# Maximum Acceleration Calculator

• OK, I changed "minimal lag requirement" to "safety factor" now and added descriptions to each line of the result calculation.

• !!Brilliant!!

Thanks for putting this together and letting me know about it.

• Hi, does it works for CoreXY machines? The worst condition is a diagonal motion as that is made by only one motor.

• @3doeste It should do. The way to treat it is to use the mass to be moved in the Y direction, which is the worse case. That is to say, the mass will be the X carriage plus the X rails and Y carriages. As you rightly say, pure X or pure Y moves will use both motors but 45 degree infill will use only one. So spec the motors as you would for moving Y on a Cartesian.

• @3doeste I have 2 Hypercube Evolutions and have been doing some testing before the calculator was available. My next Step is to weigh my X and Y moving masses. But the calculator is just a starting point. For example one of my machines is very light so I have also been wondering about jerk that the calculator doesn’t address and which smaller CoreXY s should perform quite well. When I get home from holidays I will see how high I can push jerk.
Another factor is the overall flexibility of the machine, including the belt stretch and associated resonances, X rod flex, hotend mount etc. I have no feel yet for how the belts affect performance - others might be able to comment.

I suspect the best acceleration is much lower than the calculator provides, depending on your machine construction and quality preferences.

• @garis The calculator is for stepper motors only. It is an attempt to calculate the maximum acceleration that a given stepper motor can produce for a given mass.

Whether the rest of the machine is capable of attaining those acceleration values without problems, is of course an entirely different thing.

• I have a fully machine aluminum CoreXY, I weigh my Y axis and it's around 1500gr of moving mass and my motors have 65n.cm of torque.
The calculator give an aceleration of around 4000, I tried 6000 without problems but used 4000. And with jerk I was using 1800 until I had to do a print with sharp corners and they got wavy in Z (only on corners), so I put the jerk at 600 and it came out perfect, but otherwise, 1800 works great.
I can't get more than 200 mm/s of speed though without losing steps. Perhaps that will improve with 24v and / or 20T pulleys. I have 16T currently.

• CoreXY is complicated.

X axis moves divide the force to accelerate the X carriage by two, split between the motors. Likewise Y axis moves divide the force to accelerate the entire bridge by two. That’s straightforward enough.

Diagonal moves are weeeiiiird. The CoreXY belt path is effectively a sqrt(2):1 compound pulley. IE each 1 mm of diagonal XY travel requires 1.41mm of belt travel, and you get a corresponding force multiplication effect because of that travel ratio. That’s one little-appreciated reason why CoreXY tends to produce nicer prints than an equivalent Cartesian — you’ve effectively geared down the motors.

• @rcarlyle " That’s one little-appreciated reason why CoreXY tends to produce nicer prints than an equivalent Cartesian — you’ve effectively geared down the motors. "

So is it worth it running at .9deg motor?

• @og3d depends on your PSU voltage and desired speeds. 0.9 degree motors can only spin half as fast, all else being equal, and CoreXY can only move on diagonals 71% as fast as Cartesian, all else being equal. But the typical 24v Cartesian printer with 1.8 degree steppers can do like 300mm/s without any issues as far as the motors are concerned and people seldom ever go that fast.

• Well on average the printers usual do not get close to 300mm/s, so on CoreXY using a .9 deg I wont see much improvement vs a 1.8deg motor, since by design CoreXY is geared down. Just making sure I understood it correctly.

I have a delta but at speeds 100mm/s quality is not good and seeing Rail Core II print results at 100mm/s got me inspired to build a CoreXY for my second printer.

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.