BEGIN NEW DATA CASE C BENCHMARK DCNEW-17 C 1st of 9 subcases that illustrate GIFU switch usage as described in C a story in the April, 1996, newsletter. A switch opens at peak current! C The general configuration is similar to Prof. Ned Mohan's Exercise 10. It C differs, however, in that there are 2 loads having two different voltages C (150 and 200 volts dc, respectively). The source is 100 volts dc. The C two high-voltage batteries are being charged by the low-voltage battery. C Unlike Exercise 10, this problem has only one TACS-controlled (here, C MODELS-controlled) switch. This carries the GIFU declaration in cols. C 61-64. The diodes are real, independent diodes that will be adjusted C experimentally on the same step that the GIFU switch changes status. C GIFU4.DAT --- As set up from data supplied by Prof. Yoshihiro Murai .010E-6 10.E-6 1.E-5 100 1 1 1 1 -1 399 1 405 5 425 25 500 100 MODELS OUTPUT SIG MODEL ONOFF VAR SIG OUTPUT SIG EXEC IF (t MOD 1.E-5) < 0.4 * 1.E-5 THEN SIG := 1. ELSE SIG := 0. ENDIF ENDEXEC ENDMODEL USE ONOFF AS SW OUTPUT SIG := SIG ENDUSE RECORD SW.SIG AS SIG ENDRECORD ENDMODELS 00N1 N2 1.0E-2 3 00N3 N5 1. 00N4 N6 1. 00N1 N2 1.E+4 N2 N3 1.E+3 0.01 1 BLANK RECORD ENDING BRANCHES 13N2 GIFU SIG 13 11N2 N4 13 11N2 N3 13 BLANK RECORD ENDING SWITCHES 11N1 100. 0.0000000 11N5 150. 0.0000000 11N6 200. 0.0000000 BLANK RECORD ENDING SOURCES C Step Time N1 N2 N2 N2 N2 N2 N2 N1 N2 MODELS C N2 TERRA N4 N3 TERRA N4 N3 N2 N3 SIG C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 C Switch "N2 " to " " closing after 1.00000000E-08 sec. C 100 .1E-5 100. 0.0 -200. -149.86421 9.96455754 0.0 0.0 9.81876757 -.13578997 1.0 C 200 .2E-5 100. 0.0 -200. -149.87712 19.9516477 0.0 0.0 19.8187676 -.12288012 1.0 C 300 .3E-5 100. 0.0 -200. -149.8888 29.9399652 0.0 0.0 29.8187676 -.11119764 1.0 BLANK RECORD ENDING NODE VOLTAGE OUTPUT C 399 .399E-5 100. 0.0 -200. -149.89927 39.829494 0.0 0.0 39.7187676 -.10072641 1.0 C Switch "N2 " to " " opening after 4.00000000E-06 sec. C 400 .4E-5 100. 0.0 -200. -149.89937 39.9293934 0.0 0.0 39.8187676 -.10062584 0.0 C Diode "N2 " to "N3 " closing after 4.00000000E-06 sec. C 401 .401E-5 -89.814879 189.814879 -10.185121 0.0 0.0 0.0 39.7655794 39.8238601 .049299201 0.0 C 402 .402E-5 -89.725118 189.725118 -10.274882 0.0 0.0 0.0 39.6758677 39.7340901 .049249926 0.0 C 900 .9E-5 -54.532699 154.532699 -45.467301 0.0 0.0 0.0 4.5027675 4.53815216 .029931392 0.0 C Diode "N2 " to "N3 " opening after 9.87000000E-06 sec. C Switch "N2 " to " " closing after 1.00000000E-05 sec. C 1000 1.E-5 .02232934 99.9776707 -100.02233 -50.000041 0.0 0.0 0.0 -.02229011 -.02228788 1.0 C maxima : 100. 189.814879 0.0 0.0 39.9293934 0.0 39.7655794 39.8238601 .049299201 1.0 C Times of max : .2E-7 .401E-5 0.0 0.0 .4E-5 0.0 .401E-5 .401E-5 .401E-5 0.0 C minima : -89.814879 0.0 -200. -149.89937 0.0 0.0 -.03812688 -.02813821 -.14975663 0.0 C Times of min : .401E-5 0.0 .2E-7 .4E-5 0.0 0.0 .987E-5 .988E-5 .2E-7 .4E-5 PRINTER PLOT 185 1. 0.0 10. BRANCH { Limits: (-2.000, 1.898) N2 N2 N3 N2 N4 195 1. 0.0 10. BRANCH { Limits: (-0.004, 3.993) N2 N2 N3 N2 N4 BLANK end of plot cards BEGIN NEW DATA CASE C 2nd of 9 subcases that illustrate GIFU switch usage. C GIFU5.DAT --- As set up from data supplied by Prof. Yoshihiro Murai PRINTED NUMBER WIDTH, 10, 2, { 10-digit col width including 2 blank separators 0.05E-6 5.0E-6 1.E-8 10 1 1 1 -1 79 1 85 5 90 10 MODELS OUTPUT SIG MODEL ONOFF VAR SIG OUTPUT SIG EXEC IF (T MOD 1.E-5) < 0.4*1.E-5 THEN SIG := -1. ELSE SIG := 1. ENDIF ENDEXEC ENDMODEL USE ONOFF AS SW OUTPUT SIG := SIG ENDUSE RECORD SW.SIG AS SIG ENDRECORD ENDMODELS C - - BRANCH DATA C --------- LOAD RESISTANCE AND INDUCTANCE 00N1 N2 10.0 3 00N3 N6 1.0 3 00N2 N3 1.E+3 0.01 3 BLANK RECORD ENDING BRANCHES C --------- SWITCHES BELOW REPRESENT COMBINATION OF SWITCH AND ITS ANTIPARALLEL C CONNECTED DIODE 13N2 N4 GIFU SIG 13 11N2 N3 CLOSED 13 BLANK RECORD ENDING SWITCHES 11N1 200.00 11N4 100.00 11N6 120.00 BLANK RECORD ENDING SOURCES C Step Time N1 N3 N2 N2 N2 N2 N2 N1 N3 N2 MODELS C N2 N6 N3 N4 N3 N4 N3 N2 N6 N3 SIG C *** Switch "N2 " to "N3 " closed before 0.00000000E+00 sec. C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -1. C 10 .5E-6 72.72727 7.272727 0.0 27.27273 0.0 0.0 7.272727 7.272727 7.272727 0.0 -1. BLANK RECORD ENDING NODE VOLTAGE OUTPUT C 79 .395E-5 72.72727 7.272727 0.0 27.27273 0.0 0.0 7.272727 7.272727 7.272727 0.0 -1. C Switch "N2 " to "N4 " closing after 4.00000000E-06 sec. C 80 .4E-5 72.72727 7.272727 0.0 27.27273 0.0 0.0 7.272727 7.272727 7.272727 0.0 1.0 C Diode "N2 " to "N3 " opening after 4.00000000E-06 sec. C 81 .405E-5 100. -.01993 -19.9801 0.0 -19.9801 10.01993 0.0 10. -.01993 -.01993 1.0 C 100 .5E-5 100. -.018126 -19.9819 0.0 -19.9819 10.01813 0.0 10. -.018126 -.018126 1.0 C maxima : 100. 7.272727 0.0 27.27273 0.0 10.01993 7.272727 10. 7.272727 0.0 1.0 C Times of max : .405E-5 .5E-7 0.0 .5E-7 0.0 .405E-5 .5E-7 .405E-5 .5E-7 0.0 .4E-5 C minima : 0.0 -.01993 -19.9819 0.0 -19.9819 0.0 0.0 0.0 -.01993 -.01993 -1. C Times of min : 0.0 .405E-5 .5E-5 0.0 .5E-5 0.0 0.0 0.0 .405E-5 .405E-5 0.0 PRINTER PLOT 185 1. 0.0 5. BRANCH { Limits: (-1.998, 2.727) N2 N3 N2 N4 195 1. 0.0 5. BRANCH { Limits: (0.000, 1.002) N2 N3 N2 N4 BLANK end of plot cards BEGIN NEW DATA CASE C 3rd of 9 subcases that illustrate GIFU switch usage. This differs from C the preceding subcase in that the MODELS-controlled switch used there has C been replaced by a thyristor (which is only simi-MODELS-controlled; yes C the closing is controlled, but opening is natural, on a current zero). C GIFU6.DAT --- As set up from data supplied by Prof. Yoshihiro Murai PRINTED NUMBER WIDTH, 10, 2, { 10-digit col width including 2 blank separators 0.05E-6 5.0E-6 10 1 1 1 1 -1 79 1 85 5 90 10 MODELS OUTPUT SIG MODEL ONOFF VAR SIG OUTPUT SIG EXEC IF (T MOD 1.E-5) < 0.4*1.E-5 THEN SIG := -1. ELSE SIG := 1. ENDIF ENDEXEC ENDMODEL USE ONOFF AS SW OUTPUT SIG := SIG ENDUSE RECORD SW.SIG AS SIG ENDRECORD ENDMODELS C --------- LOAD RESISTANCE AND INDUCTANCE 00N1 N2 10.0 3 00N3 N6 1.0 3 00N2 N3 1.E+3 0.01 3 BLANK RECORD ENDING BRANCHES C --------- SWITCHES BELOW REPRESENT COMBINATION OF SWITCH AND ITS ANTIPARALLEL C CONNECTED DIODE 11N2 N4 GIFUSIG 13 11N2 N3 CLOSED 13 BLANK RECORD ENDING SWITCHES 11N1 200.00 11N4 100.00 11N6 120.00 BLANK RECORD ENDING SOURCES C Step Time N1 N3 N2 N2 N2 N2 N2 N1 N3 N2 MODELS C N2 N6 N3 N4 N3 N4 N3 N2 N6 N3 SIG C *** Switch "N2 " to "N3 " closed before 0.00000000E+00 sec. C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -1. C 10 .5E-6 72.72727 7.272727 0.0 27.27273 0.0 0.0 7.272727 7.272727 7.272727 0.0 -1. BLANK RECORD ENDING NODE VOLTAGE OUTPUT C 79 .395E-5 72.72727 7.272727 0.0 27.27273 0.0 0.0 7.272727 7.272727 7.272727 0.0 -1. C Valve "N2 " to "N4 " closing after 4.00000000E-06 sec. C 80 .4E-5 72.72727 7.272727 0.0 27.27273 0.0 0.0 7.272727 7.272727 7.272727 0.0 1.0 C Diode "N2 " to "N3 " opening after 4.00000000E-06 sec. C 81 .405E-5 100. -.01993 -19.9801 0.0 -19.9801 10.01993 0.0 10. -.01993 -.01993 1.0 C 100 .5E-5 100. -.018126 -19.9819 0.0 -19.9819 10.01813 0.0 10. -.018126 -.018126 1.0 C maxima : 100. 7.272727 0.0 27.27273 0.0 10.01993 7.272727 10. 7.272727 0.0 1.0 C Times of max : .405E-5 .5E-7 0.0 .5E-7 0.0 .405E-5 .5E-7 .405E-5 .5E-7 0.0 .4E-5 C minima : 0.0 -.01993 -19.9819 0.0 -19.9819 0.0 0.0 0.0 -.01993 -.01993 -1. C Times of min : 0.0 .405E-5 .5E-5 0.0 .5E-5 0.0 0.0 0.0 .405E-5 .405E-5 0.0 PRINTER PLOT 185 1. 0.0 5. BRANCH { Limits: (-1.998, 2.727) N2 N3 N2 N4 195 1. 0.0 5. BRANCH { Limits: (0.000, 1.002) N2 N3 N2 N4 BLANK end of plot cards BEGIN NEW DATA CASE C 4th of 9 subcases that illustrate GIFU switch usage. This differs from C the preceding subcase in that there is no diode; there are two thyristors C and both are declared to be GIFU switches. C GIFU7.DAT --- As set up from data supplied by Prof. Yoshihiro Murai PRINTED NUMBER WIDTH, 9, 2, { 9-digit col width including 2 blank separators 1.00E-6 40.0E-6 1 1 1 1 1 -1 5 5 30 1 35 5 MODELS OUTPUT SIG, SIGM, SIGN, SIGO, MODEL ONOFF VAR SIG, SIGM, SIGN, SIGO, OUTPUT SIG, SIGM, SIGN, SIGO, EXEC IF (T MOD 1.0E-4) > 1.0*1.E-5 THEN SIG := -1. ELSE SIG := 1. ENDIF IF (T MOD 1.0E-4) > 3.0*1.E-5 THEN SIGM := 1. ELSE SIGM := -1. ENDIF IF (T MOD 1.0E-4) > 4.0*1.E-5 THEN SIGN := 1. ELSE SIGN := -1. ENDIF SIGO := -1.+(SIGM - SIGN) ENDEXEC ENDMODEL USE ONOFF AS SW OUTPUT SIG := SIG OUTPUT SIGM := SIGM OUTPUT SIGN := SIGN OUTPUT SIGO := SIGO ENDUSE RECORD SW.SIG AS SIG SW.SIGM AS SIGM SW.SIGN AS SIGN SW.SIGO AS SIGO ENDRECORD ENDMODELS C --------- LOAD RESISTANCE AND INDUCTANCE 00N1 N2 10.0 3 00N3 N6 1.0 3 BLANK RECORD ENDING BRANCHES C --------- SWITCHES BELOW REPRESENT COMBINATION OF SWITCH AND ITS ANTIPARALLEL C CONNECTED DIODE 11N2 N4 GIFUSIGM 13 11N2 N3 CLOSEDGIFUSIG 13 BLANK RECORD ENDING SWITCHES 11N1 200.00 11N4 100.00 11N6 120.00 BLANK RECORD ENDING SOURCES C Step Time N1 N3 N2 N2 N2 N2 N1 N3 MODELS MODELS MODELS MODELS C N2 N6 N4 N3 N4 N3 N2 N6 SIG SIGM SIGN SIGO C *** Switch "N2 " to "N3 " closed before 0.00000000E+00 sec. C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 -1. -1. -1. C 1 1.E-6 72.7273 7.27273 27.2727 0.0 0.0 7.27273 7.27273 7.27273 1.0 -1. -1. -1. C 2 .2E-5 72.7273 7.27273 27.2727 0.0 0.0 7.27273 7.27273 7.27273 1.0 -1. -1. -1. BLANK RECORD ENDING NODE VOLTAGE OUTPUT C 30 .3E-4 72.7273 7.27273 27.2727 0.0 0.0 7.27273 7.27273 7.27273 -1. -1. -1. -1. C Valve "N2 " to "N4 " closing after 3.10000000E-05 sec. C 31 .31E-4 72.7273 7.27273 27.2727 0.0 0.0 7.27273 7.27273 7.27273 -1. 1.0 -1. 1.0 C Valve "N2 " to "N3 " opening after 3.10000000E-05 sec. C 32 .32E-4 100. 0.0 0.0 -20. 10. 0.0 10. 0.0 -1. 1.0 -1. 1.0 C 40 .4E-4 100. 0.0 0.0 -20. 10. 0.0 10. 0.0 -1. 1.0 -1. 1.0 C maxima : 100. 7.27273 27.2727 0.0 10. 7.27273 10. 7.27273 1.0 1.0 -1. 1.0 C Times of max : .32E-4 1.E-6 1.E-6 0.0 .32E-4 1.E-6 .32E-4 1.E-6 0.0 .31E-4 0.0 .31E-4 C minima : 0.0 0.0 0.0 -20. 0.0 0.0 0.0 0.0 -1. -1. -1. -1. C Times of min : 0.0 0.0 0.0 .32E-4 0.0 0.0 0.0 0.0 .11E-4 0.0 0.0 0.0 PRINTER PLOT 185 8. 0.0 40. BRANCH { Limits: (-2.000, 2.727) N2 N3 N2 N4 195 8. 0.0 40. BRANCH { Limits: (0.000, 1.000) N2 N3 N2 N4 BLANK end of plot cards BEGIN NEW DATA CASE C 5th of 9 subcases that illustrate GIFU switch usage. This came from C Rajendra Naik, a graduate student of Prof. Ned Mohan in Minneapolis. It C stacks two identical power supplies in series. This is more demanding C than preceding cases in that a second experimental step is required for C to switch the second half. So, on the screen, after table dumping will C be seen two "table restoration" lines. C NAIK1.DAT --- As set up using data from grad student Rajendra Naik PRINTED NUMBER WIDTH, 11, 1, { 11-digit col width including 1 blank separators 0.025E-6 6.0E-6 1 1 1 1 1 -1 5 5 20 20 100 50 199 1 205 5 TACS HYBRID C --------- VCONTL IS THE CONTROL VOLTAGE WHICH DETERMINES THE DUTY RATIO C --------- RAMP VOLTAGE REACHES 1.0 IN ONE TIME-PERIOD OF 10 MICROSECONDS 11VCONTL 0.5 24RAMP 1.0 10.0E-6 98SIG1 = VCONTL .GE. RAMP 98SIG2 = SIG1 C --------- TACS VARIABLES TO BE OUTPUTTED 33SIG1 BLANK RECORD ENDING TACS C - - BRANCH DATA FOR TOP BUCK 00VD SWITCP 1.0E-3 00VOI VO 5.0E-3 3 00VO VOI 3000.0 00VN DIODEU 1.0E-3 1 00VO VOUTM 1000.0 1 00VOUTM VO 0.1 1 C - - BRANCH DATA FOR LOWER BUCK 00SWITCNVNEG 1.0E-3 00VLI DIODEL 1.0E-3 00VN 1.0E-3 00VLI VNEG1 3000.0 00VNEG1 VLI 5.0E-3 1 00VNEG1 VOUTM 1000.0 00VOUTM VNEG1 0.1 00VOI 10.0 5.E-02 00VLI 10.0 5.E-02 BLANK RECORD ENDING BRANCHES 13SWITCPVOI CLOSEDGIFU SIG1 11 11DIODEUVOI 11 13VLI SWITCN CLOSEDGIFU SIG2 11 11DIODELVN 11 BLANK RECORD ENDING SWITCHES 11VD 10.0 11VNEG -10. BLANK RECORD ENDING SOURCES C - - INITIAL CONDITIONS 2VO 5.0 2VOI 10.0 3VO VOUTM 5.0 3VOI VO 0.0 2VNEG1 -5.0 2VLI -10. C - - NODE VOLTAGES TO BE OUTPUTTED VO VOI VNEG1 VOUTM VLI C Step Time VOI VO VOI VNEG1 VOUTM VLI SWITCP DIODEU VLI DIODEL C VO VOI VOI SWITCN VN C C VOI VN VO VOUTM VNEG1 TACS C VO DIODEU VOUTM VO VLI SIG1 C *** Switch "SWITCP" to "VOI " closed before 0.00000000E+00 sec. C *** Switch "VLI " to "SWITCN" closed before 0.00000000E+00 sec. C 0 0.0 5.0 5.0 10. -5. 0.0 -10. 0.0 0.0 0.0 0.0 C .86736E-18 0.0 0.0 -50. .0125 0.0 C 1 .25E-7 4.70586567 5.29318057 9.99904624 -9.7049453 .293805171 -9.9990462 .953759273 0.0 .953759273 0.0 C .024264664 0.0 -49.967921 -49.993754 .025735252 1.0 C 2 .5E-7 .815038563 9.18404043 9.999079 -5.810338 4.18591392 -9.999079 .921003104 0.0 .921003104 0.0 C .038066925 0.0 -49.942927 -49.981265 .036942357 1.0 BLANK RECORD ENDING NODE VOLTAGE OUTPUT C 199 .4975E-5 2.85398023 7.14334065 9.99732089 -7.1433406 2.37896557 -9.9973209 2.67911339 0.0 2.67911339 0.0 C 2.67814375 0.0 -44.964656 -47.643751 2.67814375 1.0 C Switch "SWITCP" to "VOI " opening after 5.00000000E-06 sec. C Switch "VLI " to "SWITCN" opening after 5.00000000E-06 sec. C 200 .5E-5 2.85568452 7.14162209 9.99730661 -7.1416222 2.37837081 -9.9973066 2.69338583 0.0 2.69338583 0.0 C 2.69241791 0.0 -44.939143 -47.632513 2.69241791 0.0 C Diode "DIODEU" to "VOI " closing after 5.00000000E-06 sec. C Diode "DIODEL" to "VN " closing after 5.00000000E-06 sec. C 201 .5025E-5 -7.141608 7.13990415 -.00170382 -7.1399041 2.37777637 .001703825 0.0 1.7038246 0.0 1.7038246 C 2.6817031 1.7038246 -44.941955 -47.621278 2.6817031 0.0 C 240 .6E-5 -7.073708 7.07255339 -.00115462 -7.0725534 2.35470566 .001154618 0.0 1.154618 0.0 1.154618 C 1.29567377 1.154618 -45.885161 -47.178477 1.29567377 0.0 C maxima : 5.0 9.18404043 10. -5. 4.18591392 .001742607 2.69338583 1.74260701 2.69338583 1.74260701 C 2.69241791 1.74260701 0.0 -47.178477 2.69241791 1.0 C Times of maxima: 0.0 .5E-7 0.0 0.0 .5E-7 .5175E-5 .5E-5 .5175E-5 .5E-5 .5175E-5 C .5E-5 .5175E-5 0.0 .6E-5 .5E-5 .25E-7 C minima : -7.141608 5.0 -.00174261 -9.7049453 0.0 -10. 0.0 0.0 0.0 0.0 C .86736E-18 0.0 -49.967921 -50. .0125 0.0 C Times of minima: .5025E-5 0.0 .5175E-5 .25E-7 0.0 0.0 0.0 0.0 0.0 0.0 C 0.0 0.0 .25E-7 0.0 0.0 0.0 PRINTER PLOT 185 1. 0.0 6.0 VOI VO VOI DIODEU { Limits: (-7.142, 5.000) CALCOMP PLOT { Switch to vector plotting to illustrate superposition of plots SUPERIMPOSE 3 { Superimpose plots of the following 3 cards 2Next, illustrate the overlay of 3 plots 145 .6 0.0 6.0-10. 10. VO Superimpose 3 V-node in volts 185 .6 0.0 6.0 VOI VO { 2nd of 3 in overlay is V-branch 195 .6 0.0 6.0 VOI SWITCP { 3rd of 3 in overlay is I-branch C Ok, done with SUPERIMPOSE. Next, prove the return to normal plotting: 2Return to normal plotting 195 .6 0.0 6.0 VOI SWITCPDIODEUVOI BLANK card ending plot cards BEGIN NEW DATA CASE C 6th of 9 subcases. This illustrates switch dependence as 1st described C in the January, 1997, newsletter. This is modification of 1st subcase. C It is more efficient that using a GIFU switch because no table dumping C and restoring is involved. The solution is identical. .010E-6 10.E-6 1.E-5 100 1 1 1 1 -1 399 1 405 5 425 25 500 100 MODELS OUTPUT SIG MODEL ONOFF VAR SIG OUTPUT SIG EXEC IF (t MOD 1.E-5) < 0.4 * 1.E-5 THEN SIG := 1. ELSE SIG := 0. ENDIF ENDEXEC ENDMODEL USE ONOFF AS SW OUTPUT SIG := SIG ENDUSE RECORD SW.SIG AS SIG ENDRECORD ENDMODELS 00N1 N2 1.0E-2 3 00N3 N5 1. 00N4 N6 1. 00N1 N2 1.E+4 N2 N3 1.E+3 0.01 1 BLANK RECORD ENDING BRANCHES 13N2 { GIFU } SIG 13 11N2 N4 13 11N2 N3 13 N2 N2 N3 < Master/slave switch dependency > ONLY OPEN,CLOSE BLANK RECORD ENDING SWITCHES 11N1 100. 0.0000000 11N5 150. 0.0000000 11N6 200. 0.0000000 BLANK RECORD ENDING SOURCES C Step Time N1 N2 N2 N2 N2 N2 N2 N1 N2 MODELS C N2 TERRA N4 N3 TERRA N4 N3 N2 N3 SIG C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 C Switch "N2 " to " " closing after 1.00000000E-08 sec. C 100 .1E-5 100. 0.0 -200. -149.86421 9.96455754 0.0 0.0 9.81876757 -.13578997 1.0 C 200 .2E-5 100. 0.0 -200. -149.87712 19.9516477 0.0 0.0 19.8187676 -.12288012 1.0 C 300 .3E-5 100. 0.0 -200. -149.8888 29.9399652 0.0 0.0 29.8187676 -.11119764 1.0 BLANK RECORD ENDING NODE VOLTAGE OUTPUT C 399 .399E-5 100. 0.0 -200. -149.89927 39.829494 0.0 0.0 39.7187676 -.10072641 1.0 C Switch "N2 " to " " opening after 4.00000000E-06 sec. C 400 .4E-5 100. 0.0 -200. -149.89937 39.9293934 0.0 0.0 39.8187676 -.10062584 0.0 C Diode "N2 " to "N3 " closing after 4.00000000E-06 sec. C 401 .401E-5 -89.814879 189.814879 -10.185121 0.0 0.0 0.0 39.7655794 39.8238601 .049299201 0.0 C 402 .402E-5 -89.725118 189.725118 -10.274882 0.0 0.0 0.0 39.6758677 39.7340901 .049249926 0.0 C 900 .9E-5 -54.532699 154.532699 -45.467301 0.0 0.0 0.0 4.5027675 4.53815216 .029931392 0.0 C Diode "N2 " to "N3 " opening after 9.87000000E-06 sec. C Switch "N2 " to " " closing after 1.00000000E-05 sec. C 1000 1.E-5 .02232934 99.9776707 -100.02233 -50.000041 0.0 0.0 0.0 -.02229011 -.02228788 1.0 C maxima : 100. 189.814879 0.0 0.0 39.9293934 0.0 39.7655794 39.8238601 .049299201 1.0 C Times of max : .2E-7 .401E-5 0.0 0.0 .4E-5 0.0 .401E-5 .401E-5 .401E-5 0.0 C minima : -89.814879 0.0 -200. -149.89937 0.0 0.0 -.03812688 -.02813821 -.14975663 0.0 C Times of min : .401E-5 0.0 .2E-7 .4E-5 0.0 0.0 .987E-5 .988E-5 .2E-7 .4E-5 185 1. 0.0 10. BRANCH { Limits: (-2.000, 1.898) N2 N2 N3 N2 N4 195 1. 0.0 10. BRANCH { Limits: (-0.004, 3.993) N2 N2 N3 N2 N4 BLANK end of plot cards BEGIN NEW DATA CASE C 7th of 9 subcases. This illustrates a Type-13 TACS-controlled switch C as supplied by Dr. Tsuyoshi Funaki of Osaka University. Original data C was attached to E-mail dated May 30, 2001. The TACS-controlled switch C (BB, CC) will open at 3.25 msec while it still carries substantial C current. Of course, diode (EE, CC) must close simultaneously. This C is the "dynamic current redirection." But it was not happening with C the original data because the diode was declared to be a GIFU switch. C Removing this, operation passes through 3.25 msec correctly. But C a subsequent challenge then occurs at 3.33 msec, when the diode has C a current zero. This is not handled properly. The diode would flop open C and closed repeatedly (3 dT is the period) if T-max were extended. The C screen plots document this. About time step size dT, it is largely C arbitrary. Dr. Funaki's data used 2 usec, which WSM increased to 10 to C reduce the printout while documenting each step. Unfinished business: C the diode oscillation that begins at 3.33 msec could be eliminated by the C addition of a normal snubber circuit. This is because opening is at a C current zero. It is quite unlike opening of the TACS-controlled switch C in this respect. Final observation: CLOSED was added to the Type-13 C TACS-controlled switch simply to save one experimental GIFU step, which C requires that tables be dumped once. If removed, the simulation will C not change substantially. PRINTED NUMBER WIDTH, 10, 2, { Column width is 10, including 2 blank spaces .000010 .003380 { T-max is chosen to end the simulation once oscillation begins 10 1 1 1 0 -1 320 1 { Begin printing every step shortly before Type-13 switch opens C $DEPOSIT, D4FACT=-1. { Use SPY DEPOSIT to change STARTUP value to hold plot TACS HYBRID 14CTL 1. 60. 20. 1. BLANK card ending all TACS data AA BB 10. 0 CC GG 1. 3 DD EE 10. 0 GG 1. 0 BLANK card ending BRANCH cards 13BB CC { switch controlled by TACS CTL } CLOSEDGIFU CTL 13 11EE CC 13 BLANK card ending SWITCH cards 14AA 10. 60. -1. 1. 14DD 10. 60. -180. -1. 1. BLANK card ends SOURCE cards C First 4 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 3 output variables are branch currents (flowing from the upper node to the lower node); C Step Time CC BB EE AA BB EE CC C GG CC CC CC CC GG C *** Phasor I(0) = 9.0802438E-01 Switch "BB " to "CC " closed in the steady-state. C 0 0.0 .0117319 0.0 -10.9198 10. .9080244 0.0 .9080244 C 10 .1E-3 -.001178 0.0 -10.9003 9.992895 .9085521 0.0 .9085521 AA BLANK card ending node voltage outputs C 323 .00323 -.317139 0.0 -3.4843 3.458223 .3432148 0.0 .3432148 C 324 .00324 -.317625 0.0 -3.44524 3.422825 .3400409 0.0 .3400409 C Switch "BB " to "CC " opening after 3.25000000E-03 sec. C 325 .00325 -.318106 0.0 -3.40614 3.387379 .3368623 0.0 .3368623 C Diode "EE " to "CC " closing after 3.25000000E-03 sec. C 326 .00326 -6.67287 9.722844 0.0 3.351885 0.0 .3019074 .3019074 C 327 .00327 -5.94343 9.020945 0.0 3.316343 0.0 .2388259 .2388259 C 328 .00328 -5.29 8.388096 0.0 3.280754 0.0 .1826588 .1826588 C 329 .00329 -4.70466 7.817091 0.0 3.245118 0.0 .1326855 .1326855 C 330 .0033 -4.1803 7.301479 0.0 3.209436 0.0 .0882607 .0882607 C 331 .00331 -3.71058 6.83548 0.0 3.173709 0.0 .0488063 .0488063 C 332 .00332 -3.28979 6.413917 0.0 3.137936 0.0 .0138045 .0138045 C Diode "EE " to "CC " opening after 3.33000000E-03 sec. C 333 .00333 -2.91282 6.032152 0.0 3.102119 0.0 -.017209 -.017209 C 334 .00334 6.354539 -3.28828 -9.4208 3.066258 0.0 0.0 .34E-17 C Diode "EE " to "CC " closing after 3.35000000E-03 sec. C 335 .00335 -6.35454 9.384892 3.324187 3.030353 0.0 0.0 -.34E-17 C Diode "EE " to "CC " opening after 3.36000000E-03 sec. C 336 .00336 -2.50702 5.545732 0.0 2.994405 0.0 -.044308 -.044308 C 337 .00337 11.36858 -8.41016 -14.327 2.958414 0.0 0.0 .272E-17 C Diode "EE " to "CC " closing after 3.38000000E-03 sec. C 338 .00338 -11.3686 14.29096 8.446196 2.922382 0.0 0.0 -.27E-17 184.023.163.36 -10. 10.BB CC CC GG 194.023.163.36 BB CC CC GG BLANK card ending PLOT cards BEGIN NEW DATA CASE C 8th of 9 subcases that illustrate GIFU switch usage. This is the C same as the 1st subcase, but the switch cards have been reordered. The C GIFU switch has been placed last rather than first. New USE GIFU SWITCH C declaration at the start of switch cards avoids an error stop. Reduce C T-max by a factor of 2 so save computer time. Of course, order of the C output vector is changed since column-80 punches of the switch cards C produced 6 of the output variables (some output variables depend on the C order of branch and switch data. Finally, there will be no connectivity C output or plotting (why bother?). PRINTED NUMBER WIDTH, 11, 1, { Restore width, separation of 1st subcase (STARTUP) .010E-6 5.E-6 1.E-5 100 -1 MODELS OUTPUT SIG MODEL ONOFF VAR SIG OUTPUT SIG EXEC IF (t MOD 1.E-5) < 0.4 * 1.E-5 THEN SIG := 1. ELSE SIG := 0. ENDIF ENDEXEC ENDMODEL USE ONOFF AS SW OUTPUT SIG := SIG ENDUSE RECORD SW.SIG AS SIG ENDRECORD ENDMODELS 00N1 N2 1.0E-2 3 00N3 N5 1. 00N4 N6 1. 00N1 N2 1.E+4 N2 N3 1.E+3 0.01 1 BLANK RECORD ENDING BRANCHES USE GIFU SWITCH { Needed because no diode follows the GIFU switch C Without the preceding request which was added 23 June 2001, this would result: C Halt in OVER4. 1st GIFU switch is not followed by a 2nd or a diode (needed for the experimental step). 11N2 N4 13 11N2 N3 13 13N2 GIFU SIG 13 BLANK RECORD ENDING SWITCHES 11N1 100. 0.0000000 11N5 150. 0.0000000 11N6 200. 0.0000000 BLANK RECORD ENDING SOURCES C First 4 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 5 output variables are branch currents (flowing from the upper node to the lower node); C Next 1 output variables belong to MODELS (with "MODELS" an internally-added upper name of pair). C Step Time N1 N2 N2 N2 N2 N2 N2 N1 N2 MODELS C N2 N4 N3 TERRA N4 N3 TERRA N2 N3 SIG C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 C Switch "N2 " to " " closing after 1.00000000E-08 sec. C 100 .1E-5 100. -200. -149.86421 0.0 0.0 0.0 9.96455754 9.81876757 -.13578997 1.0 C 200 .2E-5 100. -200. -149.87712 0.0 0.0 0.0 19.9516477 19.8187676 -.12288012 1.0 C 300 .3E-5 100. -200. -149.8888 0.0 0.0 0.0 29.9399652 29.8187676 -.11119764 1.0 C Switch "N2 " to " " opening after 4.00000000E-06 sec. C 400 .4E-5 100. -200. -149.89937 0.0 0.0 0.0 39.9293934 39.8187676 -.10062584 0.0 C Diode "N2 " to "N3 " closing after 4.00000000E-06 sec. C 500 .5E-5 -81.349976 -18.650024 0.0 181.349976 0.0 31.3053236 0.0 31.358111 .044652391 0.0 BLANK RECORD ENDING NODE VOLTAGE OUTPUT BLANK end of plot cards BEGIN NEW DATA CASE C 9th of 9 subcases. This illustrates a Type-13 TACS-controlled switch C as supplied by Dr. Tsuyoshi Funaki of Osaka University. Original data C was attached to E-mail dated 31 October 2002. The TACS-controlled switch C was erroneously controlled as part of the GIFU experimental step. But C by definition, TACS controls this, so such action was not reasonable. C A 3-line addition to OVER16 makes this solution right. Only 752 steps C need be taken because step 752 was the first of many that were wrong. C One does not need to go any farther than 752 to observe the correction. C But, for plotting (now disabled), we want nice scaling, so use 800. C adp, 9, 0, 0, 0, { To see GIFU switch diagnostic, enable this comment card. 1.E-7 800.E-7 1.E-8 { Restore default EPSILN on 14 Jan 03 100 1 1 1 0 -1 0 0 0 749 -1 TACS HYBRID 11N3 50. 1. 11N1 100. 1. 14CTL 1. 1.E4 1. 33CTL BLANK card ending TACS data N2 ND .001 3 N2 N3 1. 2.5 3 BLANK card ending branches 13N2 N1 CLOSEDGIFU CTL 13 11 ND 13 C Note about the GIFU switch. With this orientation, the answer became bad C on step 752. But the answer was correct if the two ends of the GIFU switch C were reversed. I.e., (N2, N1) was bad but (N1, N2) was correct. This C is what was traced to OVER16 where 3 added lines now exempt Type-13 switch C from the GIFU experimentation. Before, it was not being excluded. Dr. C Funaki's 2 data cases were named CHOPD.DAT and CHOPD2.DAT with the C later having a different solution beginning on step 752. BLANK card ending switches 60N3 0 10. 60N1 0 10. BLANK card ending source cards BLANK card ending node voltage outputs C 185.8 48. 56. BRANCH C N2 N1 N2 N3 ND BLANK card ending batch-mode plot cards BEGIN NEW DATA CASE BLANK