Sizing a new Z axis motor

[waynec]

I'm getting ready to upgrade my ole PR Z axis, and I'm wondering about a new motor. My new Z is a precision guide, but its sorta heavy. I don't think the little PR Z motor is adequate.

How do I size a new Z motor? I'll be using a gas shock to counterbalance the weight of the motor and carriage. I'll be connecting to an Ascension 1000 controller.

I'm considering a 640 oz-in motor. I will use a Porter Cable for now, but I anticipate upgrading to a spindle later. Will the 640 oz be adequate?

And while I'm asking lots of questions, anybody know what the original PR Z axis motor is rated at? Mine is the little one attached to a ballscrew.

Thanks as always.

Parts are ordered for my new BWC rails and a new gantry.

Wayne C

[Brady Watson]

Without knowing the weight of the guide/slide assembly, the method of transmission (R&P/ballscrew etc) and the rate of resistance that the gas shock provides, there is little that anyone could suggest. You'll have to get the scale out and weigh components & tell us a bit more about your setup. It would seem to me that the Ascension people could guide you in the right direction if you are their customer. Have you tried asking them?

As a general rule if you are running R&P, you will need a stepper with a gearbox on it. I recommend looking on the Oriental Motor website for a motor with a gearbox. If using a screw, you could get by with a 1:1 motor if the screw pitch isn't too coarse. A 5 TPI screw would effectively give you a 5:1 reduction. The original PR motor is very weak. I would look for soemthing in the 500-600 Oz range with or without a gearbox depending on the load you are moving.

-B

[Towersonline]

Wayne, I have a PR with an upgraded Z and replacement motor. I will get you the numbers and post them in the AM. And no the little motor won't be up to the job.

[richards]

Wayne,
The Oriental Motor PK296A1A-SG3.6 motor will give you 22 lb*in of holding torque. The PK299-03AA motor wired as half-coil (unipolar) will give you 38.75 lb*in of holding torque. Those specs are found on the Oriental Motor web site. Be aware, however, that holding torque is reduced as much as 70% depending on whether or not you use the 'reduced current standby option' on the stepper drivers.

If I've read the specs correctly, my Alpha (1:1) has low speed torque figures that are very similar to the PK299 series of motors. At higher speeds, the Alpha is much better. My Alpha has a 3hp spindle with springs but no gas shocks. It has no trouble holding position on the Z-axis.

[Towersonline]

Wayne, The motor that is on my upgraded PR Z axes is PK296A1A-SG3.6. Hope this helps.

[waynec]

Thanks all!

I'm installing a precision ground ball screw with recirculating ball bearing guides on hardened rails. It will have a Z travel of 10 inches. I think it weighs in at about 45 lbs.

So that I'm sure I understand this, if I use a gas shock that equals the weight of my Z axis and router, the weight is effectively counterbalanced. This would mean that the motor has to deal with inertia, but not necessarily lifting the actual weight of the motor. So, I dont' really have to select a motor that can lift the entire weight of the axis/router combo? Is this right?

As I understand it, gettting too large a motor can be a problem as well. Is a 640oz in motor in that 'too large' range?

Finally, because no question is stupid, how do I convert from inch oz to lb in?

Alan from Custom CNC recommended a motor, but I'm wanting to double check and understand before I plunk down the pilasters.

Thanks all- very helpful!

Wayne

[richards]

Wayne,
That is quite a Z-axis! What are you going to be cutting that requires something like that?

Converting from oz*in to lb*in would only require that you divide the ounce figure by 16, so 640 oz*in = 40 lb*in. The figures represent the amount of torque available when the motor has a shaft or shaft/pulley with a one-inch radius.

The 640 oz*in motor is about perfect as far as a size 34 stepper goes. It is the size of the PK299 series motors from Oriental Motor. Its smaller cousin is the PK296 motor that has about 1/2 the holding torque. Remember that the 'average' 600 oz*in motor connected with the unipolar or half-coil configuration will produce about 800 oz*in if connected with the Series configuration or the Parallel configuration. The Parallel configuration requires an 8-wire motor and the Series configuration has limited 'corner speed', so, except for 8-wire motors, I use the Unipolar (half-coil) wiring configuration.

I've never experimented with gas shocks, but I know that they are very popular when used in applications exactly like the one you're doing. The same type of system is part of almost every vehicle that I've owned in the last ten years to hold the hood in an open position.

The precision ground ball screw will multiply the motor's torque significantly. Many ball screws have a 5:1 pitch, meaning that the screw revolves 5X for every inch traveled. That also means that the torque will be multiplied 5X. So a 40 lb*in motor becomes a 200 lb*in motor (or a 3000 oz*in motor). It also means that it will take 10,000 step pulses to move one inch since a Gecko stepper driver needs 2,000 step pulses to rotate a stepper's shaft 1 time and with a 5:1 ball screw, 5 X 2000 = 10,000. Speed will be affected, but I would guess that you will still have more than enough speed for a Z-axis.

One last comment about motors. There are a lot of good stepper motors available and a lot of bad motors available. Shopbot uses Oriental Motor. In my experience, Oriental Motor is thee only manufacturer that I will buy from, unless I decide to spend considerably more for a motor. There are other, more expensive, motors available, but they cost 3X to 5X more than the Oriental Motor steppers. I have experimented with a few 'low cost' steppers. My experience with those 'low cost' motors has made me decide to not use them.

[waynec]

Mike,
Thanks so much for the very useful explanation. A follow up question or two if you don't mind.
1) You say the Gecko driver needs 10K pulses to move a 5 pitch screw 1 inch. How long does the Gecko take to generate and deliver this many pulses? I assume it happens on a computer timescale- pretty dern fast. And do I also get better resolution because of this high pulse count?
2) About inertia. I'm getting ready to order some 8020 extruded aluminum for my gantry. In order to get the deflection to a reasonable amount, I'm risking a heavier gantry. The weight will be about 160 lbs with the router. At what point do I need to consider inertia of this much moving mass? Deflection with this configuration will be .007" with a single point 100lb load. I can save lots of weight if I tolerate a .04 deflection at the same weight. What would you recommend? 160 lbs doesn't seem that bad to me with the right motors.

BTW, I'm cutting softwoods and sheet goods for furniture and cabinetry and hopefully some signage. I'm hoping I can rebuild my PR to be reliable and accurate enough to forestall buying a new machine.

Thanks for the help
w

[richards]

Wayne,
You can figure that the controller will generate at least 20,000 steps per second and probably 1.5X to 2X that rate, depending on whose controller you use and whether you're using USB/Serial, Parallel or TCP/IP. Unless I'm ramping my Z-axis into the material, I never plunge faster than 1-ips, with 0.5-ips being more common in my shop. So, if you're running similar speeds, a controller that puts out 20,000 pulses per second is 2X faster than what you would be able to use.

Since you'll be using 10,000 pulses to move one inch, your resolution will be 1/10000th of an inch per step. That's about 20X better than the resolution that I'm getting (and I'm very pleased with 0.002-inches per step on my machine. Of course, having a possibility of 0.00001 inch per step and actually getting that kind of resolution might be two totally different things.

I'm afraid that I don't have the proper software to run a simulation on moving mass with a stepper. My general finding is that if the stepper can actually move the mass, you will be able to set the ramp speed to accelerate the mass within the limits of the stepper. It's kind of like a 200 lb. man pushing a 4,000 lb. vehicle. If the road is level and the man has traction, he can usually push the vehicle. Once the vehicle is rolling, the job becomes a lot easier.

When you consider the weight that the stepper motor actually moves, you have the weight of the router/spindle, plus the weight of the mounting bracket, plus the weight of anything else that is attached to the mounting bracket. The weight of the ball screw, the linear guides and the Y-car itself will be handled by the Y-axis motor and the gantry will be handled by the two X-axis motors. Of course you'll have to add in the friction of the ball-screw (minimal) and the 'stiction' of the linear rails (probably minimal) and the resistance of the gas shocks (probably moderate). There are a lot of potential pitfalls, but if you visit places like the CNCZone and other sites where hobbiests are adding CNC controls to vertical mills, I believe that you'll find that the weight that you're expecting to move is much less than the weight of some of the axes on those kinds of machines.