CA1180380A - Motor protection device - Google Patents

Abstract

MOTOR PROTECTION DEVICE
ABSTRACT OF THE DISCLOSURE
A motor protection device is provided that monitors motor currents and motor temperatures and provides protective functions whose characteristics are based on field settable data and motor operating conditions. The protective functions provided are overtemperature, overload, instantaneous overcurrent, ground fault, phase unbalance, and phase reversal.
Field settable data is entered by means of a keyboard and consists of motor trip points for the protective functions, and motor and system parameters.

Description

?3~

MOTOR PROTECTION DEVICE
This application is a division of Canadian Serial No. 358,~40, filed August 15r 1980.
This invention relates to protective systems for elec-tric motors. More specifically, it relates to overload, over-temperature, phase unbalance, overcurrent, and ground fault protection for electric motors.
Unnecessary motor shut-downs from premature relay trips can rob a critical process, industry, or utility of valuable operating time. A relay that does not trip soon enough can result in a heat damaged or burned-out motor which also results in lost opera-ting time. Critical operations that require high reliability drives need motors that have precise protection.
A bimetal type relay provides overload protection against motor overheading under steady state conditions. Bimetal relays, however, cannot accurately reproduce the temperature of motor elements under transient and cyclic conditions of short time durations.
Therefore, bimetal relays do not provide good protection for repeated starts or hot starts and may trip prematurely on cyclic loads.
Most large motors have resistance temperature detectors placed in the stator slots and near bearings.
~5 Resistance temperature detectors (RTD's) are essentially resistance elements of copper, platinum or nickel.
Measuring the resistance of the RTD ' s installed in the motor gives an indication of motor winding and bearing temperatures.
~, 3~3~

The use of a thermal analog of a motor having motor current inputs and winding temperature inputs from RTD's closely approximates the heating characteristics of most motors during both starting and steady sta-te operations. The IEEE paper Thermal Tracking - A Rational Approach to Motor Protection T7~029-5 discusses a thexmal analog using RTD's and mo-tor curren-t inputs that precisely tracks motor -temperature changes precisely as they occur;
rapidly during stall, more slowly during gradual load increases, and takes accurate account of temperatures during cooling.
A relay using an accurate thermal analog of a motor, while a vast improvement over -the bimetal type relay, has some disadvantages. One disadvantage is that lt is not readily adjustable in the field. Many times the calculated motor data, given initially by the motor designer, does not agree with the actual motor data obtained in tests in the field. These differences necessiate adjustment of the relay which requires returning the relay unit to the factory for adjustment or the use of expensive field equipment for adjustment.
Another disadvantage -to the thermal analog relays is that should an RTD open, the high resistance of the open RTD input would then be mistakenly interpreted as indicating a very high temperature which in turn would cause a premature trip of the motor.
An object of the present invention is to provide a motor protection device that has field adjustable set points, for the various functions monitored, that do not require special instrumentation or tools or enter.
Another object of the present invention is to provide a motor pro-tection device that can monitor more than one winding RTD and more than one bearing RTD.
Another object of the present invention is to provide a motor protection device that disregards open RTD's.

;3~3~

SUMMARY OF THE INVENTION
A motor protection device is provided that monitors motor currents and motor temperatures and provides protec-tive functions whose charac-teristics are based on field settable data and motor operating conditions. The pro-tective functions provided are over-tempera-ture, over-load, instan-taneous overcurrent, ground fault, phase unbalance, and phase reversal. Field se-ttable da-ta is entered by means of a keyboard and consists of motor trip points for the protective functions, and motor and system parameters.
The values of the protective functions (bearing temperature for bearing overtemperature protection or percent phase unbalance for phase unbalance protection, for example) are compared to the entered trip points for the protective functions and if the values of the protective tunctions exceeds the trip point, a trip relay is activated.
DESCRIPTION OF THE DRAWINGS
While the specification concludes with claims particularly pointing out and distinctly claiming the present invention, the objects and advantages of the invention can be more readily ascertained from the following description of a preferred embodiment when used in conjunction with the accompanying drawings in which:
FIGURE 1 is a schematic block diagram of a motor protection device in accordance with the present invention.
FIGURE 2 is a functional block diagram of a motor protection device in accordance with the present invention.
FIGURE 3 is the front panel of a motor protection device in accordance with the present invention.
FIGURE 4 is a flow chart showing the Maximum Values of RTD Temperatures program in accordance with the present invention.

~ ~`V ~

FIGURE 5 is a schematic block diagram showing the RTD thermal analog model.
FIGURE 6 is a flow chart showing the Phase Unbalanee proyram in accordance with the present invention.
FIGURE 7 is a chart showing the Phase Unbalanee Trip Characteristic.
FIGURE 8 is the overall flow chart for the motor proteetion device in aceordance with -the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENT
Referring to FIG. 1, the inputs and ou-tputs of proeessor 1 are shown. Starting with the analog inputs, up to six resistance temperature deteetors (RTD's) 2 for measuring winding temperatures and up to two RTD's for measuring bearing temperatures ean be used. RTD's are loeated in a motor (not shown) to be proteeted. Eaeh RTD 2 has two terminals with two leads A and B
conneeted to one terminal and a lead C eonnected to the other. Eaeh A lead is eonneeted to an input of eight to one multiplexers 3 and 4. Eaeh B lead is eonnected to an input of an eight to one multiplexer 5. The output of multiplexers ~ and 5 and all the C leads are connected to an RTD input bridge cireuit 6. Up to 6 winding RTD's are connected to the first six inputs of the multiplexers 3, 4 and 5 while the bearing RTDIs are always eonnected to the seventh and eighth inputs of the multiplixers 3 4 and 5.
The RTD input bridge circuit 6 has a current souree 7 connected through a multiplexer 3 to the A leads.
A resistor 8 whieh has the value of 2R is connected through multiplexer 4 to the A leads as well. The other side of the resistor is connected to the inverting input of an operational amplifier 9. A resistor 10 with a value of R
is conneeted through the multiplexer 5 to the B leads~
The other side of resistor 10 is eonneeted to the non-inverting input of operational amplifier 9. Conneeted to 3~1~

the non-inverting input of opera-tional amplifier 9 and ground are two resistors in series; resistor 11, which has the ~alue of R and resistor 12 which has the value of Ro~
The resistor 12 is connected to ground. Connected to the junction between resistors 11 and 12 are all the C leads from the RTDIs 2. Connected between the non-inverting input of operational amplifier 9 and the ground is a resistor 13 with a value of 2R. The output of operational amplifier 9 is connected to diode 14 which in turn connects to the input of comparator 37a. Feedback resistors 17, one of which is in series with line 18, are connected between the inverting input of operational amplifier 9 and the output of diode 14. A power supply 15 is also connected to the input of comparator 37a through a resistor 16.
Another analog input to processor 1 is obtained from current transformers 20, 21, 22 connected to lines Ll, L2 and L3- Line Ll, L2 and L3 feed the motor (not shown) to be protected. Each current transformer 20, 21 and 22 is connected to an isola-tion transformer 23, 24 and 25, respectively. The output of each isolation transformer 23, 24, and ~5 is connected respectively to a rectifier/filter 26, 27 and 28. All three lines Ll, L2 and L3 pass -through a window-type current transformer 30. The output of the window-type current transformer 30 is connected to an isolation transformer 31 and then to a rectifier/filter 32.
The output of each of the rectifier/filters 26, 27, 28 and 32 are connected to the first four inputs of an 8 to 1 multiplexer 33. The outputs of the rectifier filters 26, 27 and 28 are also connected respectlvely to input 5, 6 and 7 of multiplexer 33 and to diodes 34, 35 and 36. Diodes 34, 35 and 36 are connected to a peak detector 37 which in turn is connected to the eigh-th input of the multiplexer 33. The output of -the multiplexer 33 is connected to a selectable gain amplifier 35a. The output of the amplifier 35a is connected through diode 36a to an input of comparator 37a.

3~3~

The output of comparator 37a is connected to an input To on processor 1. An 8 line output from the input/output port of processor 1 is connected to an 8-bit digital-to-analog converter 38, the output of which is connected to the other input of comparator 37a.
Address latch 39 is connected by means of a bus to processor 1. The output of address latch 39 is connected to multiplexers 3, 4 and 5 as well as multiplexer 33 by four address lines. The address line indicative of the most significant bit has an inverter 42 between the latch 39 and multiplexer 33. The address line indicative of the second most significant bit is also connected to the gain selector of the selectable gain amplifier 35aO
The field setable da-ta is entered by means of keyboard 41 and an ENTER switch 57 which is connected to the read write enable line of a battery powered random access memory (RAM) 44. The latch decoder 40 provides four input lines to the keyboard 41. The latch decoder 40 also provides an enable line to address latch 39.
The keyboard 41 is connected to a tri-state driver 42 which is connected by means of a bus to the processor 1.
The tri-state driver 42 is enabled by a line from -the latch decoder 40.
A program read-only memory (ROM) 43 and the battery powered random access memory (RAM) 44 are both connected to the processor 1 by a bus. A crystal oscillator 45 is connected to processor 1 to provide an internal timing reference.
Lines Ll, L2 and L3 are connected to a J-K
flip-flop 58 and a J-K flip-flop 59 through three wave-shapers 61, 62 and 63, one for each line. The output of waveshaper 61 is connected to the J input of flip-flop 58 and the K input of flip-flop 59. The output of waveshaper 62 is connected to the clock input of flip-flop 58 and the J input of flip-flop 59. The output of the waveshaper 63 is connected to the K input 331~C~

of flip-flop 58 and -the clock input of flip-flop 59. The inverting output of flip-flop 58 and -the inver-ting output of flip-flop 59 each provide an input to NAND gate 64. The output of NAND gate 64 is connected to an input Tl of processor 1 and supplies phase reversal information.
Outputs from the processor are connected -to latches 46, 47 and 48 by means of a bus. Latch 46 is connected to a function dlsplay 49. La-tch 47 is connected to a value display 50. Latch 48 is connected to six fault indicators 51 and to a timer 52. Timer 52 is connected to a normally energized alarm relay 53. Latch 48 is also connected to a tri-state driver 54 which in turn is connected to a trip relay 55 which is normally de-energized.
The alarm relay 53 and the trip relay 55 are connected to external control circuitry, not shown. An enable signal is connected to the tri-state driver 54 from the input/
output port of processor 1. Another output from the processor is a line from the input/output port to an inactive indicator light 56.
The operation of the mo-tor protective device will now be described. First, the overall operation will be broadly described. Referring to FIG. 2, which is composed of functional blocks, the reference numerals used in this figure are used in this figure only. ~ther figures in the specification use like numerals for like elements.
The motor protection device monitors motor line curren-ts by means of current transformers in block 7~ and monitors fault currents by means of a window-type current transformer in block 71. Temperatures of motor windings and bearings are monitored by RTD's located in the motor in block 72. Protective functions are provided with characteristic based on field settable data which is made up of trip points, and motor and system parameters which are entered through the keyboard block 73. The field settable data points are stored in a RAM motor data storage located in block 74. The numerical display block 75 provides visual feedback that the proper key 3~

has been entered and received.
The ground fault functlon block 76 receives data from block 74 and ground fault current input information from block 71. If the ground fault current is 67% of the trip point an alarm condition is given. The ground fault function 76 and the other protective functions are all connected to -the relay drivers block 77 and annuncia-tor block 78. An alarm condition from a function block causes the relay driver to activate the alarm relay in block 79 which is connected to an alarm, not shown. In addition a ground fault indicator light in the annunciator block 78 flashes. r~hen the ground fault current exceeds the trip point a trip signal is given. The relay drivers block 77 is also connected to a trip relay located in block 80. A trip condition in one of the function blocks causes the trip relay to be activated and external control circuits interrupt power to the motor. The flashing indicator light in annunciator block 78 stays li-t to indicate the reason for the trip.
The phase function block 81 which includes phase unbalance, phase loss, and phase reversal functions has inputs from the current transformer block 70 and the RAM
motor data storage 74. When 67% of the trip point is reached, an alarm signal is given. When the trip point has been exceeded, a trip signal is given.
The instantaneous overcurrent function in block 82 has inputs from the current transformer block 70 and the RAM motor data storage block 74. When the overcurrent exceeds the trip point the trip condition exists. Since the device responds to a single cycle of overcurrent no alarm condition precedes the trip. The instantaneous overcurrent indicator in the annunciator block 78 is lit during the trip condition.
The overload/overtemperature functions block 83 has input from the RTD's in block 72, current transformers in block 70, and RAM motor da-ta storage in block 74. A

motor temperature analog is used in the overload/
overtemperature functions and an alarm signal is given when the motor temperature is 10 C. from the trip point. A trip signal is given when the -trip polnt is reached. I-t is also possible to have overload/overtemperature protection of block 83 without RTD input from block 72. Motor tempera-ture would be calculated from just the current inputs from block 70.
10Turning now to the specific elements, FIG.
3 shows the front panel of the motor protection device.
The front panel has a keyboard 41 which has 16 inputs which consists of zero to nine numerals 85, FUNCTION switch 86, VALUE switch 87l DISPLAY
15switch 88, SCAN switch 90, STORE switch 91 and RESET
switch 92. The switches are in the form of pushbuttons.
The panel also contains a two-digit Function display 93, a two and a half digit Value display 94. There is an Inactive light 96 and an annunciator panel 97 containing six fault indicators. In addition, there is an ENTER switch 95 located to the left of the Inactive light 96.
Each motor/system parameter is assigned a number called a Function number as shown below.
Function numbers 0 to 12 are measured motor parameters.
Function numbers 15 to 28 contain field settable data consisting of trip points and motor and system parameters which must be enetered by the user into random access memory ~4.

. .

_ _ _ . . .. _, _ .. . . .

~8~3~C~

--lo~
FUNCTION MEA~URED AND
NUMBER DISPL~YED VALUES UNITS
0 Maximum ~inding RTD Temperature C
1 RTDl Winding Temperature C

2 RTD2 ~inding Tempera-ture C

3 RTD3 Winding Tempera-ture DC

4 RTD4 Windin~ Temperature C
RTD5 Winding Temperature C
6 RTD6 Winding Tempera-ture C
10 7 RTDl Bearing Temperature C
8 RTD2 Bearing Temperature C
9 Line 1 Current % FLC
Line 2 Current % FLC
11 Line 3 Current % FLC
1512 Ground Fault Current -O Trip Current FUNCTION TRIP POINTS MOTOR/ VALUE
NUMB~ER SYSTEM PARAMETERS UNITS _MITS
Total number winding RTD's Qnty. 0-6 16 Total number bearing RTD's Qnty. 0-2 2017 Winding RTD trip temperature C 20-155 18 Bearing RTD trip temperature C 20-155 19 Full Load Current (FLC) % CT
secondary 20-100 Locked Rotor Current X FLC 3-12 2521 Stall Time Sec. 1-99 22 Instantaneous Overcurrent X FLC 3-15 23 Phase Unbalance Trip Current % FLC 5-30 99 disables function 3024 Ground Fault Trip Current % of/Amp 5-100 Ground Fault Trip Time Delay Sec. 0.10-0.99 26 Phase Reversal Number 1 disables function 2 enables function 27 Time Overcurrent Trip (Ultimate Current) % FLC 100-120 28 Reset Mode Number 51 manual reset 85 automatic reset If winding RTD's are used, Function 27 should always be selected to be 120% of full load current. If winding RTD's are not used, function 27 must be used to selec-t an ultimate current of 100-120% of full load current.

The specific data for each of the above functions 15 through 28 must be established and s-tored in -the random access memory 44 before protection can begin. For example, if a motor llad the following data ~or a particular application:
Full ~oad Current (FLC) = 400A
Locked Rotor Current = 6 X FLC or 2400A
Instantaneous Overcurrent = 10 X FLC or 4000A
Stall Time = 5 seconds Phase Unbalance Trip Current = 20% X FLC or 80A
Ground Fault Trip Curren-t = 18A
Line CT Ratio = 500:5 Ground Fault CT Ratio = 100:5 The following information will be obtainea for function No. 15 through 28.
Total number of winding RTD's 3 Total number of bearing RTD's 2 Winding RTD trip temperature 130C
Bearing RTD trip temperature lOO~C
20 Full load current (% of CT rating) 80%
Locked rotor current 6X FLC
Stall Time 5 Sec.
Instantaneous overcurren-t 10 X FLC
Phase unbalance trip current 20% of FLC
25 Ground fault trip current 90O of lA
Ground fault trip time delay 0.50 Sec.
Phase re~-ersal Yes Time overcurrent trip (ultimate current) 120% FLC
Type reset Automatic 85 ~___ __ `

.

?~

Values are calculated as follows:

Eull load current (FLC) = 500A x 100~ = 80% of CT ra-ting ~ocked rotor current (~RC) = 400~ = 6 x FLC

Qhase unbalance = 80 x 100% = 20% of FLC

~round Fault = 18A ~ 1OO x 100% = 90% of lA

Time overcurrent trip - Always set to 120 when winding RTD's are used.
Data could be entered into each function 10 through 28 by using the following data entry procedure:
1. Depress the "ENTER" switch 95.
2. Depress the "FUNCTION" switch 86.
3. Enter the desired function number (15-28~
by serially depressing the appropriate digit entry buttons on keyboard 41. The number entered will appear in the "FUNCTION" display 93.
4. Depress the "VALUE" switch 87.

5. Enter the desired value, most significant digit first, by depressing the digit entry buttons. The value will be displayed in the Value Display 94.

6. To store the displayed data, depress and hold the "ENTER" switch 95, then momentarily depress the "STORE"
switch 91, while holding the "ENTER" switch 95. Now release the "ENTER" switch. A flashing display 94 indicates the value was not stored because it was out of range or because of improper procedure.

7. Repeat steps 2--6 for all functions 15-28.
The identification number (I.Do number) is meant to provide some measure of protection against unauthorized "tampering" with the stored data. The ID number is also used to select manual or automatic reset operation.

)3~

The ID number ls the last function to be en-tered and will not be accepted unless, and un-til, all other functions have been properly stored into storage memory.
Only two numbers will be accepted as valid ID numbers.
These are: 51 for manual reset, 85 for automatic reset.
Knowledge of the ID number is required to return the unit to "ACTIVE" status following the change of any stored setpoint.
Prior to the required motor/system data being entered into the random access memory 44, the motor protection device was in the inactive mode. The inactive mode is indicated by the light 96 labeled "INACTIVE" on the front panel being lit and the alarm relay 53 being in the alarm state. The inactive mode means that motor protection is not being provided. When the device is in the inactive mode, the status of the trip alarm relay 55 remains unchanged, i.e., if -the trip relay is tripped it remains tripped or if it is not tripped it remains untripped. In addition to the device being in the inactive mode when any of the required motor/system data is not stored in memory it also becomes inactive when the supply power is lost, any of the stored motor/system data is altered without proper entry of the identification code, or when an internal malfunction of the device is detected.
The Inactive light will not be lit when the device is in the inactive mode and there is no power being applied to the device.
The internal malfunction of the device as mentioned above can occur in two ways:
1. The motor protection data stored by the user into RAM Memory 44 is des-troyed or lost. The device frequently checks the data which is stored three times in random access memory 44 and if any distortion or loss of data is detected by comparing the data entries, false tripping due to incorrect data will be avoided by assuming the inactive state. The alarm relay will be activated when the device is in the inactive mode. Also, the ke~vboard and the numerical display will not function and no motor protection is provided until the user restores operation by depressing -the special "ENTER" switch 95. This will activate the keyboard and the display functions and also clear the bad data. The user then must check all the data in Functions 15 to 28, enter and store the missing values and finally store the ID number.
2. The processor loses control over its normal operation. The processor normally generates a pulse train to keep the alarm relay in -the non-alarm state.
Should the processor unit malfunction and not produce this pulse train, the result would be that the device would assume the alarm state. The inactive light may or may not be on.
The annunciator panel 97 containing the fault indicators 51 located on the front panel consist of six light-emitting diodes labeled Overload, Phase Reversal, Ground Fault, Instantaneous Overcurrent, Phase Unbalance and searing, indicate an alarm condition when they are flashing and a fault (trip condition) when a steady indicator signal is given. Flashing overload and bearing lights also indicate an open RTD.
The keyboard 41 on the front panel can be utilized in checking motor conditions. Any function can be examined through the following steps: (1) depress FUNCTION switch 86, (2) enter the Function number in the keyboard 41 and the respective value will appear in the ~alue Display 94. The display will remain energized for three minutes after the last command is entered through the keyboard. After three minutes, the display will automatically blank in order to prolong display life.
Depressing the DISPLAY switch 88 will energize the display for an additional three minutes. The RESET switch 92 is used for manually resetting the device after a trip has occurred. The SCAN switch 90 located on the front panel activates the Scan function. This function provides an ?3~

automatic, sequential display of the motor conditions that exist in functions 1-11. Unused unctions wi]l not be scanned. For example, if no bearing RTD's are used, function 7 and 8 will be omitted Erom the Scan.
The Scan function can be hal-ted by depressing the FUNCTION switch.
Any set poin-t stored in -the memory can be changed by following the six s-teps previously discussed for entry of set poin-ts. Should any set poin-t (function 15-28) be changed, the ID number will be reset to zero by internal programming. The inactive light 96 will turn on and the alarm relay will be in the alarm state to signal that motor protection has been interrupted. To regain protection, the identification number (function 28) must be stored again into the RAM memory 44.
A flashing display for function numbers 15 to 27 indicates tha~ a selected value to be entered is outside the limits previously discussed for the various functions. A flashing display for function number 28 indicates one of the following: (1) an invalid ID number is selected; (2) all the set points (function numbers 15 to 27) are not stored into the storage memory. A
flashing display may be stopped by depressing VALUE or FUNCTION buttons.
Phase unbalance (function 23) may be disabled by storing the value 99 for function 23. The device considers an open phase as a special case of phase unbalance; therefore, if the phase unbalance function is disabled, open phase protection is also disabled. Storing function 26 with 1 disables phase reversal protection.
Storing function 26 with 2 enables phase reversal protection.
For application where no winding and/or bearing RTD's are used, functions 15 to 18 and function 27 must be selected as follows: (1) if no bearing RTD's are used, function 16 is set to zero and function 18 is set to 130 (2) if no winding RTD's are used, function 15 is set to q33~

zero, function 17 is set to 130, and function 27 is se-t to the desired value (100 to 120% of full load current) of the ultima-te trip current.
During a power outage to the motor protection device, no protection is provided; however, a recharge-able battery located inside the motor protection device will retain the user set points stored in the random access memory 44. This permits the motor protection device to begin protection as soon as the power i5 re-established without having to store all -the field settable data again. If some of the data is lost from the memory when the motor protection device is not being supplied power, the device will switch to the inactive mode and the user will have to re-enter the set poin-ts 5 before the motor protection can be restored.
The computational portion of the processor 1, the trip relay driver, and the trip relay can be tested by means of a built-in test function. This test artificially simulates a rising winding RTD tempera-ture 0 which will eventually result in actuating the trip relay.
The test will not operate when the device is protecting the motor (the "INACTIVE" indica-tor light-OFF).
The following procedure is used to perform this test:
1. Check the following:
a. All of the user data for function number 15 to 28 must be stored into RAM s-torage memory 44.
b. All the seven indicator ligh-ts (LED's) of panel 97 including the "INACTIVE" light 96 to the front panel must be off.
c. The alarm relay should not be in -the alarm state~
d. The trip relay should not be in the trip state.
2. Put the device into the Inactive mode by storing 0 into function 28.

~q,!3~

The "INACTIVE" light 96 should turn ON~ The alarm relay should go to the alarm state.
3. Not e the winding RTD trip temperature stored lnto ~unction 17.
The winding RTD alarm temperature is 10C
below the trip -temperature.
~. Se]ect function number 0.
If the winding RTD's are used (Function 15 is not zero), the value display 94 will show the prevailing maximum winding RTD temperature.
If the winding RTD's are not used (Function 15 is zero), the Value display 94 will show zero.
5. Depress and hold the VALUE swi-tch 87.
The Value display 94 will increase slowly as long as the VALUE switch is held depressed If no winding RTD's are used, the Value display 94 will first jump to 40C or higher temperature and then increase slowly.
6. When the winding RTD alarm temperature is reached, the OVERLOAD indicator will begin to flash indicating alarm condition.
The alarm relay is already in the alarm state since step 2 above.
7. When the winding RTD trip temperature is reached, the OVERLOAD indicator will come on steady and the trip relay will go to the trip state.

8. If the ~ALUE switch 87 is held depressed even after the trip temperature is reached, the displayed temperature will eventually reach 199C and then roll over to 0C.

9. Release the VALUE switch 87 and then depress and hold the RESET switch 92.
If winding RTD's are not used, the Value display 94 will jump to zero and the OVERLOAD indicator light will be turned off immediately.

~8'~'3~

If winding RTD's are used, the maximum winding RTD temperature displayed by the Value display 9~ will revert to the prevailing value a-t a controlled rate. The OVER-LOAD indicator will turn off when the temperature of 15C below the trip temperature is reached.
Even though no trip condition remains and the ~ESET switch 92 is depressed, the state of the trip relay (tripped) does not change. This is because the protection device is in the Inactive mode (no ID number in function 28 to define reset).

10. Select the desired type of resct by storing appropriate ID number into function 28.
The trip relay should then assume the non-trip state since there is no trip condition present.
The "INACTI~E" indicator light 96 should turn off and the device again assumes protective functions.
All temperature readings are displayed directly in degrees Celsius. The maximum temperatures can be displayed as 199C. Line currents are displayed as percentage of full load current. The maximum current which can be displayed is l99~ of full load current.
Ground fault current is displayed as a percentage of the set point. The maximum ground fault current which can be displayed is a 199% of a set point.
A summary of the operation of the various front panel elements of FIG. 3 is provided below:
Panel Name Type Element Function FUNCTION Switch 86 o Prepare for a new "function" entry.
o Steps a flashing display Function display 93 2-digit read- Displays the last-out entered function number which represents a unique motor parameter.

Panel Name Type Element Function ~ALUE Switch 87 o Prepares for a new da-ta en-try for the function whose number (15 to 28) is displayed in the function aisplay.
Value display 94 2-1/2-digit Displays the value of read-out the -Eunction whose number is displayed in the function display.
DISPLAY Switch 88 o Energize the Function and Value displays for a 3-minute interval.
o Displays the pre-viously entered function and value.
SCAN Switch 90 o Examines and displays the motor conditions of function numbers 1-11 sequentially.
STORE Switch 91 o Stores a new acceptable value in the memory.
o Initiates a flashing display if the value is out of range.
ENTER switch Switch 95 Used in conjunction with STORE to enter date into storage memory 44.
O through 9 Numberals 85 Digit en-try for the function and Value display.
RESET Switch 92 Allows the trip relay to be manually reset after a trip has occurred.
INACTIVE Light 96 Indicates device is (LED) protecting the motor.

3`13;~

Panel Name Type Element Function OVE~LOAD Panel 97 Fault indication - A
PHASE REVERSAL (IED's) flashing light signals GROUND FAULT an alarm condition.
5 INST. OVERCURRENT A steady indicator PHASE UNBALANCE light signals a BEARING fault (txip) condition.
Flashing Overload and BEARING lights also indicate an open RTD.

Referring to FIG. 1, the processor 1 which may be an Intel 8035, 8-bit chip microcomputer or other suitable processor has a crystal oscillator 45 having a frequency of 5.898 MHz which is used to provide internal timin~ references for an instruction cycle time of 2.54 microseconds and an internal 8-bit timer interval of 81.38 microseconds per count.
An 8-bit timer of processor 1 is used to generate timer interrupts every .016 seconds. To provide various timing references such as .25 seconds, .5 seconds~ 1 second, etc., an 8-bit memory locations of processor 1 is used as a system clock. This clock is incremented one count every .016 seconds. This cloc~ over~lows from 255 counts to zero.
A bus is used to interface the battery powered random access memory 44, latch decoder 40, address latch 39 and latches 46, 47 and 48 with the processor 1.
These devices are assigned specific addresses called external addresses which are listed below.

~(33~

Hexadecimal Address F3 Latch Column #3 4 Bits ( 4 keys) F2 Decoder Column ~2 4 Bits (4 keys) KEYBOARD
Fl 40 Column #1 4 Bits (4 keys) F0 Column ~0 4 Bits (4 keys) Not Used E8 Address Latch 39 (4-bit): Input Selector Not Used E4 Latch 48 (8-bit): Annunciator, Alarm, and Trip Signals Not Used E2 Latch 50 (8-bit): Value Bu-ffer El Latch 49 (8-bit): Function Buffer Not Used A program storage read-only memory 43 is also connected to the bus and consis-ts of the total of 2,048 8-bit bytes. There are two locations in the program memory 43 of special importance: (1) the first instruction to be executed after initialization (Reset/Power-up) is stored in location zero (RESET); and (2) the first instruction to be executed after timer interrupt is s-tored in location 7 (TMRIMT).

3~

The processor 1 contains a total of 12~ 8-bit bytes O F resident data memory. Bank 0 working registers are used during the system reset and normal operation~
Bank 1 working registers are used during the timer interrupt program. Bank 0 and bank l each contain the eight 8-bit bytes. The resident da~a memory also has an eight level stack.
The keyboard 41 is read by pulses being provided at one of the 4 outputs of latch decoder 40 which are connected to keyboard 41. The 4 output lines of keyboard 41 indicate which 4 of 16 keys are closed for each l of 4 input lines to the keyboard addressed. The keyboard can be checked for key closures and releases, 4 keys at a time.
The tri-state driver 42 is used to connect and disconnect the output of the keyboard to the bus, based on the enable signal received from decoder latch 40. The function numbers entered are stored in resident data memory of processor 1 and displayed. When field settable data is entered, depressing the ENTER switch 95 enables the write function of the read~write enable line of RA~ 44 storing the data in RAM 44.
Program memory 44 is a battery powered 192x4 random access memory. This memory is divded into three sections. Each set point or calculated constant is stored three times in the memory, one copy in each section. This makes the contents of section l, section 2 and section 3 identical. This arrangement provides a check against any undesirable changes in the contents of this memory. As was explained earlier, a special procedure is necessary to enter set points into this memory.
Address latch 39 is a four-bit latch which is used to select one of 16 analog input signals from multi-plexers 3, 4, 5 and 33. Based on the contents of ~atch 39, one of the 16 inputs shown below is selected.

INPUT CONTENT OF SELECTED

3 2 ~EX WINDING RTD 3 7 6 HEX BEARING RTD ].
8 7 HEX ~EARING RTD 2 9 8 HEX P~ASE 1 CURR~NT (HIGH GAIN) 9 HEX PHASE 2 CURRENT (HIGH GAIN)

11 A HEX PHASE 3 CURRENT (HIGH GAIN)

12 B HEX GROUND FAULT CURRENT
15 13 C HEX PHASE 1 CURRENT (L~W GAIN) 14 D HEX PHASE 2 CURRENT (LOW GAIN) E HEX PHASE 3 CURRENT (LOW GAIN) .
Up to eight RTD's can be monitored with up to six RTD's (one or two per phase normally) in the windings of the motor. If less than six winding RTD's are used, then the lower number addresses of the multiplexars 3, 4 and 5 are filled first. Up to two bearing RTD's can be connected to the seventh and eighth input of multiplexers 3, 4 and 5.
Multiplexing the RTD's allows a single RTD input bridge circuit 6 to be used. The purpose of -the RTD input bridge circuit 6 is to compensate for the RTD lead resis-tance which can introduce significant errors in the RTD
resistance readings.
A link 18 is placed in series with the proper value of feedback resistor 17 of the operational amplifier 9 in RTD input bridge circuit 6. The feedback resistor 17 is chosen to correspond with the type of RTD that is going to be used. For example, 10 ohm copper, 100 ohm platinum, 100 ohm nickel, 120 ohm nickel. Only one type of RTD
should be used.
When the address latch 3g is connec-ted to the bus by an enable signal received from the decoder latch 40, one of the addresses previously lis-ted is loaded into the latch. If address is below 8 HEX than multiplexers 3, 4 and 5 are enabled while multiplexer 33 is disabled because of the inverter 42 in the most significan-t bit line feeding the multiplexer 33~ Likewise, if the address in latch 39 is above 7 HEX, then multiplexers 3, 4 and 5 will be disabled while multiplexer 33 will be enabled.
The second most significant bit line from address latch 39 serves to select the high gain or low gain range of selectable gain amplifier 35a. A logic zero input causes the high gain setting of the amplifier 35a to be chosen while a logical one signal causes the low gain amplification setting of amplifier 35a to be selected.
The voltage output of operational amplifier 9 is proportional to the selected RTD resistance in -terms of volts per degree C, independent of the lead resistance.
The currents of lines Ll, L2 and L3 are monitored by current transformers 20, 21 and 22, respectively. Each current transformer is connected to a corresponding isolation transformer 23, 24, 25 which has an output of one volt per amp. The isolation transformers 23, 24 and 25 are connected to corresponding rectlfier/
filters 26, 27 and 28. The output of rectifier/filter 26 which is associated with line 1 is connected to the first and fifth input of multiplexer 33. The output of rectifier/filter 27 which is associated with line 2 is connected to the second and sixth input of multiplexer 33, and the output of rectifier/filter 28 which is associated with line 3 is connected to -the third and seventh input of multiplexer 33.

The analog outputs of the selectable gain amplifier 35a and the analog output of the operational amplifier 9 are coupled to one input of a comparator 37a.
The signal available at the comparator 37a is determined by the contents of latch 39. A logic level 1 output from comparator 37a signifies that the analog outpu-t o-f the digital-to-analog converter 38 is more -than the measured value input to the compara-tor 37a. A logic level zero output from the comparator signifies that the analog output from the digital-to-analog converter 38 is less than the measured signal. The output of the comparator 37a is connected to a successive approxima-tion program located in processor 1 which adjusts the digital output fed to the digital-to-analog converter 38 so that the analog value output from the digital-to-analog converter which is fed to one input of the comparator, approaches the measured analog values input to -the other input of the comparator 37a. A selectable gain amplifier 35a provides two sets of phase currents, one being the high gain phase currents having a range of zero to 195% full load curren-t which are used for phase unbalance and overload calculations, and the other being the low gain phase current which provide full range measurements of zero to 152 times full load current and are used for instantaneous overcurrent and overload protection. The 16 inputs are stored in the resident data memory of processor 1.
Phase reversal information is provided by the output of NAND gate 64 to processor 1. Two flip-flops 58 and 59 are used. Flip-flop 58 is clocked by line 2 and steered by the other two lines (Ll and L3) while flip-flop 59 is clocked by line 3 and steered by the other two lines (Ll and L2~. Should the phase sequence reverse the flip-flops 58 and 59 will no longer stay in the proper state (which results in a logic level "0" output from NAND
gate 64), rather the output of NAND gate 64 will go high indicating a phase reversal.

The software used b~ -the motor protection device will now be discussed. The software for the motor protection device is organized in -the Eollowing -functional categories:
1. Current and Temperature Inputs~
2. Keyboard and Display ~'unctions.
3. Protection E'unc-tions.
4. Annunciator, Alarm and Trip Outputs.
5. General Purpose Subroutines.
Each of these functional categories is broken down into several program subroutines. The current and temperature input related functions are organized in the following categories:
1. Analog to digital (A/D) conversions (ANALOGI).
2. The maximum value of phase currents (MAXLOG).
3. The maximum values of RTD temperatures (MAXRTD).
The A/D conversion of the sixteen current and temperature inputs is performed by the successive approximation method. Each analog input signal is selected for the A/D conversion as previously described. After performing the A/D conversion, the digital data is stored into the assigned data memory (RAM) location in the resident data memory of processor 1. After each RTD input signal is selected, time delay of 0.016 second is required before the A/D conversion can be performed.
To satisfy this requirement, the program is executed once every 0.016 second time interval.
The following scale is used in storing the 16 temperature and current inputs.
.

3~

INPUT
NUMBER INPUT DATA SCALE
1 WINDING RTDl TEMP. 1.28 COUNT=1C
2 WINDING RTD2 TEMP. ].28 COUNT=1C
3 WINDING RTD3 TEM~. 1.28 COUNT=1C
4 WINDING RTD4 TEMP. 1.28 COUNT=1C
WINDING RTD5 TEMP. 1~28 COUNT=l~C
6 WINDING RTD6 TEMP. 1.28 COUNT=1C
7 BEARING RTDl TEMP. 1.28 COUNT=1C
8 ~EARING RTD2 TEMP. 1.28 COUNT=1C
9 PHASEl CURRENT (~IGH GAIN) 128 COUNTS=5A
PHASE2 CURRENT (HIGH GAIN) 128 COUNTS=5A
11 PHASE3 CURRENT (HIGH GAIN) 128 COUNTS=5A
12 GROUND FAULT CURRENT 255 COUNTS=lA

13 PHASEl CURRENT (LOW GAIN) 16.7 COUNTS=5A

14 PHASE2 CURRENT (LOW GAIN) 16.7 COUNTS=5A
PHASE3 CURRENT ~LOW GAIN) 16.7 COUNTS=5A
16 INSTANTANEOUS OVERCURRENT 16.7 COUNTS=5A

A digital filter is provided by the software for the maximum value of the low gain phase current inputs.
The new maximum value can only be a few counts higher or lower than the old maximum value, i.e., OLD MAX. - 4 COUNTS < NEW MAX. < OLD MAX. + 4 COUNTS
To provide this digital filter, the maximum value of phaseCurrents ~MAXLOG) program is executed once every 0.016 second interval.
The device uses the maximum values of the winding and the bearing RTD temperature inputs to provide protection functions, rather than individual RTD temperatures.
Digital filters are provided for the maximum values of the winding and the bearing RTD temperatures so that:
OLD MAX - 1 COUNT < NEW MAX ~ OLD MAX -~ 1 CO~NT
To provide these digital filters, the maximum values of RTD temperatures (MAXRTD) program is executed once ~3'{3~

every 0.25~ second interval. The user stores the number of winding RTD's used in Eunction No. 15 and the number of bearing RTD's used in Function No. 16. This information is used to determine the maximum RTD -temperatures. The l~AXRTD program also determines the following modes of operation.
1. No windlng RTD's (NOWRTD).
2. Open winding RTD (OPNWND).
3. Open bearing RTD (OPNBRD).
Each mode is represented by one bit message called a flag.
A logic level 0 means that the message is false. A logic level 1 means that the message is true. Referring now to FIG. 4 the MAXRTD flow chart is shown. The first block of the MAXRTD flow chart is decision block 100. Here a check is made to see if the trip test flag (TRIPT) is equal to 1. If the flag is equal to 1, it would mean that a trip test is being conducted. If the flag is equal to zero, then a trip test is not being conducted. If a trip test is going on, then block 100 is connected to node 120 at the end of the MAXRTD program since no bearing or winding input calculation is to be made. Winding temperatures will be simulated elsewhere for test purposes. If a trip test is not going on, then block 100 is connected to subroutine block 101 which sets the no winding RTD flag (NOWRTD), open bearing flag (OPNBRG) and the open winding flag (OPNWD) ali equal to 0. Block ]01 is connected to the maximum bearing temperature (MAXBRG) subroutine block 10~. Block 102 is shown in greater detail starting with subroutine block 103. In block 103 N is set equal to the number of RTD's selected by the user, A is set equal to the maximum of winding RTD's and the decrement flag (DECPMT) is set equal to 1. Block 103 is connected to decision block 104 where if no winding RTD's were selected to be monitored by the user, block 104 is connected to sub-routine 105. Block 105 sets the no winding RTD flag 3~ 33~3~

equal to 1 or, in o-ther words, -true. Block 105 is connected to node 117. If the number of winding RTD's selected by the user is not equal to 0, then block 104 is connected to subroutine block 106 which sets M equal to 1 and count equal to 1. Block 106 is connected -to loop connec-ting node 107. Node 107 is connected to condition block 108.
The Mth RTD is checked to see if it is equal to 199.22 C
(2S5 Counts) and if the Mth RTD is e~ual to 199.22C, i-t is open. If the RTD is open, block 108 is connected to subroutine block 109 which sets the open winding RTD flag (OPNWND) equal to 1. Block 109 is connected to decision connecting node 110. If the Mth RTD is not open, then block 108 is connected to condition block 111. Condition block checks if the Mth RTD is greater than the previous maximum RTD. If this condltion is true, then block 111 is connected to subroutine block 112. Block 112 sets the decrement flag (DECP~T) equal to 0 and if COUNT is not equal to 0 then the previous maximum RTD is increased by one count and count is set equal to 0 to allow only one decrement for each check of N winding RTD's. Block 112 is connected to node 110. If the condition of block 111 is false, then block 111 is connected to subroutine block 1130 In bloc]c 113 if the Mth RTD is equal to the previous maximum, then the decrement flag is set equal to 0. Block 25 113 is connected to node 110. Node 110 is connected to subroutine block 114. The RTD to be checked is increased by one and the number of RTD's to be checked is decreased by one. Block 114 is connected to condition block 115.
Block 115 checks if all the RTD's have been checked.
If there are RTD's left to be checked block 115 is connected to node 107. If all the RTD's have been checked block 115 is connected to condition block 116. In block 116 the decrement flag is checked to see whether or not it has been set to 0. If the flag has been set to 0, then the previous maximum winding RTD has been incremented and block 3q~3~

116 is connected to node 117 at the end of the MAXWND
subroutine. If the decrement flag is equal to 1, then the N RTD's checked were not equal to or greater than the previous maximum. Block 116 is connected to subrou-tine block 118. In block 118 if the previous maximum winding RTD is not 0, lt is reduced by one count. Block 118 is connected to node 117.
Node 117 is connected to subroutine block 119.
Block 119 contains the maximum bearing RTD (MAXBRG) subroutine which is similar to the maximum winding subroutine just described. Block 119 is connected to node 120 at the end of the maximum RTD program.
The Keyboard and Display Functions will now be discussed. The primary function of the 16-key keyboard and the numerical displays (Function and Value) is to enable the user to store the motor and system data (setpoints) into RAM 44. However, using the keyboard, various measure currents and temperatures are also displayed. The functions performed are organized in the following three programs:
1. Receive commands and data from the keyboard (KYBORD).
2. Prepare Function and Value data for display (UPDART~.
3. Display the Function and Value data (DISPLY).
The user gives a command using one of the six command keys (Function, Value, Store, Scan, Reset, and Display). The user enters a digit using one of the ten number keys (0 to 9). The KYBORD program is executed once every 0.016 second interval to help provide adequate time delay for a bouncing key. Two successive key closure detections, 0.016 second apart, are required before a key entry is accepted. Also two successive key release detections, 0.016 second apart, are required before a new key entry is accepted.
The Prepare Function and Value Data for Display tUPDATE) program will now be discussed. Depending on which mode of operation is selected by the user using the keyboard, this program determines what data should be displayed. When the scan mode is selected (SCAN = 1), function numbers 1 to 11 are selected for display at regular interval of 4 seconds. Of course, the unused ~TD temperatures are skipped in this process. Depending on which function number is selected, corresponding value data is prepared for the display. When any one of the RTD temperature function numbers (0 to 8) is selected, the corresponding value da-ta prepared for display is in C. When any one of the three phase current function numbers (9, 10, 11) is selected, the corresponding value data prepared for display is in percent of full-load current (FLC). When the ground fault current function number 12 is selected, the value data prepared for display is in the percent of the ground fault trip current.
The values stored in the user data memory RAM
44 for the setpoints (Functions 15 -to 29) are in binary.
When one of the setpoin-t function numbers is selected, the corresponding value data prepared for display is in the three digit BCD value. The function number data prepared is two BCD digits and the value data prepared is three BCD digits. The UPDATE program is executed once every 0.256 second. The three minutes display timer is decremented once every second.
The Display the Function and Value Data (DISPLY) program sends the function and the value data for display.
This program is executed once every 0.016 second interval.
If the flashing display mode is active (FLASH = 1), the flashing display is generated. If the three minutes display timer has timed-out, the display is blanked.
The leading zeros in the function number and -the value data are blanked, i.e., instead of 01, 1 is displayed.
Whenever the user stores a setpoint (Functions 15 to 29), a set of constants is calculated based on the data stored by the user. These calculated constants are then ~.8~331~

stored in the RAM 4~.
The constants must be stored at this -time since the write enable line of RAM 44 can only be enabled externally by the enter switch.
The following pro-tection functions are per~ormed:
1. Overload/Overtemperature protection (CUTEMP, OVRLOD).
2. Bearing Overtemperature protection (OVTBRG).
3. Ground fault protection (GFAULT).
4. Phase Unbalance protection (P~IAUNB, PUNBAL).
5. Phase Reversal protection (PHAREV).
6. Instantaneous Overcurrent protection (INSTOC).
The necessary ~unctions to provide the overload/
overtemperature are organized in the following two programs.
1. Calculations for the copper winding temperature (CUTEMP).
2. Determination of Overload Alarm and Trip conditions (OVRLOD).
Two separate calculations for the copper-winding temperature (CUTEMP) are made depending on whether winding RTD's are used or not.
When the winding RTD's are used, -the calculations are based on the RTD model and when the winding RTD's are not used, the calculations are based on the non-RTD model. When the winding RTD's are not used, the user stores 0 into Function 15 which in turn selects the non-RTD model. Other-wise, the RTD model is selected. The copper temperature calculations assume Class B insulation, i.e., the trip temperature of 180C. The CUTEMP program is executed once every 0.256 second. One data memory (RAM) location in the processor contains the RTD temperature (1.28 count = 1C) which is used to calculate the copper temperature. When the winding RTU's are used~ this temperature is the maximum of the winding RTD temperatures, otherwise, it is the calculated RTD temperature.
When the winding RTD's are used, the model shown in FIG. 5 is used. FIGURE 5 has a current source 125 equal to KIm connected to a node 126. A maximum RTD

'3~

value 127 is connected to an input. of amplifier 128.
A resistor 129 is connected between node 126 and the output o~ ampll~ler 128. A capacltor 131 ls connected between node 126 and ground. Amplifier 12~ provides a current path to ground. ~ constant current source 132 is connected to a node 133.
The output of amplifier 128 is proportional to the maximum RTD winding temperature. The voltage at node 126 is proportional to the motor copper winding temperature.

dCU CT - RTD
10 I = KIM = C +
dt R

Where K is a constant IM is the maximum value of the phase currents, ln per unit (R~U.)~
CU is the copper winding temperature RTD is the maximum winding RTD temperature.
RC is the thermal time constant between the copper windings and the RTD.
If we apply the boundary condition:
CU = CUO when t = O
where CUO is the previous value of copper winding temperature when we have the solution.
-t/RC -t/RC
CU = (KRIM + RTD) (l-e) + e CUO

To simplify this formula, the following two assumptions are made:
1. t = 0.256 second This implies that the copper temperature calculations are made every 0.256 second.
2. RC = 65.4 seconds The RC time-constants for many motors is between 30 to s~

60 seconds. So -this time constant will slightly over protect the motor.
Substituting these values into the copper tempera-ture ~ormula/ we get: 2 CU = KRIM + RTD) /256 -~ 0.996 C~O
This formula i~ used -to calcula-te the present value of the copper -temperature (CU) from the previous value of the copper temperature (CUO) during an overload condition.
When the maximum value of the phase currents (IM) is equal to or greater than 1.95 P.U. ~195~ FLC) it is defined as the overload condition. For the lesser value of -the current, it is defined as the steady-state condition. The constant KR in the copper temperature formula is called the RTD model calibration constant.
The user stores, in Functions 20 and 21, the value of loc~ed-rotor current and the stall time respectively. The value of the calibration constan-t is determined ~ased on this information using the following formula for Class B insulation:
KR = (80 + 8871/STALL TIME) / (LOCKED ROTOR CURRENT) When the maximum phase current (IM) is less -than 1.95 P.U. (195% FLC), it is defined as the steady-state condition for the RTD model. For this condition, the copper temperature formula is modified as follows:
CU = (10 IM + RTD) /256 + 0.996 CUO
This gives the copper temperature a few degrees (0 to 19C) higher than the RTD temperature during the steady-state conditionO This modification of the original copper temperature formula prevents nuisance tripping during moderate overloads and cyclic loading. In this region of operation, the device relies on the RTD
temperature to cause a trip.

';3~
~35-When the winding RTD's are not used, the Eollowing three calculations are involved wi-th the non-RTD case:
1. The RTD temperature.
2. The copper temperature Eor overload.
3. The copper temperature for steady-state.
Since the winding RTD's are not used, the RTD
temperature is calculated based on the motor current (IM).
Every 65 seconds, a new value of the RTD temperature is determined which is either 2.34C higher or 0.78C lower than the previous value of the RTD temperature.
One data memory (RAM) location in the processor is used to determine 65 seconds time interval (0.256 second per count). One reference temperature (REFTMP) is calculated using the following formula:
REFTMP = 40C + (IM %FLC)/2.56C
or REFTMP ~ (105 -~ IM ~FLC)/2 counts Where 0 ~ IM %FLC ~ 100 If the reference temperature is lower than the previous RTD
temperature then the new RTD temperature is 0.78% lower, otherwise, it is 2.34C higher than the previous RTD
temperature.
The user stores in Function 27 a value (100 to 120%) of ultimate trip current. This value of the ultimate trip current is used to define an overload or a steady-state condition.
When the maximum value of the motor current (IM) is equal to or greater than the ultimate trip current, it is considered as an overload condition.
Otherwise, it is considered a steady-state condition.
For this non-RTD model overload condition, simple I t curves are yenerated using the following formula:

`;3~

-36~
CU = (KR IM + RTD)/256 ~ CUO
Where CU is the calculated copper temperature.
CUO is the copper temperature at t=o, IM is the maximum value of the motor current, RTD is the calculated RTD temperature.
KR is the calibration constant, given by -the formula, KR = (8960/STALL TIME)/(LOCKED ROTOR CURRENT) The user selects the value (1.0 to 1.2 P.U~) of ultimate trip current using Function No. 27. When the maximum phase current (IM) is less than the selected ultimate trip current, it is defined as the steady-state condition. For this condition, the copper temperature formula is modified a~ follows:
CU = (10 x IM + RTD)/256 -~ 0.996 CUO
Note that this is -the same formula as the RTD model for the steady-state condition.
The following two programs determine the over-load alarm and trip conditions (OVRLOD):
1. The Copper Temperature Alarm and Trip (OVLCUT).
2. The Winding RTD Temperature Alarm and Trip (OVLWND).
The Copper Temperature Alarm and Trip (OVLCUT) program will now be discussed. A data memory (RAM) location of processor contains the value of calcula-ted copper temperature in C.
The trip temperature is 180C.
The alarm temperature is 170C.
The reset temperature is 165C.
Two one bit flags, CUTALM and CUTTRIP, are used to reflect the determined alarm and trip conditions respectively. If there is the alarm condition, the CUTALM bit is set (CUTALM = l); otherwise it is cleared (CUTALM = 0). If there is the trip condition, the CUTRIP bit is se-t (CUTRIP = 1); otherwise it is cleared (CUTRIP = 0).

3~3~

The Winding RTD Temperature Alarm and Trip (OVLWND) program has a data memory (RAM) location in processor that contaiIIs -the RTD temperature (1.28 coun-t =
1C). Depending upon whether the winding RTD's are used or not, this temperature is either the ma~imum -temperature of the winding RTD's or the calculated RTD temperature respectively.
Function 17 contains the value of winding RTD
trip temperature (20 to 155C). Based on this information, the winding RTD trip temperature counts (WRTDT):
WRTDT = 1.28 x (Function 17) are calculated and stored into RAM ~ by the KYBORD program at the time when the user stored the setpoints. The trip temperature constan-t WRTDT and the RTD tempera-ture in the RA~ in this processor have the same scale (1.28 count = 1C).
This makes it easier to compare them and determine the alarm and trip conditions.
The trip temperature = WRTDT counts The alarm temperature = (Trip Temperature - 10C).
= (WRTDT - 13) counts The reset temperature = (Trip Temperature - 15C).
= (WRTDT - 19) counts There are two one bit flags, WNDALM and WNDTRP, which are used to reflect the determined alarm and trip conditions respectively. If there is the alarm condition, the WNDALM bit is set (WNDALM = 1): otherwise it is cleared (WNDALM = O).
The Beariny Overtemperature Protection (OVTBRG) uses a data memory (RAM) location in processor that contains the maximum temperature of the bearing RTD's (1.28 count =
1C). For the comparison purpose this temperature is converted in to C by dividing the counts by 1.28. Function 18 contains the bearin~ RTD trip temperature (20 to 155C).
The alarm temperature = The trip temperature - 10C.
The reset temperature = The trip temperature - 15C.

There are two one bit flags, BRGALM and BRGTRP, which are used to reflect the determined alarm and trip co.nditions respectively. If there is the alarm condition, the BRGALM bit is set (BRGAL.M = l); otherwise, it is cleared (BRGALM = 0). The beariny alarm condition caused by an open bearing RTD (OPNBRG) is determined by the OUTPUT program.
The Ground Fault Protection (GFAULT) program uses a data memory (RAM) location in the processor which contains the measured value of the ground fault current (255 counts = 1 Amp). The Function 24 contains the ground fault trip current in % of 1 Amp. Multiplying this value b~
2.55, we get the ground fault trip current in counts. 76%
of the trip value is the ground fault alarm current. The ground fault reset current is 4 counts less than the ground fault alarm current.
There are two one-bit flags, GFALRM and GFTRIP, which are used to reflect the determined alarm and trip conditions respectively. If there is the alarm condition, the GFALARM bit is set (GFALRM = l); otherwise, it is cleared (GFALRM = 0).
If there is the trip condition, it is necessary that the trip condition must exist at least for the time interval (time delay) selected by the user in Function 25.
To determine this time delay one data memory (RAM) location in the processor is used as the ground fault timer (GFTIME: 1, count = 0.016 second). After the ground fault time delay, if there is the trip condition, the GFTRIP bit is set (GFTRIP = l); otherwise, it is cleared (GFTRIP = 0).
The phase unbalance protection functions are organized in the following two programs:
1. Phase unbalance calculations (PHAUNB).
2. Phase unbalance alarm and trip (PUNBAL).
Calculations are made to calculate the phase unbalance level (V):
O _ V _ 255 3~

which will be used to determine the phase unbalance alarm and trip conditions. One data memory (RAM) location in -the processor is assigned to store the calculated phase unbalance level (V). At the time of the system initialization (power-up/reset), V = 0 is stored.
The PH~UNB program is executed once every 0.256 second.
1. The minimum value of -the three phase currents (IMIN) is found.
If the IMIN ~ 150% FLC then there is no phase unbalance.
Then the value of V is decremented by 1 count and no further calculations are made for V.
If this situation prevails for some time, V will decrease from 255 to 0 in about 255 x 0.256 = 65 seconds.
2- If the IMIN ~ 150% FLC, then the maximum value of the three phase currents ~IMAX) is found.
If IM~X > 150Po FLC then IMAX = 150% FLC is assumed.
3 The phase difference ( IMAX MIN
4. The normalized phase unbalance (U) is determined by the following formula.
U = (IMA~ - IMIN) x K
Where K = (100/FLC)/(PHASE UNBALANCE SETPOINT) 5. If U > 1 then it is defined as the phase unbalance condition, otherwise it is not unbalance condition.
6. If there is the phase unbalance condi-tion then V = V ~ u2 (1 ~ u2 255) 7. If there is no phase unbalance condition (U ~1), then the phase unbalance reference level (REFLEV) is calculated by the following formula:
REFLEV = 256 x U
8. If REFLEV ~V then the value of V is decremented by 1 count.
9. If REFLEV ~V then the value of V is incremented by 1 count.

3~

V = 255 counts is defined as the phase unbalance trip level.
When there is a phase unbalance condition (1 ~ U ~ 15), i.e., the measured phase unbalance is equal to or greater then the selected phase unbalance trip current in Function 23; then the phase unbalance trip characteristic can approximately be represented by the following formula:
Trip-Time t = 64/V where 1 ~ U ~ 15.
When there is no phase unbalance (0 ~ U ~ the phase unbalance level V will reach the level given by the following formula:
V = 256 x W where 0 ~ U C
To reach this level, the value of V is incremented or decremented only by 1 count per 0.256 second.
Referring to FIGURE 6 the flow chart for the Phase Unbalance Calculations (PHAUNB) is shown.
The first subroutine block 135 of the flow chart is to set A equal to 150~ of fuil load current. Block 135 is connected to subroutine block 136. Block 136 sets B equal to the minimum value of the three high gain phase currents. Block 13~ is connected to condition block 137. Block 137 checks if B is greater than or equal to A.
If B is greater than or equal to A then there is no phase unbalance and block 137 is connected to subroutine block 138 which reduces V by one count if it is not already 0.
Block 138 is connected to node 139 at the end of the phase unbalance calculation program. If B is not greater than or equal to A, then block 137 is connected to sub-routine block 140. Blcok 140 contains the phase maximum subroutine which is shown in greater detail in the Figure starting with subroutine block 141.
In block 141, C is set equal to the maximum value of the three high-gain phase currents. Block 141 is connected to block 142. If the maximum value of the three high gain phase currents is greater than 159~ of full load current (FLC) then the maximum value of the three high g~3~¢~

gain currents is se-t equal -to 150~ FLC. Block 142 is calculated as equal to the maximum of the three high gain currents minus -the minimum of the phase currents divided by the unbalance current trip point. The unbalance current trip point is determined by multiplying the percent phase unbalance stored hy the user by the full load current.
Block 143 is connected to condition block 144.
If U is greater than or equal to 1, then block 144 is connected to subprogram block 145. In block 145, V is set equal to U . Block 145 is connected to subroutine blocks 146 and 147 where, if V is greater than 255, V
is set equal to 255. Block 147 is connected to node 139.
If U is not greater than or equal to 1, then block 144 is connected to subroutine block 149. Block 149 sets ~ = 1 and R equal to 256 multiplied by U. Block 149 is connected to condition block 150. ~lock 150 checks if R is equal to 0. If R is equal to zero, then block 150 is connected to block 151. Block 151 lowers V by one count if it is not already 0 since there isn't any phase unbalance. Block 151 is connected to node 148. If R
if not equal to 0, then block 150 is connected to condition block 152. Block 152 checks if R is less than V; then block 152 is connected to subroutine block 153. Block 153 reduces V by one count if it is not already 0. If R is not less than V, then block 152 is connected -to block 154. Block 154 increments V by one cou~nt. If V is greater than 255, then V is set equal to 255. Block 154 is connected to node 139.
The phase unbalance alarm and trip conditions (P~NBAL) are determined by checking the phase unbalance level (V) stored in the data memory ~RAM) and the process unbalance level (V) stored in the data memory (RAM) in the processor.

~B~

-~2-V = 255 coun-ts is the trip level.
V = 171 counts is the alarm level.
(The alarm level is 67% of the trip level~
V = 12~ is the reset level.
The phase unbalance trip characteristics is shown in FIGURE
7. Two one bit flags, PUALRM and PUTRIP, are used to reflect the determined alarm and -trip conditions respectively.
If Function 23 contains the disable phase unbalance protection code (99), then both the flags are cleared (PUALRM = O and PUTRIP = O). Then neither the alarm nor the trip will take place. If the disable code (99) is not stored, then the alarm and trip conditions are determined.
If there is the alarm condition, the PUALRM bit is set (PUALRM = l); otherwise it is cleared (PUALRM = 0). If there is the trip condition, the PUTRIP bit is set (PUTRIP - l); otherwise it is cleared (PUTRIP = 0).
The user has an option to select the phase reversal protection. When the phase reversal protection is selected, the enable code (23 is stored in Function 26.
To disable this protection, the disable code (1) is stored.
As described previously, the phase reversal condition is detected (Tl = 1) or not (Tl = 0) by the hardware. The phase reversal trip condition is said to exis-t when "the phase reversal Input Tl = 1 and -the maximum value the the three phase currents is at least 50% FLC" for 0.24 second. To determine this 0.24 second time delay, one data memory (RAM) location in the processor is used as the phase reversal timer. Presetting this timer to

15 counts (0.016 seconds per count) generates the required time delay of 0.24 second. One one-bit flag:
PRTRIP is used to reflect the determined trip condition.
There is no alarm condition for this protection function.
PRTRIP = 1 for the trip condition.
Instantaneous Overcurrent Protection (INSTOC) uses one data memory location in the processor 3~J~

-~3-which contains the measured value of the instantaneous overcurrent in counts (16.7 counts = 5 Amp).
Function 22 contains the trip value (3 x 20 x FLC) selected by the user. ~ased on this data, the instantaneous overcurrent trip value in counts (IOCTRP):
IOCTRP = (Function 22) x (Function 19)/6 is calculated and stored by the K~ORD program at the -time when the user stored the setpoints. Thus, the IOCTRP and the measured value both have the same uni-ts (16.7 coun-ts =
5 Amp). So it is easier to compare them to determine the trip condition. One one~bit flag IOCTRP is used to reflect the determined trip condition. IOCTRP = 1 for the trip condition. There is no alarm condition for this protection function.
The Annunciator, Alarm and Trip Outputs (OUTPUT) program prepares and sends to Latch ~8, the fault indicator signals, the annunciator alarm and the alarm and trip rela~ signals.
There are six fault indicator signals:
1. OVERLOAD
2. BEARING
3. GROUND FAULT
4. PHASE UNBALANCE
5. PHASE REVERSAL
6. OVERCURRENT
When any of these signals is logic 1, the corresponding light turns on; otherwise, it turns off. To provide the flashing indicator ~or an alarm condition, logic 1 is maintained for half-second and logic 0 is maintained for half-second.
During a trip condition, the indicator stays on.
OVERLOAD_Signal reflects the following five conditions:
1. OPNWND: Open winding RTD alarm condition.
2. WNDALM: Alarm condition due to the winding RTD temperature.
3. CUTALM: Alarm condi-tion due to the calculated copper temperature.
~. WNDTRP: Trip condition due to the winding RTD temperature.

3~

-4~-. CUTRIP: Trip condition due to -the calculated copper temperature.
BE~RING Signal reflects the following three conditions:
1. OPNBRG: Open bearing RTD alarm condition.
2. BRGALM: Alarm condition due to the bearing RTD
temperature.
3. BOTTRP: Trip condition due to the bearing RT~
temperature .
GROUND FAULT Signal reflects the following two conditions:
-1. GFALRM: Alarm condition due to the ground fault current.
2. GFTRIP: Trip condition due to the ground fault current.
PHASE UNBALANCE Signal reflects the following -two conditions:
1. PUALRM: Alarm condition due to the phase unbalance.
2. PVTRIP: Trip condition due to the phase unbalance.
PHASE REVERSAL Signal reflects the trip condi-tion due -to the -phase reversal (PRTRIP).
OVERCURRENT Signal reflects the trip condition due to the instantaneous overcurrent (IOCTRP).
There are following two signals for the relay drivers.
1. ALMPLS: For the alarm relay.
2. TRIP: For the trip relay.
The ALMPLS signal reflects on alarm condition due to any of the protec-tion function. During an alarm condition this signal is maintained logic 1; otherwise, this signal is pulsing. This arrangement provides -the watchdog timer type action for alarm relay. The TRIP signal reflects a trip condition (TRIPF) due to any of the protection functions.
When this signal is logic 1 the trip relay will assume the trip state provided that the trip relay state-change-permit signal (P24) is logic 1 too. During the inactive mode (Fl = 1) the change in the state of the trip relay is not permitted. However, during the trip test (TRIPT = 1), the change in the state of the trip relay is permitted.
Referring to FIGURE 8, the program for the processor one is organized into the following main modules:

3~!~

l. The Rese-t program (RESET) 156.
2. The Loop program (LOOP) 157.
3. The Timer Interrupt Program (TMRINT) 158.
At the time of the system initialization, -the program control vectors to the beginning of the Reset program. The end of the Reset program marks -the beginning of the Loop program.
The Loop program is a program-loop.
At fixed intervals of 0.16 seconds a timer interrupt is generated to perform some time dependent functions. The execution of the Loop program is momentarily interrupted and the program control vectors to the beginning of the Timer Interrupt program. The program control returns back to the Loop program at the end of the execution of the Timer Interrupt program. The program control stays with the Loop program until the next Timer Interrupt is generated.
The Reset program performs the followin~ functions:
1. Clears the interval data memory of the processor.
Contents of the processor RAM are lost on power loss.
2. Presets the display timer to 3 minutes. The display timer is used to determine 3 minutesof active display time for the function and the value data display. One internal data memory location of the processor is used as this timer.
3. Presets the value of the maximum temperature of the winding RTD's equal to 40C. One internal data memory location of the processor is used to store this value.
4. Preset the value of the calcula-ted copper temperature, which is used to provide the overload protection, to 85C.
5. Starts the internal hardware timer which is used to provide timer interrupts.
THE LOOP PROGRAM
(LOOP) The Loop program performs the following functions:
l. Sends command to permit timer interrupt. Without this command the timer interrupt will not be ~enerated.
(EN TCNTI) 3~

2~ Checks for the ground faul-t alarm and trip conditions.
Prepares messages to reflect present conditions, (GFAULT) 3. Checks for the phase reversal trip conditions. Prepares a message to reflect present condition.
(PHAREV) Checks for the phase unbalance alarm and -trip conditions.
Prepares messages to reflect present conditions.
(PUNBAL) 5. Checks for the instantaneous overcurrent trip condition. Prepares a message to reflect present condition.
~INSTOC) 6. Checks for the overload alarm and trip conditions due to the calculated copper temperature. Prepares messages to reflect present conditions.
(O~LCUT) 7. Checks for the overload alarm and trip conditions due to the maximum of winding RTD temperatures. Prepares messages to reflect present conditions.
(OVLWND) 8. Checks for the bearing overtemperature alarm and trip conditions. Prepares messages to reflect present condition.
(OVTBRG) 9. Checks for the Identification Number stored in the external data memory. Sets present ACTIVE/INACTIVE condition.
(IDNMBR) 10. Prepares and send s the following output data reflecting present conditions:
a. The six LED outputs.
b. The alarm relay output.
c. The trip relay output.
(OUTPUT) 3~

TIMER INTERRUPT PROGRAM
(TMRINT) The Timer Interrupt program performs the following func-tions:
1. Sends command to select the reyisters of Bank 1. The Timer Interrupt program uses the working registers of Bank 1 while the Loop program uses the working registers of Bank 0 to prevent their interference with each otherO
(SEL RBl) 2. Saves present accumulator data, being used by the Loop program, into one internal data memory (RAM) location of the processor. Restores this data back to accumulator at the end of the Timer Interrupt. Thus, the accumulator content will be the same before and after the Timer Interrupt program.
3. Presets the hardware timer to have next timer interrupt every 0.016 second. This results in having timer interrupt every 0.016 second.
4. Updates, every 0.016 second, the ground fault timer which is used to generate the ground fault trip time delay.
One internal data memory (RAM) location of the processor is used as the ground fault timer.
(GFTIME) 5. Updates, every 0.016 second, -the phase reversal timer which is used to generate the phase reversal trip time delay. One internal data memory (RAM) location of the processor.
(PRTIME).
6. Converts, every 0.016 second, the analog inputs to digital data and stores them into the internal data memory (RAM) of the processor.
(ANALOGI) 7. Determines, every 0.016 second, the maximum of the three low gain phase currents input data and stores it into one internal data memory (RAM) location of the processor.
(MAXLOG) 8. Generates one, 0.016 second time base, clock to determine various times such as 1~4 second/ 1 second etc. One internal data memory (RAM) location is used as this clock.
tCLOCK) 9. Updates, every 1.024 second, the display timer which is used to determine 3 minu-tes of active display time for the function and the value data display. one internal data memory (RAM) location is used as this timer.
(ONESEC) 10 10. Performs, every 0.256 second, caLculations to determine present value of the copper temperature which is used to provide overload protection.
(CUTEMP) 11. Prepares, every 0.256 second, the function and the value data to be displayed.
(UPDATE) 12. Determines, every 0.256 second/ the maximum temperatures of the winding RTD's. Prepares messages to reflect open RTD.
MAX~TD) 13. Performs, every 0.256 second, calculations -to provide the trip characteristics for the phase unbalance protection (PHAUNB) 14. Performs, every 0.016 second/ the keyboard service functions~
(KYBORD) 15. Prepares and sends, every 0.016 second, data for the function and the value display (DISPLY) 3Q While the present invention has been described with reference to a specific embodiment thereof it will be obvious to those skilled in the art that various changes and modifications may be made without departing from the invention in its broader aspects.

Claims (8)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. An apparatus adapted to interrupt the coupling of a three phase AC source to a three phase motor responsive to actual phase unbalance exceeding a maximum allowable phase unbalance comprising:
(a) sensing means adapted for providing three phase current signals, each being representative respectively of the current magnitudes in each of the three conductors coupling the source to the motor;
(b) comparison means for deriving the maximum one and the minimum one of said three signals;
(c) first means responsive to said maximum and minimum ones of said three signals for deriving an actual phase unbalance magnitude signal which is a function of the difference between said maximum one and said minimum one of said three signals, (d) second means for operator entry of parameters representative of maximum allowable phase unbalance;
(e) third means responsive to said first and second means for deriving a normalized unbalance current signal representative of the ratio of said actual and said maximum allowable phase unbalance magnitudes;
(f) fourth means responsive to said normalized unbalance current signal exceeding a predetermined magnitude for generating from said unbalance current signal squared, an unbalance level signal approximating the heating characteristic of said motor responsive to phase unbalance;
(g) switching means responsive to said unbalance level signal attaining a predetermined magnitude adapted to interrupt the coupling between the source and the motor.

2. The apparatus of Claim 1 wherein said sensing means comprises means for obtaining the rectified average of said currents in each of said three conductors.

3. The apparatus of Claim 1 wherein said first means further comprises means for comparing said minimum of said three signals to a predetermined value and permitting an unbalance level to be determined only if said predetermined value is not exceeded, whereby premature motor interruption during motor start-up is avoided.

4. The apparatus of Claim 3 wherein said first means further comprises means for limiting the maximum one of said signals to a predetermined value in deriving said actual phase unbalance magnitude signal.

5. The apparatus of Claim 1 wherein said fourth means comprises means for time integrating said squared unbalance current signal to derive said unbalance level signal, wherein said time integrating is accomplished with a predetermined time constant.

6. The apparatus of Claim 5 wherein said fourth means comprises means responsive to said unbalance current signal exceeding said predetermined magnitude, to increment the magnitude of said unbalance level signal by the sequential addition thereto of an incremental signal representative of the square of said unbalance current signal.

7. The apparatus of Claim 6 wherein said fourth means comprises means responsive to the magnitude of said unbalance current signal being below said predetermined magnitude, to decrease the magnitude of said unbalance level signal by sequentially decrementing the magnitude of said unbalance level signal by a predetermined magnitude, whereby motor cooling is simulated.

8. The apparatus of Claim 5, 6 or 7 wherein said second means comprises operator entry of parameters representative of full load motor current and of the ratio of maximum allowable phase unbalance to the full load current, and further comprising means for generating a maximum allowable phase unbalance signal which is a function of the product of the aforesaid parameters.