Hello, I am currently in the process of upgrading my CNC. Currently I am with a MPCNC Primo and am looking to move up to the LR3. However, one of the problems I am having is the stepper skipping. I would like to know if anyone has done a conversion from Nema 17 to Nema 23 motors. I believe this would greatly help the power of the system.
Thanks
Ryan considered the nema23 in the design. But AFAIK, no one has tried it. Ryan may have some parts he could share. But they probably haven’t even been printed, let alone checked for their ability to be assembled.
Are you sure the nema17s are the limit you’ve reached? A lot of skipping step problems are something else. If you’re driving it very hard, at fast speeds and deep cuts, then the motors might help. But I would eliminate everything else first. The nema23 machines I’ve seen haven’t moved the bar as much as you would hope.
Thank you for your reply. Actually I use a Blackbox from openbuild to drive my MPCNC and the difference in Amp help in a design from another. I think it might help to have a better speed and more agressive. I am welling to try a nema 23 version and share my results here or in private.
Thank you
share my results here or in private.
Here. People on the forum will be interested in any information about upgrades. You might want to consider developing a test for before and after to get some real info on the impact of your changes. We see a number of improvements reported on this forum, but there is little testing to verify their impact.
Substantial torque improvements have been reported with NEMA17s using external drivers run at 40V and a 32-bit processor. I’ve also seen NEMA17 motors with substantially higher torque than the common ones.
Im with Jeff on checking other things out. Nema 17’s can be more than stout enough for the mechanics we are working with.
How fast are you going now? Besides the black box, do you have any other custom changes?
Ryan is less available this week, but if you ping him next, he can talk to you about what he thinks about the nema23 LR3 files. He would know better what is needed to try those files out.
On the occasions when steps are skipped with a standard LR3 (or LR2 for that matter), do you all think it is a result of a belt slipping past teeth on the gear cogs, or a motor getting held back by more force than it could torque through?
Skipping steps on the motors is definitely possible. I’ve never heard of the teeth jumping cogs. It sounds like crunching gears, but it is really just the motor quickly skipping to another position. There is no slipping teeth.
10mm belt teeth ought to be stronger than the torque available to the motors, so it shouldn’t skip, unless your belt is way too loose, which will come with a host of other problems.
I was looking at some 92oz.in NEMA17 motors, but they’re a bit long for the LR3 application at 60mm bodies. Still easier to adapt than NEMA23s
I’ve never observed a stepper skipping issue. Instead, the belts will slip past the teeth and grind away if there is an operator error such as attempting to move in xy while the bit is in a drill hole. An even more dramatic scenario occurs when one of the y axis belt holders pops off the fastener. I replaced all the pla-printed belt holders with petg and drove 3 inch wood screws through the top. Problem solved.
Here’s my lr3 setup:
Motor: 84oz nema 17
Driver: tmc2209 set at 0.95 amps, one driver for each motor
Belts: 10mm polyurethane with steel reinforced wire
Spindle: Bosch 1617EVS, 2.25hp
The Bosch router attached to the lr3 can accurately cut through hemlock at 1" deep with a 1/4" straight cut router bit.
The steel reinforced belts tend to fatigue and break over the tight radius of these small pulleys, so keep an eye on that.
And if you’re cutting an inch deep of anything harder than foam with a 1/4 inch endmill, post a video, pat yourself on the back, and crack a beverage, because I’m pretty sure you win.
That’s a great result, right there.
Are you sure the belts are slipping? You should mark the pulleys and see if they are still moving when you skip. Because skipping steps looks and sounds like gears jumping from on tooth to another, but it is all internal to the stepper motor and nothing is grinding. It is the stepper adjusting 3.6 degrees to align with the next magnet (and often happens a couple of times at once, crunch crunch).
The steel belts will get fatigue fractures while maneuvering around the pulleys and the steel breaks, then the rubber stretches. The fiber reinforced ones don’t stretch and the fiber won’t break like steel.
Thanks, I will get a set of the fiber reinforced belts. You could be right that the motors are actually skipping rather than grinding. I can see marks on the belts if they were not sufficiently tight so there could be grinding occurring as well.
Do you have any recommendations for the driver current settings for 84oz nema17 and tmc2209?
I have a set of external drivers that were originally used with nema23 motors, a 48 volt power supply. The mks robin nano has 4 signal pins per driver (enable, step, dir, ground). I believe you connect +5v to pul+, dir+ enbl+ and the signal pins to the pul-, dir-, enbl-.
Start with something close to 900mA and then raise it until your motors or drivers get too hot. 50C on the motors or so is safe. Much more and they may deform in the heat.
Here’s the requested video:
https://youtu.be/f9R3hqmp1wE
My apologies, it needed two passes at 13.25 depth instead of one, plus a finishing pass.
The bit is 1/4in straight cut.
My guy, that’s still really impressive! By all means, continue to hot rod this thing and push the envelope, but I hope you understand that you have some really great performance right there!
I wonder how much more that router can take!
Good job!
1" inch cutting depth with a straight bit is impressive.
I haven’t seen a mount posted for the Bosch 1617EVS. Are you willing to post the 1617EVS mount to Printables?
Yes, the Bosch mount is posted on thingverse:
Impressive deep cuts, curious what feed rate and speed was used?
Well I was pushing it to the limits, I believe the speed was 22k and feedrate was ~60mm/sec.
