Originally Posted by
hanermo
Yes, and no.
As you said at 12 mm travel you get to max speed in your example.
If one is making traditional parts, blockish, with some corners, and some little chamfers, this will make relatively-little difference, in steel, if the top speed in cutting is limited for some reason like tool length, machine rigidity, power, etc.
But..
A servo system gets to the same top speed in approx 0.5 mm of travel.
About 24 times faster.
IF one is doing contouring in 2d, or 3d, or shapes of any kind, the moves are usually 0.01 mm - 0.1 mm.
Very short moves, lots of them.
The servo system will be == 20x faster, and leave better finishes.
I used to use the "best" stepper drives, geckos 203V, at 68V DC.
With the "best" hw pulse system at the time, a centipede (smoothstepper, pokeys, others).
Driving a very good gantry, on linear rails (low friction, low stiction),
A stepper based system takes about 1 sec to get to top speed, if I want high speed.
(I don´t, so far, in 12 years. Top speed is irrelevant.)
Because of the torque curve.
At better settings, about 0.3 secs to top speed, around the 12 mm travel you mentioned (5-10 mm).
But..
my current system with same-size cheap nema 23 sized ac brushless servos, does the same speed in approx 5x less time.
In about 1 mm of travel.
And it could do so at 20x less time, if I wanted to, but I don´t, because it is too fast and wears things out.
Examples:
A best-system stepper setup took about 3 secs to drive the stepper to 12.000 rpm no load.
0.5 secs "good" load, to 1000 rpm.
Suitably inertia-matched.
The servo goes to 3000 rpm in 20 ms.
To 1000 rpm in 10 ms.
0.1 secs "good" load, to 1000 rpm.
Would do 0.05 secs, but this is bad for the mechanicals.
The same-motor-sized servo system of *same or similar cost*,
is 10x faster in the real world for work done, and
5x more accurate in the real world.
Running the roadrunner on my small-step size machines, goes about 5x faster with the servo systems.
You spend about the same money, but get 500% better results, and 500% more accurate results.
All my stuff is about accuracy, I have never cared about speed.
I use about 1/3 top acceleration,and 1/3 top speed, in sw settings, to reduce risks and costs in errors.
Example:
A cheap nema 23 ac brushless servo system = 290 €, EU, taxes paid 22% VAT.
400W, 60V, 5000 counts, 3000 rpm, motor, drive, industrial cables.
I will sell one to anyone, if they need, as I buy them in boxes of 10.
A gecko 203V + nema 23/34 motor will cost about the same, maybe 250€ with cables (with 22% VAT).
The servo system will run 3000 rpm vs 800 rpm,
at 2.3 Nm cont. torque at 3000 rpm vs .5 Nm at 800 rpm (stepper, typical), 5:1
20 ms to 3000 rpm vs 100 ms to 2000 rpm, 5:1
2/5000 accuracy in positioning vs 1/400 accuracy, 6:1
Lathe runs 7.5 m/min or 7500 mm/min.
But at 0.25 microns step size, or 4000 steps/mm.
And a single step 1/10.000 x 1:2 belt drive has 10 Nm x 2 ==> 2000 kgf push force to make it position.
Error readout on the servo drive led indicators is always 0.