[bdillard]

I need to build a power supply to power some 70 volt servo motors for my cnc router project, and I'm trying to get my ducks in a row prior to buying the components. Not having built a power supply before, any input from experienced/knowledegable readers is welcome.

I have a variable transformer that outputs 0-140 VAC and is rated at 4.5 amps and .63 KVA that I am considering (primarilly because I acquired it at auction with some other stuff some time ago, and it is not currently being used) so that I won't have to spring for a toroidal transformer. I'd like to get the full rpm from the servos. They are rated at 2.27 amps at Tc stall. I hope to power two of these servo motors with my power supply.

First, does it make sense to use a variac (since I already have one), perhaps mounted externally so that I can gradually increase the voltage? Second, does the aforementioned variac sound up to the task of powering the two servos? Third, should I stop now and go back to school?

All knowledgeable responses very much appreciated.

[Al_The_Man]

Here is previous posts on using a variac supply for servo projects .
http://www.cnczone.com/forums/showth...ghlight=variac
Not always a safe method due to the variac having no isolation from the incomming AC supply.
Al.

[bdillard]

That sums it up pretty succinctly. No variac for me.

I guess everyone is right: Al is the man.

[gar]

060219-1611 EST USA

bdillard:

You do NOT want to use the output of a Variac without isolation for your application. In other words you need another transformer between the Variac output and your rectifer circuit. This transformer for your power supply must have its secondaries electrically isolated from its primary.

There are various ways to build transformers. One is called an auto-transformer. An auto-transformer secondary shares some of the same wire as the primary and thus there is no electrical isolation between primary and secondary. The Variac (trademark of General Radio) is a variable auto-transformer. This is a current limited device rather than volt-ampere limited.

Yes a Variac in front of an isolation transformer can be very useful for you.

Yes you and many others need to find some books on the basics of electrical engineering. The "Radio Amateurs Handbook", and maybe some US Navy Electronic Technican training manuals, might be starting points as well as high school and college physics books. Understanding of fundamentals and how to ask the correct questions, and test these questions is of vital importance. There are far too many un-necessary questions asked on this forum because the questioners do not have a fundamental understanding of electrical circuits, components, and physics.

On your Variac question, without knowing anthing about the device, you could use an ohmmeter to determine that there was no isolation between input and output. Also if you study and understand the electrical characteristics of a diode, and then analyze how a bridge rectifier works, then you will understand why you would be in deep trouble putting a bridge rectifier directly on the output of a Variac.

.

[bdillard]

OK. Basically I'm lost. I think I know how they all go together, but I do not know how to size the components for an appropriate PS. I have searched unsuccessfully for a site that will guide me on selecting a toroid, bridge rectifier, capacitors, fuses, etc of the correct parameters to power my servo motors. Any takers?

I will be using Geckos driving servo motors (eventually four) with the following specs:

NOM MIN MAX
Torque, peak stall - - 300
Torque, continuous 52.20 46.98 57.42
Torque sensitivity 23.00 20.70 25.30
Armature resistance 3.15 2.68 3.62
Armature inductance 6.62 4.63 8.60
Amps at Tpk - - 13.04
Amps at Tc stall 2.27 2.04 2.50
Volts back EMF 17.02 15.32 18.72
Volts at Tpk 43.09 - -
Static friction torque 4.5 - -


Max terminal voltage = 70VDC
Max sustaining voltage = 70VDC @ 3627 rpm @ 34 oz-in.
Max no-load speed = 4113 rpm.

RPM Trated EMF Current
0 52.2 11.7 2.4
363 51.0 17.6 2.4
725 49.6 23.5 2.3
1088 48.2 29.4 2.2
1451 46.6 35.3 2.2
1814 45.0 41.1 2.1
2176 43.2 46.9 2.0
2539 41.3 52.7 1.9
2902 39.2 58.4 1.8
3264 36.9 64.1 1.7
3627 34.0 69.7 1.6

[bdillard]

Thanks, Gar, for your subtle admonishment.

Yes, I admit I am informationally challenged when it comes to electronics. I am at the learning stage and will not undertake to actually build the PS unless and until I get confirmation from someone qualified to say "Yup, that is how it is done." Your response is a valued part of my education. If, in the final analysis it makes more sense for me to buy one, I'll shell out the cash, or pay someone qualified to build what I need. Regardless of which route I ultimately choose, I want to know how it is done; the appropriate components, and how to determine their appropriate capacities to achive the desired outcome.

Thanks again.

"Yes you and many others need to find some books on the basics of electrical engineering. The "Radio Amateurs Handbook", and maybe some US Navy Electronic Technican training manuals, might be starting points as well as high school and college physics books. Understanding of fundamentals and how to ask the correct questions, and test these questions is of vital importance. There are far too many un-necessary questions asked on this forum because the questioners do not have a fundamental understanding of electrical circuits, components, and physics."

[gar]

060219-1744 EST USA

bdillard:

To size components you need to look at your total system from an energy point of view (power --- power is the rate of expending energy).

What is the instantaneous peak power required by the load, what is a short time power requirement, and a long time average. The thermal capabilities of various components will determine which of these time periods are important for the particular component. For example transformers, motors, and copper wires can tolerate very large overload for a short time. Also large voltages for a short time.

Input power is related to output power by the efficiency of the system. This may not be the same at different power levels.

So for example if your mechanical load was 1 HP (746 Watts) and the efficiency of the system was 60%, then the required input power from the AC line would be 1243 Watts. Suppose the power factor was 0.8, then the VA input would be 1554, and at 120 V your current is about 13 Amps. In any rectifier circuit with a capacitor input filter the AC line current will be non-sinusoidal even through the source voltage is a sine wave.

Diodes, and other semiconductors have a very short thermal time constant in comparison to transformers, motors, and wiring. Semiconductors have critical voltage, current and temperature limits. Thus, you design for peak instantaneous values for these devices. Transformers, motors and wire allow more averaging.

You really need to evaluate the combination of the drive and motor together with your expected load. Until you can define your mechanical load anything is a guess.

It may not be obvious, but regenerative dynamic braking may be a different consideration for you. In other words stopping a lot of mass quickly.

.

[gar]

060219-2105 EST USA

bdillard:

The large capacitor at the ouput of your DC supply is an energy storage device. As such it can moderate the peak power required from the AC line and it can absorb energy from braking the motor load.

As a starting point disregard the specific motor specifications you listed and treat the controller and motor as one black box, This assumes you are not going to overload either component of the black box.

You may not have any easy way to determine what your actual servo mechanical load will be, but try to estimate in watts. Then assume an efficiency of motor and controller. Determine the watts required for this condition. Then knowing your power supply voltage calculate the current. Make some estimate on how much simultaneous power all your motors will require and calculate this power. Then make a judgement on how large to make your power supply.

.

[gar]

060220-2157 EST USA

bdillard:

More input for you.

One horsepower will vertically move 550 pounds one foot in one second. One foot/second is 60 feet/minute, and 720 inches/minute.

If you can judge your maximum desired cutting force in one axis, and define your maximum feed rate at this load, then you can determine the required horsepower to feed the part. This assumes no acceleration. Acceleration will require additional horsepower.

HAAS mentions on small machines, VF3, a peak axis capability of 4 horsepower. But this is for a very short time.

Suppose we take my previous example of 1 HP and assume it is your maximum for all axes, and the 60% efficiency. As viewed from the filter capacitor lets assume the effiency is 70%. Thus, power from the capacitor point in the system is 746/0.7 = 1066 watts. Next assume your power supply provides 60 V DC average and the capacitor is very large. Then for a 1 HP on your motor you need 1066/60 = 17.8 A. A realistic bridge rectifier would have 35 A diodes.

The VA rating of your transformer should be at least about 1066/0.8 = 1332 VA.

Possibly you should use an AC input fuse at 120 V AC of 1500/120 = 12.5 A. Probably slow-blow. When you select a type and size of a thermal fuse you have to think about inrush current, the characteristic curve of I squared * T, and tolerance. You may want an negative temperature coefficient thermister in series with your primary to limit inrush current. Thermal and magnetic fuses are not precise devices. You might need something larger than 12.5 to avoid erroneous blowing. Also you have to work with standard ratings.

.