The housing for the spindle is in two parts. There’s an outer sleeve (which was too thick by 0.4mm so I had to reprint it), and an inner sleeve made to give the spindle support, but allow lots of airflow around it.
The outer sleeve is 112mm long. The inner sleeve slightly shorter.
Why two sleeves? The outer sleeve stays in the M4 and ensures that it is easy to slide the inner sleeve in and out, without having to jiggle the arms about to get everything back in.
I would consider having the inner and outer shells not just there as spacers, but have the outer shell act as a receiver/socket for what end up being modular cores that you can easily swap out.
Very nice build! Especially the cooling chanels being open on the inside.
I am looking at spindles myself, not sure yet on what to buy; a 800W 3-phase with a frequency drive inverter up to 400Hz (might make a lot of electronic noise?) , or maybe your spindle, which is 500W I think, and I saw only 2 wires for the motor, so not a frequency drive inverter. Can you please share how well the spindle behaves?
Thanks,
The three phase spindles are RADICALLY better than the DC ones. MUCH better bearings. I’ve used both on my CNC router and I’m much happier after the switch to the three phase!
The DC motor ones use kinda sorta bearings and a bit of rubber for a sorta-preload. The three phase actually has real bearings and proper preload springs.
I spent WAY too much time trying to true up the ER on my DC spindle
That is actually the idea. The outer shell has a tiny lip at the bottom to support the inner shell, and I could have made it with a tiny top lip, facing outwards for it to support itself.
Also with the tiny bottom lip you could realistically print simple spacers to hold any inner shell up higher in the assembly, while still providing sufficient support for clamping against
The three phase spindles are RADICALLY better than the DC ones. MUCH better bearings. I’ve used both on my CNC router and I’m much happier after the switch to the three phase!
The DC motor ones use kinda sorta bearings and a bit of rubber for a sorta-preload. The three phase actually has real bearings and proper preload springs.
I spent WAY too much time trying to true up the ER on my DC spindle
but can you get them small enough? all the ones I could find were 80mm diameter,
we need <67mm
I’m looking forward to hearing about this. I was thinking in terms of chopping up some 5/16 or 8mm bolts to make bearings to go between the spindle and the arms and printing new top/bottom clamps.
I wonder if the DC spindle will result in less RFI than the routers do??
If the DC spindle is brushless, it should be far superior. The problem with all AC routers is that they use brushes, and the commutation process generates arcs that in turn generate RFI across the spectrum.
When I worked for a large US manufacturer of robot vacuums, we had to add disk varistors to all the brushes of all the DC brushed motors (5 per robot) in order to pass FCC part 15.
I wonder if it would be worthwhile to do the same for the (AC powered) router? Since the router motor is a universal type motor, I think it should work, but I’m an ME not an EE.
…and I would add, the varistors have to be mounted as close to the source of the arcs. We had the motor manufacturer mount them inside the motor case right across the brush mounts. Some people solder them across the motor terminals on the outside of the motor but I believe this did not work well enough.
If the DC spindle is brushless, it should be far superior. The problem with
all AC routers is that they use brushes, and the commutation process generates
arcs that in turn generate RFI across the spectrum.
these cheap spindles are not brushless (I wish they were, I’ve thought about
building one, but never gotten past doing the research)
I can confirm that the two wire spindle referenced above is a brushed DC motor. You can put a small capacitor across the leads to reduce the EMI. A ferrite ring on the two wires (maybe two turns) could also help.
Here’s what’s inside the DC motor version.
THe other end has a rubber cup around the bearing that is the preload “spring”. It also makes the back end sloppier; I made a custom tube to take the angular slop out, but it wasn’t a big improvement.
My first brushless spindle was only slightly better, with a wave washer “preload”
I’ve printed and installed these new anchor point mounts, with a significant improvement in fitness, from 0.489 to 0.55.
I reckon that’s there’s probably another >2mm total flex (mostly across the width of the frame) for me to deal with. But for now, I’m good with just passing calibration on my modified frame.
