Bob, I've been to parts of Florida in the summer; I can see why it is done that way *grin*. How'd the Xylotex work out for you? Did you ever encounter mid-band resonance/are the motors fitted with dampeners?
I`m putting up another wall of text, just in case this is useful for anyone else is still following along.
Steppers have a rated holding torque that drops off with speed increase, so the torque needed has to be calculated at the operating parameters expected. Here's a
sample curve. Faster operation means less rotary torque available to be converted into linear force.
Servo motors (excluding a encoder equipped stepper, which technically is a "servo") generally have a rated torque which is run-all-day-long value and a peak torque, typically 3 times the rated for only a limited time period that can decline with increasing speed but remains fairly flat (brushed DC motors drop off faster than BLDC or AC Servo).
Motor Sizing is dependent on what the machine is expected to do: how much weight and cutting forces must be moved at what rate and what acceleration, as well as what mechanical advantage is supplied by the rotary to linear motion mechanism (pitch of the screw). The overall goal is to make certain the torque available is
sufficient during acceleration and nominal operations. On a servo, the intermittent region of the curve is for acceleration and deceleration, with the average needs within the rated torque envelope; on a stepper, the holding torque drops with speed, sometimes precipitously.
Now, before I run on further and not address Bob's comments....
I don't know how to compare the ratings of the servo motor at 400W to the 425 in-oz stall torque on the Xylotex steppers, or if 425 in-oz is a reasonable torque for the G0704.
Warning: I did make attempts to check my math, but I`ve been multi-tasking this all day *scribbles*, and this prime ground (for me) for mistakes, what with all the conversions.
Is the 425 in-oz a reasonable torque for the G0704?
Hoss states his
bare recommended minimum is 381 oz-in steppers on the X and Y, and a 465 on the Z, and then states that 570 oz-in steppers are a
great match (on the X and Y axis; 906 oz-in on Z) with stock screws (which have huge efficiency losses compared to ballscrews, and will traverse slower) or ballscrews. So, that gives us a range. Others builds he references also seem to use similar motor ranges.
Xylotex does not seem to publish a torque vs. speed graph for the motor, so I looked at a
Automation Technologies 425 unit.
The graph is a bit daft, apparently missing the top value (425 oz-in is 3 N.m), and is in Pulses Per Second, not Rotations Per Minute; makes it tricky to compare.
1.8 degree motor at 8x micro-stepping would be 1 600 steps per rotation. 8000 PPS = 480 000 pulses per minute, 480 000 PPM / 1600 steps = 300 RPM.
So, according to the curve, 42.5 oz-in is remaining at 300 RPM into a (picked out of the blue) 5mm pitch (~0.2 inch per rotation) ballscrew will
produce ~75 pounds of force at 53 Inches Per Minute (IPM) of travel; this is just first approximation, and disregards friction from the gibs, dovetails, opposing cutting forces, acceleration losses, etc.
Again, this is assuming the two motors are in any way similar; I have no way of knowing, and measuring in PPS is...unusual, to put it mildly.
Looking at the
570 oz-in Hoss recommended, it has a
Torque-Speed Curve claiming at 1000 RPM it will produce ~63 oz-in translating into ~111 lbs. of linear force.
So the 570 stepper provides more rapid speed and more force. Why 1000 RPM? Using the above screw at 1 000 RPM will produce 200 IPM travels. Why 200 IPM?
Feeds and Speeds
I set out to see what the maximum rapid IPM people have been achieving. Again, the ubiquitous Hoss has achieved
300 IPM rapids (with the 570 oz-in motor). 200 IPM seems to be quoted as a more common number, so I used it in the example above. He also claims in one
thread he has hit 200IPM with a 0.5 endmill during High Speed Machining operations, so it seems a reasonable number to look at for actual use (moving 200 pounds at 200IPM might be asking a wee bit much, however).
To do some basic cross-checking, I decided to take what I'd seen and compare the quoted numbers to G-Wizard rendered ones. The G0704 comes stock with a 1 HP spindle motor, end mill capacity of 0.75 inch (according to Grizzly). I took that information and built a machine in G-Wizard, and started in.
One
video has Hoss demonstrating with a 0.8HP spindle motor driving a 0.5 inch four flute HSS endmill at 2250RPM, He managed a full width cut at 0.375 inches depth in 6061 Aluminum at a stated 12-15 IPM with no excessive audible motor loading noted.
G-Wizard, using this information and that from the mill specifications, gives a value of 18.4 IPM for a very aggressive cut, and 9.7 for a fine finish. The IPM is limited on the aggressive setting by available HP, and tool deflection on the fine finish. So, the theoretical roughly corresponds to the actual achieved numbers.
I then cross-checked using a
milling power calculator (this is an approximation only) on 6061 Aluminum and lowered the feedrate to 13 IPM (due to the calculator claim this is 1HP motor load), I did the same using 4140 steel which nets a feed rate of between 2 (calculator) and 2.6 (G-Wizard) IPM. Hoss`s numbers look good (not that I doubted them; he knows more about practical machining than *I* do, for certain).
What I take away from the above is: Yes,
depending on the motor's characteristics and mode (low impedance, high driving voltage, bipolar parallel more usable torque at higher RPM) and its torque-speed curve, a 425 oz-in stepper would work on the G0704
X and Y axis with high rapid speeds, as shown by theory, actual examples and cutting rates.
Comparing the ratings of the servo motor at 400W
Power is work over time, either in HorsePower or Watts. Torque x RPM = power.
Quick and dirty power formula for motor : Power: Watts = (heaviest object in lbs. * Inch Per Minute) / 531 (the 531 constant is just cutting out the conversions)
Example: ~113 Watts required to move 200 pounds (assume: 75 pounds for table and attachments, plus 125 pounds for work-piece, holding) at 300 IPM.
The 570 oz-in motor in the previous example would be generating ~46.5 Watts, and the 425 oz- in generating ~10 Watts. The G0704 1HP spindle motor, for contrast, would produce 448 oz-in at 2250 RPM to rate at 1 HP.
The DMM 640-DST-A6TS1 (DYN2, 60mm fram, 400W, Medium Inertia, 16 Bit Encoder, 60V max, Straight Shaft) has a rated torque of 1.27 N.m (
180 oz-in) and a peak torque of 3.82 N.m (540 oz-in). Here are the
Torque-Speed Curves.
As such, while 180 oz-in is considerably weaker than the holding torque of the aforementioned steppers, at speed (200 IPM rate * 5 rev. per inch = 1000 RPM) the motor is still generating 180 oz-in (18x and 3.8x respectively for the previously mentioned steppers) and ~318 pounds of linear force, or 133 Watts, with 3x that much in reserve for intermittent use at that speed.
Edit: Acceleration is a critical part of system design, and I will actually (gasp) refrain from getting into it at this time; suffice it to say there should be ample reserve capacity to
put the pedal to the metal, so to speak.
Additional Resources:
- here's a handy guide put together by
CNC Cookbook for motor sizing. In that post is a link for a Excel spreadsheet that has assorted variables to contemplate.
(for the above Excel file)
Mill Table Weight: I also could not find the stated weight of the mill table, so off I went to a calculator with the dimensions for a guesstimate: 73 lbs.
Part and Fixture Weight or Table Capacity: The G0704 has a table capacity of 125-150 lbs (also according to
Hoss).
Cutting Force: That can be determined via
calculators found on the web.
Linear aka Rapid IPM:
Hoss again. 300 IPM (utilizing steppers)
Cutting IPM: Hoss claims in one
thread he can hit 200IPM with a 0.5 endmill. This is in-line with comments others have made.
Leadscrew: Not much information I could quickly research, sorry.
Lead/Pitch: I looked quickly and could not seem to find what the leadscrew is that comes with the G0704, one claimed it was 5mm.
Diameter and Length: Again, no information.
Lead Efficiency: I could also not find reference to the lead efficiency, but it is somewhere between 20-40% generally. .4 is the value I run across most often.
-
Kollmorgen Motioneering Online : online (register access) utility for linear motion system design
-
Choosing Stepper or Servo: Gecko Drives
-
MIT on Speed-Torque Curves
- Maris Freimanis from Gecko on a CNCZone thread running some numbers as an
example of sizing, gearing, power, resolution for a gantry router