Title is the problem. I send it across the board at about 9200mm/min, no issues. From 9200 to about 9500 it sounds the same, so it probably isn’t picking up speed. When I try to send it at 10000 it starts moving then says “up yours, dude”. I cut the acceleration from 200 to 100 and it was definitely slower to get moving, but still jammed up. Sorry about the aspect ratio.
Damn! I’ve never seen them lock that hard before!
2 Likes
Nice!
This really isn’t an issue for running my primo, but I have another project in mind that I want to run with grbl and my other set of steppers that’s gonna need about 500 rpms. At 16t, I think this works out closer to 300.
Weird is that what happens when a GRBL board locks up? I would try it with mine but I literally just unplugged it, I had it hooked up for at least two weeks and unplugged it 10 minutes ago to test some more Marlin.
I don’t know what’s going on, honestly. I spent about an hour tonight searching and about an hour or so when I first set it up. I figured with all the brains here, I’d be stupid not to at least ask. If nobody knows by next weekend, I might dig out the teeny tiny and see how it behaves. It has the rambo, this guy has a mega/ramps 1.6 plus.
158mm/s is pretty fast.
I think if you really need that you should get much larger pulleys.
I have the marlin builds maxed at 50mm/s, but with the laser we are limited to ~30.
They won’t be driving belts. Is this the sort of thing that would happen for sending steps too fast? I’m reluctant to mess with this one too much because it’s running, but if taking the microsteps down from 1/16 to 1/4 or 1/2 might do it, that’s a good solution. Otherwise I’ll have to pick up some different steppers and see if they behave the same way. After I try the rambo.
At 35mm/sec we are at much less than half peak power of the steppers and it falls off fast. The CPU determines what the max steps/sec it can handle but the stepper power is a direct factor of RPM so lowering the step count will not increase your power, but might get you into the realm of possible.
The right thing to do is gear it correctly. You are trying to race formula one in first gear in a honda civic.
1 Like
Do you have a 24V PSU?
If it is a limit on the cpu (it can’t bit bang fast enough), then smaller microstepping would help. If it is just skipping steps because it is just trying too hard, I would expect it to stutter, not just completely stop.
As a datapoint, on my 24x24 Primo I’m running grblmega5x an a mega+ramps1.4 with 5@ purple drivers at 1/16th micro, 16T pulley, at 150mm/sec with 300mm/s2 acc, at 12V, and she purrs like a kitten during rapids. It has never skipped a step at 1.4A.
So, I doubt microstepping and cpu are issues here. I have had behavior like that when I was dealing with intermittent motor connections though. Looking at the video, it does seem like the driver is entering some unknown state. Seems to happen when not accellerating, so acc can probably go back up. Does the problem tend to occur in the same area of your table? That would point to wiring. You have an oscope handy?
Happens wherever I send the fast command, and no oscope.
150/s is 9000/min, which I can do. Ever try to go 10k?
12v I think.
OK you piqued my curiosity that there may be some limitation kicking in beyond 9k. I kicked mine up to 10k, and it just did 3 separate laps around the workspace at rapid, and no issues. Should I try 11k, lol… sounds fast enough already.
What drivers are you running?
[edit: For S&G’s, I tried a quick run at 11k and no problems cropped up. The rpm’s sounded really high, but it didn’t lose step or lock up. I went back to 9k for usage though, just in case.
1 Like
Don’t bother overthinking too much, the easy fix is really to lower the microstepping. Try 1/8 or even 1/4, it won’t really make a difference in terms of accuracy but it ill be really a whole lot faster plus it’ll give you more torque. It is not necessarily much louder as well, but this depends on the drivers you’re using.
On my printer (which is not using grbl but reprap firmware, and uses a duet board, but the same principles apply) I can go for more than 20k without any problem if using 1/4 step. Cannot do that with 1/32, otherwise the motors will do the same thing as yours. I suppose this has to do with the motors reluctance, they can’t just switch the coils poles that fast.
Oh, and as Kev said earlier, intermittent motor connection is a very good candidate also, I had a similar behavior whith a defective wiring harness too, in which case the issue will happen at a specific spot, and one axis will be locked while occasionally trying to move while the other will work just fine.
In any case, lowering the microstepping is always a good move to reach high speeds so maybe start by doing that, it’s the easiest.
1 Like
Thanks Dui! I was hoping someone had seen this before. I run the primo for hours with no issues so I doubt it’s wiring.
In fact, running it for hours is the whole reason I don’t want to start messing with the boards and changing things around.
I used to have a sign… “if it ain’t broke, fix it until it is!”
I’m really not concerned with torque though, so that’s just a bonus.
The 8825 purple ones. I also noticed today that my steps/mm are at 200. I think I may have set my microstepping to 1/32. I’m going to call that the problem
1 Like