Brian Popp published a new build:

Read more about this build...

 

Anybody have any ideas as to why my rig is moving so slowly?



Max feed rate ($110/$111) is set to 1000mm/min. I didn't see anything limiting speeds in the gcode, but as you can see at the end, even non-cutting movement is slow. Amp draw is less than half an amp at 24v. I realized as I was writing this that I haven't really tweaked the amperage on the drivers (A4988).

 

I'm using stock grbl firmware, and my settings are as follows. Really the only thing I've changed are my steps ($100-102) and the max feed rates ($110/111):

>>> $$
$0 = 10 (Step pulse time, microseconds)
$1 = 25 (Step idle delay, milliseconds)
$2 = 0 (Step pulse invert, mask)
$3 = 0 (Step direction invert, mask)
$4 = 0 (Invert step enable pin, boolean)
$5 = 0 (Invert limit pins, boolean)
$6 = 0 (Invert probe pin, boolean)
$10 = 1 (Status report options, mask)
$11 = 0.010 (Junction deviation, millimeters)
$12 = 0.002 (Arc tolerance, millimeters)
$13 = 0 (Report in inches, boolean)
$20 = 0 (Soft limits enable, boolean)
$21 = 0 (Hard limits enable, boolean)
$22 = 0 (Homing cycle enable, boolean)
$23 = 0 (Homing direction invert, mask)
$24 = 25.000 (Homing locate feed rate, mm/min)
$25 = 500.000 (Homing search seek rate, mm/min)
$26 = 250 (Homing switch debounce delay, milliseconds)
$27 = 1.000 (Homing switch pull-off distance, millimeters)
$30 = 1000 (Maximum spindle speed, RPM)
$31 = 0 (Minimum spindle speed, RPM)
$32 = 0 (Laser-mode enable, boolean)
$100 = 400.000 (X-axis travel resolution, step/mm)
$101 = 400.000 (Y-axis travel resolution, step/mm)
$102 = 629.920 (Z-axis travel resolution, step/mm)
$110 = 1000.000 (X-axis maximum rate, mm/min)
$111 = 1000.000 (Y-axis maximum rate, mm/min)
$112 = 500.000 (Z-axis maximum rate, mm/min)
$120 = 10.000 (X-axis acceleration, mm/sec^2)
$121 = 10.000 (Y-axis acceleration, mm/sec^2)
$122 = 10.000 (Z-axis acceleration, mm/sec^2)
$130 = 200.000 (X-axis maximum travel, millimeters)
$131 = 200.000 (Y-axis maximum travel, millimeters)
$132 = 200.000 (Z-axis maximum travel, millimeters)

The gcode I was printing starts out like this (exported from ArtCam):

T1M6
G0Z5.000
G0X0.000Y0.000S50000M3
G0X3.654Y-7.641Z5.000
G1Z0.000F1000.0
G1X3.578Y-7.615F1500.0
X3.546Y-7.542
Y-5.707
X3.567Y-5.645
X3.621Y-5.610

Let me know if any other settings would be helpful and thanks for the help.

 

Increasing the acceleration ($120-122) made a ton of difference. It still seems a little slow, but that may just be inherent with leadscrew designs. I currently have acceleration at around 250 and was able to crank my max feed rates to 3500. Anything over that and my stepper motors start barking and not spinning. Not sure what this means? I currently have my microsteps set at 1/16, so maybe reducing that would allow me to push them faster?

 

I probably wouldn't try to launch a space shuttle with it, but it's surprisingly rigid and accurate. I haven't done any in depth testing with machining tools or anything (I'm obviously very new to this), but the few parts I've made fit perfectly. I suspect that it will eventually tear itself apart, though, so I'm very quickly trying to get all my plates cut out without putting the machine under too much stress. For this plate I used .3mm cut depth, around 250mm/min (very slow), and I think around 15-18k RPM (around 2 on my Makita router).This is 1/4" 6061 aluminum.




I do have two problems right now that I'm trying to work through. I don't think they are plate related, but they might be. The first is that my Z axis seems to plunge inconsistently. The bed is relatively level and the gantry is level with it, but it consistently cuts deeper the further right on the X axis I go. Inevitably if I put my z axis on the top of the stock, set zero, home, and then return to zero, my cutter try to dig 1-2mm into the plate. It almost never goes back to exactly the same z position (the other two axis are spot-on).

The other problem is that I'm getting a lot of resonance at high RPMs. I tried to cut some MDF last night and it was mind-piercing. I think this is caused by looseness in my Z axis gantry, but when I tighten the mini v-wheels down, the wheels stop rolling and instead drag. This causes a stutter since the weight of the router is putting uneven forces on the gantry plate.

The CBeam design with the mini wheels for the z axis is kind of clever, but so far I'm not really a fan. It takes up too much space and it really seems unbalanced and has a lot of deflection. I'm definitely going to spend some time trying to improve it.. either by fixing my mistakes or redesigning it entirely.

 

I tried the $1=255, but didn't notice any difference. I did a little more experimenting, and what is happening is that each time the Z axis moves up and then back down, it seems to lose between 1 and 2mm. I created a part with 10 slots, and each slot is slightly deeper than the next by about 1-2mm. I typically noticed this when the cuts went from left to right, but that was just a coincidence. Cuts along the Y axis do exactly the same thing.

 

Looks like I was losing steps on the retract. I changed the gearing and adjusted my microsteps down and it appears to be holding the depth correctly now. The first two columns were before the change, and the last column is after:

 

No, the support rail goes underneath the X axis cbeam (slightly darker rail in the shot below):



I joined the two together with blind joints. I honestly don't know if this extra reinforcement is even required. Those C-Beams are already pretty beefy. I'd talk to one of the guys that have built 1000mm CBeam machines and ask if they've noticed any deflection in the middle. I honestly doubt it, but for $15 bucks and no negative impact to the design, I thought it was easily worth it. It also has the added advantage of moving the wheels further apart, adding a little extra weight, and increasing the surface area where it joins to the Z axis cbeam.

Just throwing this out there, but what might be interesting would be to move the CBeam down to where I have the 20x40, and then replace the 20x40 with a 20x60 and move it above the cbeam. This would allow you to use 3 rows of wheels instead of two and that would definitely have a positive impact on reducing z axis torquing. You could also put a 20x20 next to the bottom cbeam to allow for 9 wheels on the bottom. I'll mock that up when I get home tonight and upload some more plates if it looks like it would work out.

By the way, I mentioned earlier that my 3d printed wheels didn't work out. That's not entirely true. I am using using 3d printed wheels (polycarbonate) for the large wheels and it seems to be working fine. You can see them above in the Z axis gantry shot.

 

Justin, I beefed up the design a little bit and it's actually working much better. Aluminum would be better, but so far it's taken whatever I've thrown at it:



I did another plate last night and got no chatter and was able to double my cutting speeds (450mm/min+). Halfway through I started boosting the feed speed to 200% and didn't seem to have any issues. I'm going to experiment a little more tonight and see how much faster I can go.

You can see I have my router at the top and I'm having to cut way down at the bottom of the Z axis, though. When I rebuild the X/Z axis, I'm going to mount the Z axis lower so that I get more range and can mount the router further down. It would be much better if I could choke up closer to the bit.

 

Yeah, I've got this thing in my office right now, so dust/debris management is high on my list. I'm planning on attaching some plexiglass to the cbeams left and right of the cutter. Honestly it's more about making it easier to keep clean than protecting the wheels. In my experience, stuff just falls harmlessly into the slots of the cbeam. I also plan on adding a dust shoe at some point soon.

Just curious, why do you want a longer z axis? I never make it more than 6" up my z axis and can't imagine why I ever would.

 

Problem you will have with the Sphinx design (and honestly, most designs) is that the bottom of the Z axis is going to collide with any tall object you would want to cut unless it just happens to be along a 1" perimeter of the object. If you look at the shot above, that black plate at the bottom of the Z axis does not move. You *could* permanently raise the Z axis up a little bit, but it's going to make working with shorter end mills more painful (you'll have to use a tall wasteboard or vice). Either way, unless I'm missing something, with this design you will never get more clearance than the bottom of the X axis (in my case, that's around 125mm from the table).

 

Thanks Rick. I'm embarrassed to say I didn't even consider that option. I'll definitely check out the XL gantry plate. I like the aesthetics of the small wheels, but I've had constant issues with slop.