Stepper Motors

Fast slewing speeds are highly desirable, for obvious reasons. When I first built this scope, I was getting less than 1 degree per second slew rates when running 12 volt motors on 12 volts. Any attempt to go faster resulted in a stalled rotor.

The usual procedure to get more speed is to run the steppers at 2x or 3x their rated voltage. It seems that this approach must have worked fine for some people, but in my case I ran into a major problem. The problem was rotor resonance which occurred at slow to medium speeds because the motor was being driven beyond its current limits.

My recent experience with high speed stepper drives for small CNC machine tools has revealed that resonance is a common problem when steppers are driven beyond their rated voltage if a current limited source if not used. The effect of rotor resonance is that the motor runs very rough and then stalls in the low to medium speed range, with the problem only getting worse if more voltage (and current) is applied.

If we limit current to a point less than the maximum ratings of the motor, resonance will not occur. As the speed of the motor increases, its effective impedance increases. In order to maintain a constant current, we need to dramatically increase the voltage across the motor windings as it spins faster.

There are several ways to accomplish this. The most efficient method would require a chopper drive to vary the duty cycle of a high voltage source in order to maintain the desired average current through the motor. It's important to note that Mels program allows you to tune the current drawn by the motors in microstep mode using this same technique. However, in half-step mode, we need to be able to sense current and then change the duty cycle on the fly. A fixed pulse width factor would not do us any good.

The quick and dirty fix is a series current limiting resistor for each motor. At slow motor speeds a large percentage of power is wasted and is dissipated by the limiting resistors. However, slow speeds in half-step mode are very brief and are only encountered during ramp-up and ramp-down modes. We can afford to be a little inefficient for these brief periods of time. The speed ramping is normally required in order to avoid slippage of the drive, or stalling problems that could occur if tried to instantly move the scope from a dead stop to full speed.

The actual value of the limiting resistors depends on several factors, which include the current and voltage ratings of the motor, and the voltage of the supply that we are going to use. In order to approach the highest speeds that are possible from most stepper motors, supply voltages of not less than 5x the voltage rating of the motor are required. It's not uncommon to see significant gains in maximum RPM with voltage levels as high as 10x that of the motor ratings. There is supposedly a general rule of thumb that doubling the supply voltage (within reasonable limits) will result in a 50% increase in the maximum speed that can be attained.

The resistance value of the series limiting resistor can be calculated with this equation:

  
                 Ro = ((Vs - Vd) - Vm) / Im

WHERE: Ro = Series resistor value in ohms.
       Vs = Motor coil supply voltage, larger than Vd + Vm.
       Vd = Voltage drop from transistor & diode, about 1 to 2 volt.
       Vm = Rated voltage for motor when stopped.
       Im = Rated current (amperes) for motor when stopped.

                 Rw = ((Vs - Vd) - Vm) * Im

WHERE: Rw = Series resistor value in watts.
       Vs = Motor coil supply voltage larger than Vd + Vm.
       Vd = Voltage drop from transistor & diode, about 1 to 2 volt.
       Vm = Rated voltage for motor when stopped.
       Im = Rated current (amperes) for motor when stopped.

With this information in mind it becomes clear that stepper motors with very low voltage ratings (Vm) are highly desirable. It turns out there are many stepper motors available with voltage ratings of around 2 volts or less. There is a reason why these motors are designed with such low (locked-rotor) voltage ratings - it allows the use of a reasonably low supply voltage (Vs) to achieve a wide operating RPM range.

In an actual application with a 12 volt battery as a power source, the optimum stepper motor would have a voltage rating between 1.2 and 2.4 volts. If you are using motors with voltage ratings between 2.4 and 4.8 volts, then the best performance will probably be achieved by using two batteries in series for a supply voltage of 24 volts.

One good source of surplus stepping motors is C&H sales (800-325-9465). Here are two motors that are listed in their catalog that should provide outstanding speed and performance if current limited in the manner just described:

SLO-SYN #MA-61FS-80078 - rated at 1.4 vdc/3.9 amps per phase. C&H stock #SSM8954. Cost $25 each.

RAPIDSYN #23D-6106FA - rated at 3 vdc/1.5 amps per phase. C&H stock #SSM8504. Cost $30 each.

Both of these are NEMA 23 frame 8-lead motors, easily connected as a 6 lead.

The SLO-SYN motor, when powered with a 12 volt battery, would need a current limiting resistor of about 2.5 to 3.0 ohms. The calculated wattage rating would only be required for constant operation at very slow speeds, a condition that we normally don't have. A 10 to 15 watt rating should be adequate in our application.

The RAPIDSYN motor, when powered with two 12 volts batteries in series (24 volts) would require a current limiting resisitor of 13 to 15 ohms. Once again, a 10 watt rating should be sufficient.
When the motors are running at full speed in half-step mode, very little power is dissipated by the resistors - the vast majority is absorbed by the motors.

Running stepper motors at high voltages or using motors that have very low winding resistance can place heavy demands on the power output devices (darlington transistors) and the power supply/batteries. Be careful to not exceed the ratings of the devices. If in doubt, use multiple devices in parallel to increase current capacity, or use a different device with higher ratings.

Cabling and Wires

This is an area that deserves careful planning. Wires are very easy to trip over in the dark, even when you know they are there. If there will be people walking around your scope who don't know about the wires, you are almost guaranteed that people will trip over them. I've had my computer knocked to the ground, cables ripped out while the system is running (requiring a complete re-alignment of the scope) and lots of embarrassment and apologies.

Here as some of the things I've tried, plus some ideas that I plan to implement on the 20 inch scope.

1) Run all cables through a hole in the middle, or near the azimuth axis of the scope. If you have wires running to things directly on the side of the mount, you will have problems with wires getting tangled or pinched in the azimuth drive as the scope is moved in a complete circle. Running all cables through a hole in the bottom of the center of the scope will allow it to be turned several rotations without getting things all wound up.

2) Put all wires going to and from the scope in ONE harness. The fewer cables you have, the easier it is to keep them from being tripped over.

3) Keep this harness of wires a least 12 feet long. I personally feel that 18 feet is much better. If you cover the wires with carpet, people won't be able to trip over them as easily. A wide carpet if also much easier to see in the dark than a thin wire harness.

4) Keep the cable between the handpad and scope short enough that it won't drag on the ground. Put a hook somewhere on the side of the scope where you can hang the handpad when you aren't using it.
I made the mistake of running the handpad cable directly from the back of the computer with a separate and very long wire. I had nothing but trouble with this. Keep the handpad cable reasonably short, and run it directly to the side of the mount.

Don't underestimate the importance of keeping wires out of the way.

Gear reduction thoughts

Until recently, most people were using some type of gearbox in order to get the necessary gearing reduction.
Finding affordable gearboxes has been a problem for many folks, there doesn't really seem to be anything good that is much cheaper than about $150 each.

A problem with almost all gearboxes is that they have multiple gear mesh points, and therefore have noticeable amounts of backlash. Mel's software has backlash compensation built-in, but slop in the drive train can still cause problems. If the scope is used under breezy conditions, it may jostle around a bit. For visual use, it's just a minor annoyance, but if you are attempting any kind of imaging, you might have to wait for the wind to die down.

Ideally, we'd like to get rid of all the backlash. The only simple way that I know of to do this is with a 100% friction drive. In a permanently mounted instrument, it might be acceptable to have two or three large discs and rollers to yield the necessary total reduction ratio (something on the order of 2600:1 is needed for a 200 step/rev motor). For a scope that is intended to be portable, we probably only have room for one large driving disc or plate for each axis. What is needed is a friction drive that can give about 100:1 or 150:1 reduction in a very small package. At this time, I don't know of a design that would be suitable for this application. If anyone knows of a compact friction reducer that meets these criteria, I'd sure like to know more about it!

Although a bit of a compromise, I rather like the idea that Mel and several others have used which involves the use of a single worm and worm gear (only one gear mesh point). The wormgear then drives a friction roller to provide the final reduction. This approach seems like a major improvement over the standard gearbox. It may require the fabrication of a custom wormgear, but it would be simpler, tighter, and probably a lot quieter. Now that I own my own mill and indexing table, I plan to fabricate a pair of lightweight worm gears and use a similar approach on the 20" scope that I'm now working on.

Vibration and rigidity

Vibration problems can be caused by factors other than coarse gearing of the stepper motors. Even when geared for movement of only 0.25 arc seconds per microstep, serious vibration can occur and wreck any chance of getting a stable high-power view.

The biggest potential problem area for vibration I've found is the location of the altitude axis motor. This mounting point needs to be very rigid. If the side boards on your mount are even slightly flexible, you may have vibration problems that will be difficult to solve. The best way to avoid this problem is to design the mount with very low side boards that are at least 1.5 inches thick.

Even with a very rigid mounting point, it is still helpful to isolate the motor from the rest of the rest of the drive train with a slightly flexible coupling. It may also prove helpful to mount the motor itself on a very thin sheet of soft rubber, with the motor housing being secured by nylon screws.