[CaspianCmonster]

Hi everyone, I am designing a small CNC lathe attachment that will go on my Schaublin 102 lathe. (Mid size (102mm from bed to center and 400mm between centers) If you don't know the machine they are accurate Swiss tool room/instrument lathes that do not use a lead screw. Anyway I will keep the manual functions and add a CNC option with a totally separate system for cutting threads etc. The machine has some nice features such as full length T-slots on the the top and both edges of the bed. The mechanical side of things is under control.

For electronics I have ordered a pair of G723-400-4 (5amp 400oz-in Gecko motors and a pair of G203V drives and will be running with a 48v power supply. I also have a Mesa 5i25+7i76 for interfacing with the drives and plan on using Linuxcnc.

I am kicking around a number of different designs but need to pick a motion solution. My preference is to run or at least have the option of running the steppers directly on screws (so what I am calling a 1:1 drive ratio)

Looking at the documentation It seems clear I don't want to spin the steppers too fast while cutting (say less than 17Hz (1000rpm). This isn't the issue, the question is what is the minimum stepper speed I should be aiming to cut at? I am far more concerned with finish quality, accuracy etc than production rate as this is just for personal use. Once I have this detail I can figure out the ideal ratio between the stepper motor and movement but I can't run my optimization without this lower bound. Reading the Gecko site it looks like maybe this lower bound is 5Hz but I would like to get more info.

The drive design (at least on the Z axis) is open so I can use lead screw, rolled ball screw, belts etc. I am not going to pay for ground ball screws but everything else is open.

thanks

Luke

[doorknob]

Looking at the documentation It seems clear I don't want to spin the steppers too fast while cutting (say less than 17Hz (1000rpm). This isn't the issue, the question is what is the minimum stepper speed I should be aiming to cut at? I am far more concerned with finish quality, accuracy etc than production rate as this is just for personal use. Once I have this detail I can figure out the ideal ratio between the stepper motor and movement but I can't run my optimization without this lower bound. Reading the Gecko site it looks like maybe this lower bound is 5Hz but I would like to get more info.

While I'm not sure that I have the answer to your question, I am a bit confused by the units that you are referencing.

You state that 17 Hz is 1000 rpm. Is it possible that you arrived at the 17 number by dividing rpm by 60 in order to get revolutions per second? If so, revolutions per second would not normally be stated as "Hz" (cycles per second), as that is not a unit that is typically used to describe stepper motor motion.

Your motor has 200 steps per revolution (when using full steps - my understanding is that the Gecko G203V driver operates with 10X microstepping, or 2000 microsteps per revolution).

200 steps per revolution at a speed of 1000 rpm translates into roughly 3333 full step pulses per second (or multiply that by 10 for 10X microstepping to get the step pulse rate).

I did not find a torque vs. step rate plot for your stepper motor, but 3000 full steps per second would be on the high side for a motor with a torque-speed curve like this one:

http://www.kelinginc.net/KL23H2100-30-4BT.pdf

(Note that the chart in the link above shows pulses per second and/or rpm when operating in half-step mode - which would be equivalent to 6000 half steps per second at 1000 rpm, which is off of the chart.)

So, if the torque-speed curve for your motor is anything like the one linked to above, 1000 rpm is a very high speed (and there are various factors that may prevent such a motor from reaching such a speed).

As for minimum speed, stepper motors can easily handle very low speeds - for example, one step per day (or per month) with no problems. So the minimum speed limit is not going to be motor-dependent, rather it is probably going to depend on the cutting tool geometry, the spindle speed, the leadscrew pitch, and the workpiece material characteristics, as far as getting the desired finish.

But then perhaps I have misunderstood your framing of the question.

[CaspianCmonster]

Hi lets see if I can clear things up, I like to think in Hz for a few reasons, To start it is a proper unit (seconds are the SI unit of time, not minutes) also if you look at the Gecko webpage they also use revs or steps etc per second in their more technical documents and charts so it makes using and understanding those more direct. If is easier to use RPM then no worries.

Here is the motor torque curve http://www.geckodrive.com/catalog/pr...d/25/image/62/ that is for half steps so 400 steps per rev. For a working range I would revise what I said before and say that somewhere around 5Hz (2000 half steps per sec) or 300 RPM which gives about 200 oz-in might be a good max, higher speeds would be fine for unloaded rapids. I don't know if that is exactly correct as that chart shows the data for half steps not micro steps with morphing etc but it should be close anyway.

I went and revisited this document: http://www.geckodrive.com/support/ap...ing-works.html I think I missed something important the first time I looked at it, the issue with 4 or 5 rev per second is not that I want to be above that but rather than micro steps are an advantage below that speed. This would imply that the real limit for stepper speed lower bound along with getting sufficiently high max speeds is the positional resolution and that I will have plenty of power at any usable rotational speed no matter how low. For the sake of the discussion if I want a 0,001mm Z axis resolution I need an effective drive ratio that will give me no more than of 2mm travel per stepper rev. (2mm/2000 steps=0.001mm or probably a real stepper resolution of 0,002mm but anyway) This means either I use reduction gearing of some sort or a very fine lead screw to get this sort of resolution. For this project if I use high-end trapezoidal screw or inexpensive ball screws it looks to me that a pitch under say 4mm or 5mm per rev. which would give plenty of accuracy with direct drive and have a large usable range of cutting speeds.

I suppose screw efficiency comes in here as well when picking the lead screw pitch, if I use a ball screw with a higher pitch and high efficiency (90%) and pitch of at least 5mm rather than a tighter pitch 2mm lead screw with lower efficiency (>50%) the usable torque from the motor could be the same or even higher with a ball screw although the resolution would be lower. The ball screw also has better pitch tolerance and probably lower backlash so a system built with a ball screws can better utilize the high resolution and because the efficiency is higher can provide the same force with a lower input torque (higher stepper speeds). So really with ball screws on a lathe you would want to run them at a higher reduction than lead screws to see the full benefits. I suspect that on a mill the feed speeds are so much higher than this becomes a non issue. From the point of simplicity, cost and ability to easily machine the ends of lead screws they make sense for this project but it is hard to make the call.

Does this all sounds reasonable?

thanks

Luke

[bhurts]

I'm no expert on steppers but I think you may be overestimating there usability at high speeds. Steppers are rated for holding torque at there rated current. When you run them the faster you go the weaker they get. That means the more you gear them down the faster you have to turn them to reach a desired feedrates. If you run them at those higher speeds they will have little to no usable torque and they will stall.

Ben

[RCaffin]

I am not sure that you can have high resolution AND high speed at the same time out of a simple stepper drive. Sounds more like a expensive servio motor with a rather high encoder count, plus maybe a very high speed encoder signal divider which can be switched in and out. More $$.

Cheers
Roger