Help me understand why my steps/mm don't jive.

Hi all,

I'm using a Workbee 1000 that is driven by lead screws at a pitch of 2mm/rev (I confirmed this with a caliper too).
My drivers are the DMT542T set at 400 pulse/rev (microstepping).
These are hi-torque Nema steppers and are spec'd at 200 steps/rev. Clearly 200 fullsteps would advance the gantry 2mm. Simple math. Oddly enough, my GRBL settings for $100, $101, and $102 (steps/mm) are each set to 50 to work at the proper portion of the NC code. I would have guessed this is a misconfiguration somewhere. I would have thought 400 pulses/rev on these 2mm displacement leads screws would equate to 200 steps per mm. So why is $100,$101,$102 working out to 50? If I set it to 200, it's moving 4x further.

Setting it to 50 works just fine. But it seems like a poor resolution. Right? OR something else is amiss.


BTW. The GRBL v1.1 and is on a Protoneer Pi Hat on my Rasberry Pi.

I look forward to some enlightenment. Thanks all.

Cheers,
Frank

 

OMG... I measured the thread distance but didn't compensate for the angle of the thread! So....one full rotation equates to 8mm of travel on the gantry. Now it makes sense. Sorry to bother with this. Maybe this will be helpful for someone else overlooking some important physical properties of a screw drive like I did. hehehe... Don't just measure the lead screw thread distance...Measure it by rotating the stepper (with the unit off!) a full rotation with a ruler. I can't believe I overlooked that. Ugh.

 

Wow. Thanks Peter!
Is that still advisable on a motor with only 917g-cm of detent torque? This motor's spec says 2.45N.m of holding torque and 90mN.m of detent torque. I fear I would have to route really slowly if I up the microstepping. I could be overcompensating and it's fine to do so but I fear running into a surprise of lost steps. Would you still recommend that adjustment with this motor?

 

2500mm/min? Whoa! I've never run it that fast. I fear the bit would snap right out of the workpiece! lol. The fastest I have run it is 700mm/min. On average I have been running at 150 and 300 but that's to avoid heavy tearout on plywood stock. It's gonna take more experience (and guts) before I can feel confident to run her that fast on any solid stock. Maybe when using 1/4" shank bits? You guys are clearly pros.

Thank you both for the confidence to increase the microstepping to 1600.

Cheers,
Frank

 

If you can help improve the unit, I would be so greatful!

It's a Workbee 1010 purchased from Bulkman3D. The drivers were originally TB6600's but I learned my lesson and upgraded to the Stepperonline's DMT542T's. The unit came with a MachV controller but I didn't want to be stuck with needing to use Windows so I purchased a Protoneer Pi CNC (v.2.6) hat that attaches to a Raspberry PI 3b and it runs Linux with bCNC on it. The spindle is a Makita mini router (SKU: 590466 at Home Depot). I made sure all wiring is shielded and wired to earth ground. I even re-ran the wiring from the Makita to be shielded as well. The unit came with a single 24V power supply but that prevented the unit from going past 1500mm/min without stalling. As a result, I updated the power supply to a qty of 2: 36V power supplies. One runs the X and Z axis while the other runs the Y and A (secondary Y) axis as the Y axis has two motors. For an idea of what the machine looks like, you can checkout my instagram account: @franklyrustic

I've got the leadscrews lubed properly. Table is square and wastboard is planed flat. I opted for the lower bed configuration in case I decide to work with stock thicker than 35mm. Maybe that results in too much travel down to the stock and weakens the router's axis?