Since there is some good info coming out of this thread I want to add some pertinent information about comparisons between steppers and servos. Mike uses the SB alpha (Oriental AS/AR series) 7.2:1 gearmotor as a baseline. There are some erroneous assumptions made about these motors. First, the available torque. The mfgrs spec sheet shows ~1280 ozin in their spec sheet. Holding torque. Assume a minimum 30% decrease when moving. (very slowly) Torque chart is given at FULL steps (sometimes 1/2) in any case neither are real world numbers as 1/10th microstepping is employed in the drives.
Tenth microstepping is a useful tool to reduce noise and smooth out a motor, especially for environments where noise would be an issue, like medical equipment, but it may be the worst choice in CNC routers where variable and unpredictable loads are present. I have seen numerous guys here order servos based on what torque they THOUGHT their steppers produced. See here:
https://www.faulhaber.com/en/support/te ... ostepping/
For the most part you can assume that a given stepper with 1/8 or 1/10 microstepping is producing 30% or less of its holding torque while cutting and 20% or less while at rapid speeds. Take that SB oriental alpha motor. !280 ozin thru a 30T pinion (1.5" dia, .75" radius) will yield holding torque of 1280/.75 or a bit over 1700 ozin at the rack. Adjusted for 1/10th Ms (25% avg) around 425 ozin at the rack.
The 3411Sservo has a CONTINUOUS rating of 150 ozin, 750 thru a 5:1 gearbox and using the 24T (1.2" dia, .6" rad) yields 1250 ozin at the rack. By my calculations that's 3 times the torque. And this is without taking into account that the 3411S has over 4 times that available as peak torque for up to about 3 seconds. (638 *5 /.6 = 5316 ozin) well over 10 times the torque of the OEM motor. That's why I do not recommend larger motors for SB PRS frames. I will not even put anything but a NEMA23 on a PRT frame.
Back in the day the SB alpha motor was touted as one of the most powerful.......yada yada. reality is in this day and age it is by far the weakest NEMA34 motor sold in the US. The closed loop NEMA23 motors in my "alpha replacement" closed loop system shown here:
viewtopic.php?f=62&t=2480 have no problems outperforming the larger appearing, more elegant looking 7.2:1 alpha gearmotors, and they have been doing it every day in commercial applications for close to a couple years.
Keep in mind that these machines are bolted together with questionable tolerances. The V rollers can "climb" the rail with excessive side torque and the top heavy gantry design places undue stress on the pinion shaft under high force and accel/decel. Add in the fact that the motors and rails are not mounted on machined surfaces and may not "run true" is it surprising how they wear pinions on a far too frequent basis? There is no benefit gained from overpowering a frame. None from underpowering one either, but paying for power that you cant use is not a plus. But you can play: "My servo is bigger than your servo!"
Power supplies: Be aware that all the calcs are correct, except that no one talkes into account the real denominator for calculations. How much of the available power you are going to use. And are you aware that the IPC5 will put out close to 900W for a few seconds, if you need it? And remember, the motor draws ~200w when outputting its rated continuous torque. IF you use the Clearview software you will see that on a light frame router they seldom draw over 100 on a light frame router and that is during rapid (aggressive) acceleration. Around 50 most of the time.
So yes, along with running smaller motors than most DIY guys, I use many less watts of power supply too! My current machine is set up for high torque motors on 16:1 gears plus a monster on 25:1 rotary and I use 2 IPC3 power supplies. Guess what they put out when cooled properly??