Quote Originally Posted by Torchhead View Post
True unipolar connection is where the full + volts are wired to the center tap of a winding and the two legs are pulled sequentially low . its sometimes referred to as a Push-Pull circuit . The unipolar driver is easier and cheaper to make and the early ones just used a big series resistor in the + leg to limit that current. With the cost of chopper stepper drivers now almost as cheap as the unipolar drivers and the distinct advantages you get with a bipolar setup there is no logical reason to try and move back to the older technology.

The other factor you are not thinking about is the loss of torque with RPM . Just because you can spin a motor faster does not mean you can use all that extra RPM. Yes, inductance does predict the slope of the torque curve (and where the torque "knee" happens ) but as stated you have a lot of connections options with an 8 wire motor. A true unipolar motor only has 6 wires: 3 on each winding.

The correct question is what level of torque and RPM go you NEED. You can get a ballpark on paper but there are too many variables to get really accurate ones. The max theoretical RPM of a motor is a combination of many things including the applied voltage, torque curve at that voltage , series or parallel connection and even the "purity " of the pulse (step) signals. Then you have "microstepping" and step morphing that changes the dynamics of the RPM / torque curve at higher RPM (some drivers have it, some don't) . We have seen a 25% increase of top RPM just from moving from a PC generated step to a pulse train generated by a separate pulse card.

I use the 1/2 rule. You can usually get about 50% of the max numbers in real live moving a load. Its conservative but when you start to push boundaries you start to see problems. With 1605 ballscrews your linear feedrate will be 1/5 of your motor RPM. That RPM has to be at a point where you still have enough torque to both accelerate and decelerate the motion of the toolpath and handle the inertia of the total mass of the object you are moving and any cutting forces .

So---- at 48VDC you can expect to see from 800 to 1000 RPM unloaded motor RPM before it self stalls. USABLE RPM can be 1/2 that . Lets take 500 as a good number. Put that though the 1605 and you are at 100 IPM . Torque and speed and on opposite sides of the equation so while you can get more RPM that often results in less torque (especially through a transmission) so its faster, but won't pull a sick cat out of a litterbox.

The best advice I can give is to look at designs that are similar to what you want to do and use that as a baseline. As they say "your results will vary". There are no simple formulas to predict the overall performance of a set of electronics and motors on a specific mechanical structure
Firstly, thank you very much for answering my question. It's also much appreciated the detail you've gone into and the time you spent to type it.

I did suspect Uni-polar was something like you have described but I couldn't find a definitive answer. I didn't know anything about the resister on the "tap" and couldn't find anything about the "tap" being ground or positive. From your advise, I'll dismiss any thoughts of using a Uni-polar setup.

I am aware of the more speed means less torque thing. It was primarily for rapids that I started to think what I had was too slow. The first motors I had intended to use gave me a theoretical speed of 30IPM (I usually work in metric. Aussie). In my build I initially wasn't too worried about speed, I was more worried about having enough torque hence the 1605 ball screws instead of 1610 or higher. That and they were easier to find and cheaper. This was also prior to knowing about feeds and speeds. I thought that I could just cut slow. Now I know too slow can be as bad as too fast. The first motors that I had intended to use were over 9mH so max RPM was way down. They only came with 36v power supplies as well. I'm still not concerned about the pitch of the ball screws being too slow. I figure I will use toothed belts from the motors to adjust ratios if it doesn't work out. I am trying to give it best chance I can in the first place though. There has been plenty of research done. I have found about 2NM is a good starting point. I haven't read much on speed though. What started out as being a sub $1000AUD build is at $1200 with rails, ball screws, motors, power supplies and drivers. Now the decision has been made to go with Bi-Polar Parallel, I need to find an appropriate driver. The MA860H is the current choice but I am still researching

Thanks again for the replies.

Happy Days