240 feeding 208 volt motors

Status
Not open for further replies.

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150925-1349 EDT

Attached is a graph of speed vs torque for a wound-rotor induction motor with rotor resistance being a parameter. From "Alternating-Current Machinery", Bailey and Gault, McGraw-Hill, 1951, p 248. P 242 has a plot of speed-torque vs applied voltage, not shown here.

PICT3845.jpg


The curve is monotonic with a sufficiently high rotor circuit resistance. Thus, any speed from 0 to near 100% is obtainable by control of the torque load. Or speed changes with torque load, a lighter load has a higher speed.

A fan blade produces a mechanical load that increases with speed. For a fixed high resistance rotor the RPM of the motor will increase with increasing motor voltage. Load power goes up and input power increases as voltage goes up.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150926-1132 EDT

For comparison the measurements on a 115 V 60 Hz 1/3 HP 1725 RPM single phase split phase motor with no mechanical load except bearings and windage are:

Code:
V In      Amp     Watts     VA       PF

100      3.59       97      361     0.26
110      4.16      118      459     0.25
120      4.94      152      591     0.25
130      6.16      211      801     0.26
140      7.59      299     1064     0.28

Ratio Increase per change in input

1.10     1.159    1.217    1.272    0.962     110 to 100
1.09     1.188    1.288    1.288    1.000     120 to 110
1.08     1.247    1.388    1.355    1.040     120 to 120
1.077    1.232    1.417    1.328    1.077     140 to 130

% Increase per change in input

10.0     15.9     21.7    27.2     -3.9       110 to 100
 9.1     18.8     28.8    28.8      0.0       120 to 110
 8.3     24.7     38.8    35.6      4.0       120 to 120
 7.7     23.2     41.7    32.8      7.7       140 to 130


Motor speed was close to synchronous speed because bearing friction and windage do not create much mechanical load, and the rotor resistance on this motor is probably as low as is reasonable to achieve.

I have double checked some of my measurements and calculations, but not all. The current and VA for 140 V were not what I expected, but the measurements seem correct.

Motor acoustic noise seems to increase substantially above 120 V. Note motor is rated at 115 V.

Since speed doesn't change much the increases with input voltage are from gradually going into core saturation and IR losses.

Note: from 100 to 140 V the power input to a constant resistance would go up by a factor 1.4 * 1.4 = 1.96 times, but in the above experiment the power input went up by 3.08 times.

.
 
Last edited:

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Tell me more about Saturation.
Saturation, by definition, for an iron core is when the flux versus current (external field) stops rising more or less linearly because the iron cannot magnetize any farther. At that point you are essentially making a gradual transition to an air core inductor instead of an iron core inductor.
The average inductance is lower so that the percent increase in current is much higher the increase in voltage. The current waveform also departs greatly from a sine wave.

The back EMF remains essentially the same, but the impedance (both inductive and resistive, but with the inductance dominating) goes down, causing the current to increase out of proportion to the voltage.

PS: When you add load to a transformer you do not have saturation effects since the primary and secondary load currents cancel out in terms of magnetic field in the core, leaving only the magnetizing current. The latter will get you into saturation if you increase the primary voltage too far.

Instructive example from the audio world: They make compact adaptor fittings to convert between 150 ohm balanced line and 10,000 ohm unbalanced cable.
The cheap ones can only handle microphone level voltages without saturating. If you try to use the same adaptors for upper end line voltage levels, the output goes way down and the distortion is horrendous.
The real line level adaptors use a lot more iron in the transformers.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150925-1610 EDT

GoldDigger:

I looked at the current waveform, and I did not see a noticable indicator of saturation. The load is highly inductive, as we know it should be, under light load conditions and the low PF also shows this.

However, note, that in my data the current goes up much faster than the voltage. Probably mostly from hysteresis and eddy current losses that grow faster than linear with a voltage increase

The motor full load rated current is 5.9 A at 115 V. At no maechanical load we reach 5.9 A around 128 V.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150925-1635 EDT

ActionDave:

If the motor was mechanically fully loaded, then as voltage is changed the power input would be close to constant because the mechanical load is constant, but modified by what you see in core losses at no load conditions. So at full load you really should see a decrease in current as voltage increases.

I don't have the equipment for an easy setup of an experiment for a fully loaded motor.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150928-0758

ActionDave:

I have been thinking about what you said about current increasing with applied voltage.

In the data I showed above the power input current for the unloaded motor increased more rapidly than a linear relationship with voltage. Voltage ratio was 140/100 = 1.4 and the current ratio was 7.59/3.59 = 2.11 . For a linear load the power ratio should increase by the square of the voltage, or (140/100)^2 = 1.96, but the measured results increased by 299/97 = 3.08 .

In a linear circuit the flux density would be proportional to current, and in turn voltage. Hysteresis losses should vary by about the 1.6 power (exponent) of the ratio of flux density, while eddy current losses (basically resistive) should vary by the square (power of 2) of the flux density ratio.

I ran another test at 90 V. Current read 3.11 A and power read 79.1 A. Between 90 V and 110 V the change in power followed the square-law including a check at the 100 V point. Current was not as expected between between 90 and 100, but was linear between 100 and 110 using two points.

Above 110 V the increases are greater than linear and the square-law. This seems to imply a saturation effect that would cause a reduction in inductance, but my waveform observations did not show that, and power factor doesn't. More to be learned.

A mechanical 1/3 HP load is 746/3 = 249 W. Most of the power input at no load is core losses plus some windage and I^2*R, but windage and I^2*R should be substantially less than core losses.

As load is applied to the motor windage won't change much for this motor. I^2*R will increase with the load. At 115 V no load the power input is 133 W. Add 249 W + some I^2*R and total power input is 382 + I^2*R. Probably about 50 to 150 W for I^2*R.

I really need to do a load test on this motor over the range of 90 to 140 V. I believe over this range we will see, for a constant full load, that the current will initially drop with increasing voltage, but as we get to higher voltages the greatly increasing core losses will cause current to increase with increasing voltage.

.
 
Status
Not open for further replies.
Top