Workpiece and spoil board combined, I think- often more than 20mm.
I was assuming that you calibrate on the spoil board, so it’s height is already
included (and is pretty much a fixed value)
so I think the question for the user is ‘how much extra height is there for this
cut setup’
When it comes to calculating the Z offset’s effect on the belt length, if we’re not going to use as accurate an offset as possible for everything under the sled, what difference does it make if we accurately measure the Z carriage height above the sled?
An alternative approach would be to create some standard spacers that could fit around each lead screw. Since they’d be a known height the calibration could be to raise the Z to the top limit (or close), insert the spacers (which are a known size) and let the machine calibrate by lowering onto them. This would be pretty foolproof and not require driving the installed bit below the bottom of the sled.
An alternative approach would be to create some standard spacers that could
fit around each lead screw. Since they’d be a known height the calibration
could be to raise the Z to the top limit (or close), insert the spacers (which
are a known size) and let the machine calibrate by lowering onto them. This
would be pretty foolproof and not require driving the installed bit below the
bottom of the sled.
Yes, this should be supported as well, but it’s the same method, except instead
of setting Z=0 at that point, you set it to some other value.
@Bob_Craig - much better idea than mine. Tall enough to clear a bit, and by putting them on the lead screws, you could square the two steppers against them.
Why not just loosen the belts and lift the sled when you need to home with a bit installed? Afterwards you can set your z=0 with a conductive plate and you should be good to go… Am I missing something?
I agree you probably need to enter your stock thickness somewhere in that process (assuming you calibrated on the spoilboard).
I’m imagining that homing the z-axis would happen before even extending the belts so it should be possible to lift the machine up to do it with a router bit in place. Just not sure how much of a hassle that would be
Depending on shop layout, though, it may not be easy to be at the UI controlling the process and also lifting the sled during the Z-axis homing. I don’t like the idea of driving the bit into the work surface if I’m not able to lift it off during the process (both to protect the surface and the bit).
Depending on shop layout, though, it may not be easy to be at the UI controlling the process and also lifting the sled during the Z-axis homing. I don’t like the idea of driving the bit into the work surface if I’m not able to lift it off during the process (both to protect the surface and the bit).
This is where a browser based system shines, you can just connect from a phone
for this sort of thing while still using a real computer for most control.
I got the basics of being able to save and load values from the NVS (non-volatile storage) system working along with being able to update the values within fluidNC (which takes some finagling because we need to synchronize the stepper drivers, the gcode planner, and the machine position with the new values).
More testing is needed to see if it’s stable, but I think it should be totally possible to preserve values through power cycling
I got the basics of being able to save and load values from the NVS (non-volatile storage) system working along with being able to update the values within fluidNC (which takes some finagling because we need to synchronize the stepper drivers, the gcode planner, and the machine position with the new values).
More testing is needed to see if it’s stable, but I think it should be totally possible to preserve values through power cycling
when you save them, don’t just save the length in mm, also save the encoder
angle.
If the sensor angle is very close to what it was before (within the error of the
sensors, which I belive is ± 1 degree), and the belts are tight (I think we
would have to pull to know), then we can say that we probably are at the same
position we were before powering off.
If the encoder angle is incorrect at startup, we have one of two situations.
if we are within ±1/2 sensor rotation, the spools haven’t moved, this is
just slack belts, and we can adjust accordingly.
if we are off by more than ±1/2 rotation, we can’t recover and have to fully
retract.
unfortunantly, we have no way of knowing which situation we are in, so we will
have to ask the user.
I’m running into a situation where I would like to pause a job while cutting and lower the z home position like .02” to more consistently cut through material that turns out to be irregular or when surface is building up chips which raise z. I thought I could do this on old Maslow. Can we do this on 4?
Sequence something like
Pause job.
Lift z to zero and have a look.
Move z to desired new height.
Define z home
Lower to previous cut path depth(plus new depth)
Hit play and let job keep running.
This would be handy. I believe it should work as you describe. Just remember to shut the router off and turn it back on in there somewhere.
The only issue I see is that if the idea of Z-zero is the tip of the bit is flush with the bottom of the sled, if you need the bit position adjusted lower, you need to have the bit sticking out below the sled. I believe it currently lifts up a small amount before moving so it’s above the work surface. That might be a feature of the CAM software you use as well. I feel like with the jobs we were running where the plunge depth was 9.7 mm the retract amount was some odd mm value i.e. not 5mm even, and I meant to look up if that was an inch value translated to metric somewhere.
Anyway, if your z-adjustment leaves the bit below the sled then it will hang up when moving, but my experience has been that there would be a few mm to play with as long as your Gcode is doing something reasonable like retracting slightly above zero during moves.
I don’t know my CNC but I think what I’m talking about is referenced below:
I don’t believe it will. I think pausing the gcode sends a ‘!’ which puts fluidnc into “Hold” state which if I’m reading right, also disables jogging. I could be wrong on that though. seems like it would be simple to test with a small gcode file and no bit though. If I was set up right now I would do it, but I’m not near a maslow right now.
Yeah I can’t test because I’m midway through a complex cut job and it’s not getting all the way through. I think my bit might have slipped. Anyway. We all need a bag of tricks to get through jobs when things go south.
Can you mess with the z without disrupting gcode using the fluidnc window?
I don’t want to throw hardware at this( yes I really do ) but on my 3040CNC with no endstops, I just know the upper limit and never really get close anyways but the real useful part is the zero probe.
Back to the Maslow4, I’m just getting started and noticed the calibration asks to bring it to bottom and I made sure to do that a couple of times just in case I was only hearing one motor grind. ie wanted to be sure both motors were in step. For my own understanding I set the movement to 10mm steps and counted how many until it hit the top. I recall it was 7 or 8 so I know for sure I’ve got 60mm of vertical travel.
But if there is a free IO pin being able to plug in a probe, place a clip on the tool then lower the tool until it hits the calibrated disc on the work piece then know your z=0 is your current height minus your probe disc height. It’s what comes with the cheap Chinese 3040tCNC machines.
And if there is an IO pin for a probe, it could also double as an upper endstop input when cutting since you only use the probe input before every tool change.
BTW, Mitch Bradley over at FluidNC maintains CNCjs and is where I got to know how to use my 3040tCNC and the probe capabilities.
