionableEngineering i was researching the new motors they used on the b-2w model, but there's zero information on the m107 units. I'd assume they're similar to these ones, direct drive with mild gearing.

@@Flyingwigs Yeah, they're oddly inconsistent in what they publish.

@@PORTUGAL-HOBBY-CLUB I have no plans to reassemble mine. Hard to say if you should or not, don’t know what your end goal is!

Not elastic actuators, just geared motors. It's several loops of a flat spring and ribbon cable, covered in grease, to transmit data and power. None of the motors has to rotate more more than maybe 200°, but I guess there’s not really room between these two motors, which are sandwiched together, for a communications cable that can repeatedly twist that much.

@@QuestionableEngineering Ah okay. Thank you for clarifying.

I did a little bit of stair climbing. I found it awkward, in that you have to switch modes to up- or down-climbing, and it struggled a bit. I also played a little with obstacle avoidance disabled, in order to get through a narrow passageway (furniture was too close to a wall for it to be willing to pass, even though there was ample room). I also briefly tried the leash mode, where you can tug on it and it'll go in that direction. I haven't tried any autonomous stuff.

@@QuestionableEngineering thank you for the information! And sorry I just saw this response now. I have one on order, I will experiment with it soon.

Yeah, my repo is a mess, but I'm using the stuff down in `router3`. I need to clean this up: github.com/jetforme/router-table

I took a quick look. It’s a PITA. You'll have to remove the entire sensor assembly, which I do at around 1:08 in this video. Once you've done that, there are a bunch of screws to remove to gain access to the camera board. I didn't go that far, but you'll probably have to remove the LIDAR sensor like I do in the video above to see where all the internal screws are. I actually didn't end up needing to remove the LIDAR for my purposes, but at least you'll be able to see how it comes out.

@@QuestionableEngineering thanks, yes I can see a shitload of screws there. I will go thru the lidar and use some angle side screwdriver instead to fix it in place, only 2 screws this way :)

I did disassemble one partly, but I guess I did that off-stream. If I find some time I'll throw some video up.

Ha! I guess that’ll depend on how well the building was constructed. The robot definitely makes tippy-taps for everything it does.

Not a good idea man, I got the robot and got complaints on the first day I got it, don’t use it in the house, just take it outside to the park, people will like it

I can’t say if it’ll be good for you, but it’s a lot of nice hardware for an amazingly low price. There’s an effort out there to make an open-source OS for it, which I think will be a big boon.

@@QuestionableEngineering im thinking of the go 2 air

@@DadTrim That’s what I got.

@@QuestionableEngineering thanks for your advice

most likely sensors

It’s the LiDAR.

@@QuestionableEngineering oh wow that is high tech thank you

The brake is definitely recommended; without it, the spindle will crash into the bed when e-stopped or powered down. Not shown in this video (but maybe on the z-axis teardown) is a little one-way bearing that tries to take some of the load off the z-axis servo. I took that out, and my servo has to work pretty hard to keep the z-axis from falling when operating. It was overheating and shutting down until I added a fan to it. It is probably undersized for the mass of my spindle. If you’re using simple steppers, you’ll want to oversize it for sure. I do experience ringing, generally seen as ripples in the cut just after 90° turns. 80/20 is not the most rigid structure, but I’m sure I’m pushing reasonable limits with this spindle (and I still want to add a retractable dust shoe!). I could probably stand to lower the acceleration (my table is scary fast), and I’d love to someday put an accelerometer on it and see if it can be finely tuned. The ripples have so far not been a problem, as they’re minor enough to sand out. Off the top of my head I can’t remember which axis shows the ringing. I would expect them to be in Y, as that would be the least rigid one, but I just can’t remember (it has sadly been a while since I’ve used it). I will recommend Teknic SDSK servos over steppers. They are amazing!

@@QuestionableEngineering Thanks for the quick response! I'm going to certainly follow your advice and oversize the stepper/ add the brake. Thanks again!

Yeah, I think it was an Ardupilot Mega with DIYDrones Hotel 1.0 board from 2010. I don't think we had upgraded to the PX4 yet, but we definitely used that on the next wing (which eventually took a dump in the SF Bay, and we bought an inflatable boat to go get it. I should try to find that video 😂).

I did a quick google search on that router. I don't think it will be easy to add a 4th axis. You don't really have enough z-axis clearance, and I don't think your controller can manage a 4th.

Thanks! The spindle is overkill for this router, but it’s a 4.5 kW HSD ES330 questionable-engineering.com/blog/spindle-motor

I don't know exactly what it’s made from. Some kind of fabric lamination. With the red metal covers it forms a dust shield over the assembly. It came from AvidCNC, so you might be able to message them about it; they’re usually very responsive.

It's got a fair bit of flex compared to a welded tubular steel frame, so I can't really push them to their limit. I reduced the accelerations a fair bit in LinuxCNC.

@@QuestionableEngineering How rated power are those servo motors you used?

I love this spindle!

I can try to post it somewhere but I’m currently out of the country and won’t be back till the 18th. Also, it’s specific to the Chinese tool holders I got, so it may not work for you.

@@QuestionableEngineering what you already have is perfect, much better than starting from scratch

If you have Fusion 360, you can download the CAD file here: share.getcloudapp.com/Jrun5PAZ If not, lemme know and I'll try to export 3MF or STL, but I *highly* recommend working with the Fusion file.

I tried to strike a balance between allowing tool clearance over the tops of stored tools, and giving enough room for the bristles. In my case, the red and black blocks are 45 mm tall. The could probably be a little shorter, since the only real interference with the bristles is the 8020 crossbeam.

Thank you. That’s a fair request, and I did intend to write a companion article with more details. I’m swamped at work for the next couple of weeks but I’ll try to get one posted after things cool down.

No, I had that shim stock separately. You can buy it in varying thicknesses and materials from McMaster.

I replied via email but here's the link: kzbin.info/www/bejne/apS2c2iuar9meLM

I believe that the stepper motor would then have to over come spring tension thus negating the benefit causing wear and tear along that axis and should spring or something go out of alignment that spring could help push or further hinder that alignment. Plus that spring would also be pushing up or outward aka away from the spoil board putting untold tension in that direction too. I could be wrong. Just my opinion having spent many years repairing equipment with tension springs and adjusters with idlers and so forth. That old say for every action there is a reaction. Now if the system was designed with the tension system built in that is a different story altogether. You maybe onto something.

I'm sorry I missed this comment! It would be nice to have something to offset the weight of the z-axis, and that was the purpose of their semi-one-way bushing. I'm using a Teknic SDSK NEMA 34 servo, but it was getting hot and hitting its thermal cutout until I added a fan to it. It could probably stand to be upgraded to a larger servo.

For a 4x9’ gantry machine, that’s asking a lot. You would need a different design. This one was designed and built by AvidCNC.

@@QuestionableEngineering Did you measured actual displacement of this design? I am just curious how much it is in lower/upper Z position and in the middle of the gantry

@@L2Max1 No, I don't have a good way to measure it. But there is definitely flex in the gantry.

Thanks for reply

@@QuestionableEngineering Could that be for metal fatigue? Sagging ? Both steel and aluminum have a tendency to bow downward when stretched in the horizontal I.E. as in a suspension Bridge for example basically that is what the gantry is doing suspending or bridging the distance across. It can also happen in a vertical orientation also. There would be some deflection or flex and one possible way to measure would be from the carriage home position , then far left far right or two equal points in between to spoil board and compare the distance to a set height of say the spindle collect in the tightened position. Being that you have a true flat surfaced spoil board. or a level that could bridge width of unit or length of unit. Just some random thoughts if deflection was considered to be an issue.

Thank you!

It's a “stud-driven hole punch”: www.mcmaster.com/hole-punches/metric-stud-driven-round-hole-punches-for-conduit/

I wish it were clear to me…

It took me a long time to get a brake because I could only find expensive ones with long lead times. Then Tormach offered a new machine with a brake, and I asked them what part number it was. Ended up getting it from Automation 4 Less, $272.73: www.automation4less.com/store/proddetail.asp?prod=MPC034-24-0500

Added a link to the description.