Problem: BlackBox goes idle (stops) in the middle of the work and few other questions

Have you checked your power settings and made the recommended changes: OpenBuilds BlackBox 4X Documentation

 

where are your home switches?
industry standard is positive ends of all axes. in this case offsets to avoid hitting them have to be negative.
if your switches are negative ends of travel, then your machinehomeX/machienhomeY woudl have to be positive to avoid hitting the limit.

 

this depends on where you put your part zero in Fusion. if it is in the middle of the part then some moves will be negative.


why are you starting a job from machine home anyway? that is very unusual.
normally one would mount the work on the table then jog the tool to the WCS point and use the appropriate 'setzero' button to set work zero at that point, uually X and Y first then Z with the tool at the work surface,

depends where your home switches are, as previously indicated. 0 will probably still crash.

you need to select WCS 1 in Fusion
right click the setup
select edit
select the 'post process' tab

take the collet out and measure it. and if the box says 6 to 8 then there should be a 6 and an 8, and possibly a 7 in the box. a 6mm will expand to hold a 1.4" tool, and a 1/4" collet will compress to hold a 6mm, but an 8 will never compress far enough to hold a 6mm.

watch more video! (-: (seriously,. watch all the FusionFriday videos, you will be pleased you did)

 

little bit = half the endmill diameter so the cutting edge lands on 0,0...?
(it moves along the edge of the cutter, not the centerline, else parts would be incorrectly sized)

 

v1.0.235 should be coming to you in an online update in the next couple minutes - I've made some changes to the code, test and let me know if that took care of your IDLE problem (or if it did not, please also let me know)

 

Use a better, shorter, USB cable.
move the USB cable away from all power cables, that is, AC cables, from PSU to blackbox, and especially from blackbox to motors.
if they must cross, then they must cross at 90 degrees.

if you are using homing switches, then unplug them, disable homing in the GRBL setup, and 'fake the home' (search this forum for how to do that)
in fact do this first, it is very easy to do and if it then works you know the source of the EMI.

 

That shows corrupted serial data

100% EMI the cause. Try a different USB cable, and post some pictures of your bench setup, wiring etc so we can check for sources of EMI you could address, but basically, it boils down to spacing motor, spindle, power cables far from low voltage cables

 

I say this because home switches are the biggest route of electrical noise (EMI) ingress. When a user has persistent EMI disconnecting and disabling homing will prove either that it is the home switch wiring to blame, or that it is the USB cable. ie if you still get interference with the home switches disconnected then the problem is the USB cable.

the why is industry standards which result from safety experiences over many years . The standard is 'machine home at positive ends of travel'.
For Z this is obvious, it is high up away from the work and clamps etc, and also furthest away from the operator, so is a safe place to 'park' while doing other moves.
For machines with tool changers, Z has to raise up to change the tool and the exact position of the tool changer is relative to the high Z home position.

The X and Y positive ends come mostly from moving table milling machines. Moving to home at positive ends means the table comes left and toward the operator, which makes changing out parts easy/possible. Remember the global picture here, some of these machines are big, and the parts are loaded with cranes or forklifts. In the early days every manufacturer did their own thing but operators got hurt when machine B homed in a different direction to machine A and so this changed to a standard so humans don't get hurt AND the machine is still convenient to operate. Imagine borrowing a friends car and the brake and accelerator pedals are reversed! Yep, early cars had every possible variation of control positioning, but this caused accidents so standards were put in place.

Early machines had very primitive controllers. Paper tape was fed in with the program and there was no cleverness at all, everything was an absolute position from home, no work offset!
This is obviously a royal pain to program and very quickly work offsets were introduced. This means the part zero can be anywhere on the table and the programming is 100 times easier and less error prone. This saves time. time is money. (-:

So having a work coordinate system is great, but what if we have more than 1? We have six in GRBL. This means we can put 6 vices on the table, load 6 bits of material, set 6 offsets, then run the exact same program 6 times to get 6 parts before having to unload/load material. enormous time saving, and with growing automation it means machines can run thousands of parts on multifaceted fixtures without a human anywhere near them.

BTW: A benefit of having part Z zero being the top of the material is that when you read the code any negative Z movement is a CUTTING move.

ok, so we have a standard that can be relied on, and programmers being lazy buggers (I know , I am one) will quickly RELY on that standard and programs like Fusion360 will assume adherance to the standard and generate Gcode from that starting point. Yes it has a million settings, but Z home = high is non-negotiable, it assumes this and counts on it in the way it does safety moves.

Now your machine with home at front left.... this means that any parking coordinate offsets must be positive to prevent triggering the switches. but the Fusion post defaults to -10 because of the standards. I would love to see a flyer included in every machine bundle that explains this simple thing you need to know - when the machine build instructions place the switches in a non standard manner, the user needs to know that they need to set the home offsets positive and why, and how to do it for the common CAM software we see here.
Put the switches the other side on X and Y and you don't need to know this, it will all 'just work'.

that is because machine home was not Z high and you have not set work coordinates (ie machine home was work home, not a good idea).
the software assumes Z home is high and is also NOT work zero, because that is the standard.

Why I can't just home my machine with "auto home" and switches. Reset to zero. Then jog my machine into work position, use paper sheet with Z axis, get the same stock point that I set in my CAM, reset to zero again and just use the machine as it is?
[/quote]

of course you can, but you must first have a machine that operates reliably. removing the home switches from the equation simplifies finding and fixing the EMI problem.
home switch wiring must be done correctly with the right shielded cables.

• the shields must be connected correctly (star grounds)
• the wire must be routed correctly (some cables emit so much noise the shield fails, so those cables must run far apart)
• EMI filters must be used. always. (Blackbox has this, and the openbuilds switches also have it, so both ends are protected)

I have been running my OX for some years now and have NEVER had an EMI issue. it has no home switches, the router power cord runs far away from the motor power wires (motor power off the back, router power off the front). I just plugged it in like this and it has worked ever since. I have purchased cable chains and nice 4 core wire with the aim of rewiring it all neatly. someday... meanwhile, I can cut parts without worry. I use the fake home method and a 'goto MCS' macro before I shut down and it all 'just works'. I am in South Africa and power outs are quite common, so I often have to 'home' by pushing the gantry against the stops and resetting the controller. Takes about 6 seconds (-:

I believe that there is a certain amount of complexity that is essential and unavoidable.
Running without home switches is just as simple as running with them, if you are told upfront that you must 'turn on with Z high and X,Y always in the same place, preferably right/rear'.
And if the home switches are not in the standard place then you need to know that and what the sideeffects are. After you know those few things, it all just works.

Here are the basics of Gcodes.


Also watch Joe Piezynski, NYCCNC 'Fusion Fridays', Grimsmo Knives. For more advanced stuff Peter Stanton does some excellent videos.
Yes I know those guys are not running machines like yours, but knowledge is power. I know what I know because I watched videos for years before even building a machine.

this guy is also useful

 

Sounds like bad
Spindle: Yeah I like suspending mains from the ceiling - definately don't run it in the cable chain

Best practice: - a quick google for "EMI Wiring best practise" sure, but its on the one hand pretty logical, and on the other a bit of black magic. All wires generate EMI, but the heavier the amount of current and how more fluctuating the power through that wire, the worse it "broadcasts" the interference. So the spindle, power and motor cabling are the usual worser offenders.
But other nearby sources can be the culprit too: Microwave oven in the shop? 800w 2.4ghz radio transmitter right there, etc

Play with the setup a bit, earth the frame, seperate low voltage signal cables away from bad ones a bit, try a different USB cable, etc. No easy "ah thats the one" suggestion we can make, each guy's setup is just a bit different.
Play the elimination game - spindle off, and it still happens, great (still remove the spindle power, even if something else is the cause, it will eventually be the cause if you keep it in the cable chain) - so it's not the spindle (yet) - lets try moving the motor cables, see if it improves, etc etc.

Have even seen cases where the PC was picking up EMI via a keyboard, mouse, webcam cable

Until it stops corrupting data - you'll see the <Idle\Run...> messages never appear under normal circumstance (the parser catches them and parsers them to update DRO, etc), as long as there's "garbage" in the serial log, the data flowing between PC and controller is getting corrupted

One fix is to buy some clip on ferrite cores and clip them around USB cable on both ends, and a good quality shielded USB cable with ferrites on both ends, will help too

 

yes, but plastic will make the bits blunt quite quickly so maybe you want to keep a set for wood for good finishes

in plastics conventional often gives a better finish. in insulation foams and foamrubber conventional is the only thing that works properly.

 

Where did you buy them from?

 

Possibly the case, that's why I asked. Unofficial sources lots of times are out of tolerance. QA Processes (;
You could slightly drill out the holes in the plates where the solid wheels go, and sorta "nudge" them closer to the middle before tightening, but if you can get the seller to sort it out (plates or beams out of spec) it would be better