I’m curious as to the limits people have been able to push their X-Carve as far as increasing the speed at which it rapids. I’m interested in maximizing the speed of the XY axes the most. It is a question of setting both the GRBL max speed and max rate of acceleration.
Thus far I’ve been able to max out the X axis at 48000mm/sec at 2000mm/sec^2 acceleration. This is actually a drop from the highest reliable point I found (up around 56000mm/sec, any acceleration above 2000mm would pretty much stall out at any max speed rate). I could probably get it up further but I don’t want to turn up the power going to the motor any further, I’d like to keep it as low as possible to avoid burnout as much as possible and prolong the life of my machine.
Unfortunately I haven’t been able to come nearly as close with the Y axis speed, and the issue is not due to the motors becoming overburdened and the power setting on the trimpot being too low. With the machine entirely off and steppers disconnected from the gshield I simply cannot get the gantry to move very quickly before it begins stuttering in an ugly fashion when pushing it by hand.
The only thing I have discovered is that if I disconnect the two Y axis motors from eachother then the gantry can move freely and quickly, just like the X axis carriage. But once the two motors are connected back together (as they are to receive the same Y axis stepper signal from the gshield) then that’s where the trouble starts… So, it has nothing to do with gshield Y axis trimpot setting or anything, because the stuttering occurs when moved by hand while completely disconnected from the gshield. It’s purely a mechanical issue involving the connection between the two Y axis motors.
I am curious if this is just an inherent problem with the design of the machine - having two steppers operating like this where they are wired backwards from eachother, on a belt-drive setup. Or if there is just a problem with my machine itself that needs tweaking/adjustment/fixing/etc.
No takers? Has anybody else even attempted to push their machine to go as fast as it can (without a load)? The application I am using my x-carve for essentially is the dragging of a fine bit across the surface of some aluminum, and so there is virtually zero load, and getting it to go as fast as possible would be ideal, but this Y axis issue is just bringing my project to a literal crawl.
It could be an issue with the gantry not being properly parallel to the rails. Because of the design it easily for one side to get a little ahead of another when the machine is powered down.
One forum poster has some set blocks he uses to make sure the y axis is properly aligned on startup.
Another poster had a homing switch mod that also did this:
A mod to protect limit switches:
You may want to talk with someone who has upgraded to the XControler. As it has each y motor on a separate driver chip so they are no longer tied together.
Because you are running as such high accelerations you might want to look at other controllers. It might make a difference for your application. (I am not sure as Acceleration and Torque curves are still pretty confusing to me.)
GRBL does linear acceleration curves.
TinlyG does S acceleration curves.
I believe the smoothie board does S curve acceleration too.
Thanks for the reply. The issue is not with GRBL or the gShield as it occurs when the machine is completely disconnected from the gShield and only when the two Y motors are still connected to one-another. The gantry simply will not move faster than 10000mm/sec without stuttering horribly, whether it be powered by the gShield or by hand while disconnected from the gShield. Looking at an alternative that powers the two Y motors independently seems like it should be a solution, but I’m wondering if there is a fix that doesn’t cost money as I’ve already dumped all my funds I had acquired for this project into the machine itself.
Yes that is “working as intended” because stepper motors also work as generators. So when you free move the gantry the stepper is generating voltage.
Because the Y motors are wired together in reverse they start to power each other.
So when you free move them fast enough they act as a break.
It also puts power back into the gshield which can make the leds light up and other strange things happen. (In my case it can trip the spindle on relay - scary!)
It is unknown if this can potently damage the gshield or not. So most people try not to move the motors too fast by hand.
It should not be a problem when the motors are being powered by the gshield as they are both being powered at the same time to work together.
That is tricky.
Your best bet is to see if anyone who has independently powered Y motors will do some test runs at your acceleration rates and see if it can handle it. That will let you know if changing the controller will make a difference.
But you will have to spend a little $ to make a change.
If you are comfortable with soldering electronics yourself you could wire in a 4th stepper driver chip.
But unless you are comfortable with that level of modification and have a lot of the necessary boards pins and wires on hand already you would be better off going with something else.
Here is the cheapest 4 driver option I have seen at around $60
GRBL shield alternative.
But if you do wind up spending $ on this I would take a hard look at if S curve acceleration will help. Saving $100 won’t help if it doesn’t fix the problem.
Read up on acceleration curves. You may be approaching the max acceleration of the motor. I know if it goes too high the motor can’t keep up and will start to miss steps. But how high is too high has to do with the motor torque rating and gearing ratio on the axis.
(I think the Nema 23’s in the kit are 140 oz-inch)
But I guess if the X axis is handling it then it is not a motor torque issue as they are the same motors on the same gearing.
If it is just the Y axis (and you checked that no V wheels, belts or pulley set screws are loose or too tight) my best guess it is the motors being tied together and at those high speeds the reverse current being generated (usually not enough to be an issue) is starting to be a problem?
How big of a project are you working on?
If you are doing lots of BIG moves you may not need a huge acceleration.
Acceleration being how quick you get to max speed.
If you are doing all 12 inch moves does it mater if you are getting to max speed in .08 inches instead of .01?
What if you just try the higher speeds without the high accelerations. Yes you won’t actual be running at max speed 100% but that is the only adjustment I can think of without moding the machine.
Thanks for the suggestions involving a 4th driver, this seems to be the only solution. It’s clearly an issue arising from the two motors being wired together in reverse and has nothing to do with acceleration. Once the gantry reaches a certain speed - regardless of acceleration and whether or not its even powered or wired to the gshield - the motors stall out, and it’s distinctly different from when a motor is trying to accelerate too fast when powered and steps are missed.
I’m working on a project that involves scoring aluminum using a conically-ground carbide bit, and the goal is to produce the smoothest, shiniest, and finest grooves possible, which works best on the X-Carve the faster I push it. The work size ranges from under a square foot to over 2x2 feet. The greatest portion of the load is just the inertia of the machine itself, and in this case there is an upper limit to how fast I can move on the X-Carve because of the single driver driving both motors on the Y-axis. The grooves can range in size from less than an inch to being the whole size of the work. For the most part the speeds I’ve been able to achieve seem to be working well enough, but there’s a very clear difference when I run the machine at a comparably slower speed and after a point increasing the speed provides diminishing returns, so I may just keep my machine the way it is.