So you disagree with the @AlexN Belt Stretch and Stepper Holding Measured post findings and the wisdom of using less stretchy GT3 or Kevlar belts?
1 Like
Nope,
That post was most informative and what set me off doing the measurements.
What Iāve found though is that the V wheels are responsible for considerably more unwanted deflection than the belts are. There is a backlash mechanism in the machine, but itās not just the belts.
Bigger better belts provide a number of benefits, less deflection on dynamic loads when cutting, less chatter etc. but that wonāt necessarily have any effect on final accuracy in almost unloaded finishing passes.
The downside of the bigger belts is that they move the load further away from the stepper motor front bearing and increase the radial load on the stepper shaft which means you need to be even more careful to not gorilla tighten your belts trying to reduce backlash or deflection by winding up the static tension.
Iāve got a bunch of stuff to write up and post, which I need to do whilst I can still take V-Wheel measurements.
7 Likes
Awesome - looking forward to that! 
3 Likes
In the meantime, the Shapeoko Pro upgrade to the slightly wider belts and the MGN rails is, in my opinion, a really good choice based on price, size and utility. Itās close to what Iām doing based on the deflection measurements Iāve taken. The inclusion of a few other things like the flipped motors, wider X beam, Y rails outside the Y beams and the proper T slot bed make a really nice package which I expect will work really well for Shapeoko sort of jobs.
People who want to chase maximum MRR on a Shapeoko with a 2kW spindle will probably still want to mod it more, but even then, itās probably a better place to start.
2 Likes
It will also not surprise you that Iām in favor of the double drag chains 
1 Like
I pray they mount the electronic enclosure on the side rail of the smaller machines as well and have drag chains. I really dislike the x gantry mounting spot on the standard so3 as I dont want my power and usb cables dragging along the machine edge. Plus itās a pain to spin the machine around to access the control board in my enclosure.
wishful thinking.
1 Like
It wouldnāt be a large modification project to buy a Ā£10 drag chain on eBay or AliExpress and fit it to your machine with some 3D printed or sheet Aluminium brackets to carry the USB and DC power cable up to the controller.
Just a thought.
1 Like
Except that the wiring isnāt long enough to reach a different position ā in order to modify the placement of the electronics you usually need:
- Molex KK pins and connectors
- crimping too
- 4 conductor 18ā20 gauge shielded wiring ā ideally a sort suited to bending
2 Likes
Is that the DC power cable thatās short then?
I wasnāt thinking of moving the controller, just managing the incoming USB and power leads.
OIC.
For the USB and power leads a pair of drag chains at either side of the machine makes a lot of sense if one is having issues with their runs.
1 Like
Actually it does, since spindle motor E/M coupling could be virtually eliminated by routing its wiring perpendicular to the other wiring and keeping it as far as possible from the electronics. I.E. by routing it alongside the dust collector hose. 
2 Likes
Wellā¦
Firstly, Iām already guilty of the double drag chain to keep the spindle wiring and plumbing separate;
Secondly, in the double chain, facing in opposite directions install, a large chunk of the chain length is not running close and parallel, even where it is we have the inverse square law in our favour so even just a couple of inches gives us a huge drop-off in EMI.
If we were really concerned to the extent of fully diverse routing then the real fix would be to use properly screened cables instead, as already done for most VFD spindle power cables, a simple vertical plate between the two drag chains on the X rail would cure any cross-contamination concerns.
If I were fighting VFD noise pickup in my signal and motor cabling, Iād just use screened cables for those or wrap the whole bundle in the drag chain in a copper braid and ground it.
2 Likes
Screened cables suppress the radiated electric field component but not the magnetic field component. But, twisting the power conductors would as would enclosing them in a magnetic shield. It seems to me that since its basically a low impedance system, the magnetic field component would far exceed the electric field component - right?
This is a reasonable summary of the noise source behaviours;
(Although itās a bit of a cable advert)
There is a shed-load (new colloquialism for Julien) of electrical noise coming off the VFD or nasty commutated universal motor wire thanks to the high frequency interruptions of either the VFD chopping or the commutator brushes. Whilst there is some magnetic field from the power cable, thereās just not that much current being delivered so itās comparatively small.
As I understand it weāre much more concerned with the low voltage, low current cabling as this is most susceptible. There can be issues with EM pickup on things like stepper power cables but this needs to be several orders of magnitude worse to show up, basically injecting noise back into the drive electronics. In most cases suppression at source is preferred (e.g. easier to stop a tool using Twitter than stop all the followers reading).
For the Shapeoko wiring;
- The wires to the old homing switches are twisted, which is good as these are signal level. Iām not sure about the new proximity switches (still not in the shop, grumble, grumble), hopefully lessons were learned and those are TP and shielded. For the old microswitches I considered going screened by swapping the wiring for some CAT6 Ethernet STP as itās dirt cheap but I donāt have an interference problem to fix so left them alone
- The wiring to the steppers is power delivery and a source of EMI itself, the current flows should be sufficient to outweigh any pickup, if not then you should turn the plasma cutter head off and think carefully about your upgrade urges
2 Likes
I guess the lack of output filter inductors on the VFDs would cause an increase in high frequency impedance at the motor because of the motor inductance. But, IMO it would be better to add filters to the output (and input) of the VFDs to prevent the E-field radiation rather than try to suppress it with stiff and expensive shielded cables.
Good point about the universal motor routers. Their untwisted and unshielded power cables would carry more current than higher voltage HF spindles for the same power draw. Also both their PWM drivers and brushes would generate E-Fields. So, it seems prudent to run their cables perpendicular to and away from others and to keep sensitive things away from the router and/or properly shielded.
BLDC routers run at even lower voltage and proportionally higher currents for the same power output. So, theyāre likely to generate higher M-fields and lower E-fields on the cable between the motor and its driver.
An effective KISS (I think @Julien already knows that one) solution for all types is to keep things well separated with perpendicular spindle cable routing.
2 Likes
That was what I hoped too, but unfortunately not, according to the manufacturers (the competent ones) output filtering the VFD reduces efficiency substantially. They went to faster switching electronics to reduce the losses in the system but these are the source of the really nasty high frequency noise, especially the reflection of the voltage transient from the big impedance change when it hits the motor inductance at the end of the cable.
I think the better manufacturers do pay some attention and take the really high frequency stuff off at the drive but mostly they just seem to say āfilter the inputā, āwire it up properlyā and āRTFMā along with ātry to keep the wiring shortā.
Anyway, back to the separation discussion, yes, orthogonal would help, but you need to have enough pick up to start with, a user can easily route the router cable down the vac hose but thatās not a part of the machine Carbide supplies so the secondary drag chain is a big improvement over ācable tie the router power cable to the outside of the signal wire drag chainā which I was never a big fan of.
1 Like
Well, it looks like Inventables has lost the plot. They just announced their Xcarve Pro for a whopping $6,495ā¦ (actually $11,995 without discount) What a joke!
Shapeoko Pro is a way better value per $.
On the flipside, I hope Carbide sees the opportunity to offer a 4x4 Shapeoko Pro option.
6 Likes
Hmm,
It looks nicely and expensively finished but at that price you could be buying an Avid and really go big. Although I agree that being able to feed 1220mm / 4ā stock would be very handy.
At that price I would certainly expect more than a measly ER11 collet in the 2HP VFD spindle too, I like being able to run grown up bits in my ER20 collet, it was a major reason for the upgrade.
4 Likes
Carbide can definitely crush the competition. Looking forward to seeing it unfold. During the Inventables live stream live chat I told people about the Shapeoko Pro.
Please 4x4! 
2 Likes
They probably should have went away from the āX-Carveā name with this announcement of the Pro. They are completely different machines, different leagues for different markets. The Shapeoko āProā doesnāt even match IMO. We are talking about 25mm HG series profile rails and ball screws vs 15mm MGN series X/Y, HG Z and 15mm belts with Z leadscrew, 48in vs ~30in work area. Completely different leagues, use cases and customers in mind. Honestly the price isnāt completely unreasonable, though the pre-order price is a much more aggressive and would benefit them better, especially with that spindle choice (seems limiting compared to everything else).
4 Likes