In this video I’d like to describe the

use of power meters for measuring load on the spindle and stepper motors of a

CNC mill. This is an update of a previous video where I used a current and voltmeter to estimate the power. The previous approach

also measured apparent power instead of actual power. To understand the

difference will have to examine some definitions. Mechanical power is measured

in horsepower while electrical power is measured in watts. One horsepower is

defined to 746 watts. To measure the power delivered by an electrical motor

we’ll measure its consumption of watts. Power going out can’t be greater than

power going in and we’ll estimate the load on the motor as a ratio of its

delivered to its rated horsepower. Unfortunately since there’relosses in the

motor, belts, pulleys, bearings, and so on, the system is not 100% efficient and not all supplied power will be delivered at the spindle.

Electrical power can be divided into actual, apparent, and reactive power.

Actual power is what the electrical company charges for and, over time,

measures the work done. Actual power is measured in watts. Reactive power is power that does no

work. It is power borrowed from the electric company to charge electric

motors coils and it’s later returned. Apparent power is the total power

delivered including actual and reactive power.

Apparent power is measured in KVA or kilovolt-amp and is used to size

wires and circuit breakers. Unlike watts, basic multimeters can be

used to calculate kilovolt amps by measuring and multiplying RMS voltage

and current. To measure watts a specialized meter is required. The ratio between actual and apparent

power, or the ratio between useful and total power, is called the power factor. Ideally, a load would have a power factor

of one. Where all delivered power is used. This is ideal because of benefits for

production and distribution electrical power. As the motor’s load increases its

power factor increases. The motor becomes more efficient to a point. For a milling

machine we can use load as an estimate of reserve capacity, and to guide us to

take heavier or lighter cuts. While our previous meters behaved like a

multimeter, where we measure current and voltage separate, and multiply them

together to produce apparent power, our new meter is able to measure these

plus actual power, in terms of watts, and power factor. These meters are produced by Murata and

consists of a panel interface with a selection button. On the back, there is a

current transformer and screw terminals to attach to the mains. Current is measured

by induction from a wire that passes through the current transformer, while

voltage is measured at the screw terminals. A microprocessor inside of the

power meter is taking into consideration the phase difference between the voltage

and current waveforms. From this it is able to measure watts and power factor. The top

meter will measure the spindle motor while the bottom meter will measure the

stepper motors. Both meters are fused with an external half amp fuse. There is no internal fusing. Here the

spindle motor is a DC motor and the current is measured as its passed to the

driver. Power delivered to the stepper motors is measured as power delivered to

their switch mode power supply. A small sub panel for these meters was cut from

wood on the CNC mill. This graph measures watts and the volt-amps delivered to the

spindle motor as various RPMs. It also shows the power factor as the

RPMs are changed. The spindle was turning a half inch drill bit at the time. A repeat of these measurements without any

tool inserted into the spindle showed no difference. This graph might be useful for

describing the power necessary to simply turn the spindle. In that case, at 2,600

RPM, approximately one-tenth of a horse power is necessary simply to turn

the spindle. For most of this graph the power factor is between 0.5 and 0.6. However, under significant load, the

power factor will increase to 0.7 to 0.8. If the power factor remained within

this tight band between 0.5 and 0.6, it may be reasonable to use a current

meter as an indication of load. However using a power meter removes all

doubt by removing the reactive component of apparent power. While a power meter may be a better

approach for measuring power, and thus load, over a current meter, in practice, as

we’ll see, they each have their advantages. The milling machine that the power meter

is attached to is a modified Precision Matthews 25-MV (PM25MV). The spindle motor is a

one horsepower DC motor. The maximum speed of the spindle is 2600 RPM. The

advertised drilling capacity is one inch, but to drill 1 inch without step

drilling is overly optimistic. In some CNC mills the drilling capacity is

not limited by the spindle, but instead by the thrust capacity of the axis motor.

With this CNC mill, horsepower delivered to the spindle is the limiting factor. An

advantage of the current meter over the power meter is faster response time. The

power meter is only update every second or so. In this case, the spindle motor was

not able to deliver the necessary horsepower to drill the hole and

ultimately stalled. The current meter spiked to 15 Amp or 1,800 volt-amps of

apparent power at 120 volts. The power meter spiked to approximately a thousand watts. I had hoped to use a power meter as an

indication that the spindle is about to stall, but the slow response time of the

power meter may not allow this. The combined power consumed by the three

stepper motors is rarely above 70 watts. I found this to be surprisingly low and

if it turns out to be representative there may have been no need to install power

meter for the stepper motors. The second drilling test uses a slower spindle

speed and feed rate. While the spindle is under considerably

less load this is not an efficient drilling operation. The stringy chips in

addition to being dangerous are representative of the feed rate that is

too low. We can estimate the horsepower

necessary to drill a half-inch hole in the aluminum at 1000 RPM. These

equations and constants can be found in the Machinery’s Handbook. While this method has been derived from

the experiences of drill operators over decades, it is important to recognize this is

still an estimate. We’ll first estimate torque necessary, then power necessary of

the cutter, and then power necessary at the motor. The constant of 63,025 is

the torque necessary to produce one horsepower in one revolution. And one

horsepower was originally defined as the energy required to lift 33,000 pounds

one foot in one minute. It was an estimate derived from experience

with horses. Needless to say the SI units for power are a

lot cleaner. Inserting the constants into the equations, we arrive at 11.3

inch-pounds of torque required. This is 0.179 horsepower at the cutter

and assuming an efficiency of eighty-five percent 0.211 horsepower at

the motor. Given a horsepower is 746 watts, 157 watts are necessary to drill a

half-inch hole in aluminum at 1000 RPM. On this mill, 35 watts are necessary to

spin a spindle without a tool. This is the spindle’s overhead. Adding 35

watts to the 157 estimated watts, we arrive at 192 watts of estimated power necessary to drill a half-inch hole in aluminum at

1,000 RPM. A rough average of the measured power is 200 watts. These two values are surprisingly close

and this experiment will be repeated to verify consistency between estimated and

actual horse power requirements. To make measurements of power truly

useful they need to be captured by the control computer. And unfortunately these

meters have no means to feedback information to the computer. Because of

the slow update rate of these meters, they may be more useful for steady state

operations than as an early indication of spindle stall. I have a few other ideas of how these

meters might be useful and if those ideas work out I’ll make videos on those

topics as well. If you have ideas for experiments or noticed a mistake that

I made please make a comment below. I hope you enjoyed this video. It has taken

the longest of any of the videos I’ve made to date. If you liked the video

please hit the like button and consider subscribing. More videos are coming. Until

the next video, thank you for watching.

Good video. Im trying to work through these calculations too, I like the meters idea, think I'll get some of them too. Thanks

interessante!!!

Check out my most recent video: https://goo.gl/Jj7cU1

Awesome video and great explanation! I've been looking for a power meter like that to estimate forces on a CNC router. Those meters are surprisingly accurate – thanks! Too bad they don't have any sort of external interface. This other approach looks too hard and expensive http://www.ti.com/lit/an/slyt545/slyt545.pdf

It appears that your high speed drilling example (2600 RPM? and IPR?) stalled due to inadequate spindle power. Is that right?

Are speed changes on that spindle accomplished via drive belt pulley changes or the motor controller?

If Main Spindle Power of CNC1 is 5.5/7.5 kW and CNC2 is 9/11 kW, a same machining job is performed on both the machines, will the power(energy) consumed by both the machines will be same ? Or if different, then approximately how much difference ?