====== VATT Pseudo Hexapod (Legacy) ====== The VATT secondary is focused and collimated with a Pseudo Hexapod. Motion between the axes are coupled but as little as possible and we get uncoupled positions with some linear algebra and iterations. The pseudo hexapod is to be retired for a PI hexapod which will be installed summer 2017 ===== Software ===== The VATT secondary is controlled using a c program that runs on vatttel. You can find the source in our mogit repository. The most important source file is secmove.c, which controls the autofocus, autocollimation and closed loop motion. A block diagram and the source code for this module are below. {{:vatt:secmove_flowchar.png?800|}} /* VATT secondary move routine * The six user motions are all cross coupled * into the six motors on the VATT secondary. * Hence, all the matrix math. * This will be a replacement for winsec.c */ #define abs(a) ((a) < 0 ? -(a) : (a)) extern void sprintf(); extern void printf(); extern int sec_checkVoltsOK(); extern void movev(); extern void vme8_kill(); double sec_dv = 0.03; /* maximum allowable voltage difference */ double sec_dvmax = 0.2; /* maximum before kill */ double sec_min_temp = -40.0; /* minimum sane temperature */ double sec_max_temp = 50.0; /* maximum sane temperature */ double sec_min_cosel = 0.0; /* minimum sane cosine(elevation) */ double sec_max_cosel = 1.0; /* maximum sane cosine(elevation) */ double sec_lastgood_temp; /* last sane temperature */ double sec_lastgood_cosel; /* last sane cosine(elevation) */ int sec_max_mloops = 5; /* maximum motor loops */ #define V 6 /* number of motors/sensors */ #define M 6 /* number of user motions */ #define TIPX 0 /* arcseconds */ #define TIPY 1 #define TIPZ 2 #define DECENX 3 /* microns */ #define DECENY 4 #define FOCUS 5 double sec_threshold[M] = {2.5, 2.5, 2.5, 4.5, 4.5, 1.0}; char sec_errormessage[100]; /* error message */ double sec_u0[M]; /* previous user values */ double sec_u[M]; /* user values */ double sec_c[M]; /* corrections */ double sec_m[M]; /* target motions */ double sec_t[V]; /* target voltages */ double sec_v[V]; /* true voltages */ double sec_p[M]; /* true positions */ /* The immediately above positions are derived from the lvdt readings * and multiplied by v2m, then the zero offset is removed. */ #define C 4 /* number of corrections */ #define FTC 0 #define FLC 1 #define CTC 2 #define CLC 3 /* the original Steve West matrix */ double sec_m2v_sw[V][M] = { {0.0029680, 0.0051070, -0.0001555, -0.0000931, -0.0001510, -0.0069780}, {0.0016000, -0.0007957, -0.0033810, -0.0033320, -0.0058000, -0.0003111}, {0.0029300, -0.0050660, -0.0000175, -0.0000407, 0.0000389, -0.0069420}, {-0.001475, -0.0010590, -0.0032270, -0.0031640, 0.0057680, -0.0004334}, {-0.005346, -0.0000617, 0.0000176, 0.0000674, 0.0001066, -0.0063050}, {-0.0001785, 0.0017840, -0.0034270, 0.0065690, 0.0000873, -0.0002048} }; /* All new values per cromwell 10/6/97 * Most of these are now updated in startup.cmd */ double sec_z[M] = {0.0, 0.0, 0.0, 0.0, 0.0, 0.0}; double sec_z0[M] = {0.0, 0.0, 0.0, 0.0, 0.0, 0.0}; double sec_m2v[V][M] = { {0.0029680, 0.0051070, -0.0001555, -0.0000931, -0.0001510, -0.0069780}, {0.0016000, -0.0007957, -0.0033810, -0.0033320, -0.0058000, -0.0003111}, {0.0029300, -0.0050660, -0.0000175, -0.0000407, 0.0000389, -0.0069420}, {-0.001475, -0.0010590, -0.0032270, -0.0031640, 0.0057680, -0.0004334}, {-0.005346, -0.0000617, 0.0000176, 0.0000674, 0.0001066, -0.0063050}, {-0.0001785, 0.0017840, -0.0034270, 0.0065690, 0.0000873, -0.0002048} }; double sec_v2m[M][V] = { { 55.4830132, -3.6541212, 58.2489128, -0.3046370, -125.3519897, 0.4331740, }, { 98.1537399, -3.2015131, -98.2364426, 0.0305685, -0.2867458, -0.8237946, }, { 3.4616468, -100.5291290, 5.1929183, -99.6347046, 5.4746866, -98.9561157, }, { -25.0618916, -50.5131302, 29.3149300, -53.1427689, -1.6702332, 100.8553619, }, { 16.7931938, -85.4085464, 12.2465286, 88.4728470, -33.9353485, -0.0462751, }, { -47.9787979, 0.8650305, -47.8929024, 0.9076155, -52.8920479, 0.4418988, }, }; double sec_c2m[M][C] = { { 0.0, 0.0, 0.0, 0.0, }, { 0.0, 0.0, 0.0, 0.0, }, { 0.0, 0.0, 0.0, 0.0, }, { 0.0, 0.0, 0.0, 0.0, }, { 0.0, 0.0, 0.0, 0.0, }, { 0.0, 0.0, 0.0, 0.0, }, }; void sec_update_Z (m0, m1, m2, m3, m4, m5) double m0; double m1; double m2; double m3; double m4; double m5; { int m; double mm[M]; mm[0] = m0; mm[1] = m1; mm[2] = m2; mm[3] = m3; mm[4] = m4; mm[5] = m5; for (m = 0; m < M; m++) { sec_z[m] = mm[m]; sec_z0[m] = mm[m]; printf ("%.7f\n", mm[m]); } } void sec_update_nelsonCT (m0, m1, m2, m3, m4, m5) double m0; double m1; double m2; double m3; double m4; double m5; { int m; double mm[M]; mm[0] = m0; mm[1] = m1; mm[2] = m2; mm[3] = m3; mm[4] = m4; mm[5] = m5; for (m = 0; m < M; m++) { sec_c2m[m][CTC] = mm[m]; printf ("%.7f\n", mm[m]); } } void sec_update_nelsonCL (m0, m1, m2, m3, m4, m5) double m0; double m1; double m2; double m3; double m4; double m5; { int m; double mm[M]; mm[0] = m0; mm[1] = m1; mm[2] = m2; mm[3] = m3; mm[4] = m4; mm[5] = m5; for (m = 0; m < M; m++) { sec_c2m[m][CLC] = mm[m]; printf ("%.7f\n", mm[m]); } } void sec_update_nelsonFT (m5) double m5; { sec_c2m[FOCUS][FTC] = m5; } void sec_update_nelsonFL (m5) double m5; { sec_c2m[FOCUS][FLC] = m5; } void sec_engzero () { int m; for (m = 0; m < M; m++) { sec_u[m] -= sec_z0[m]-sec_z[m]; sec_z[m] += sec_z0[m]-sec_z[m]; } } void sec_userzero () { int m; for (m = 0; m < M; m++) { sec_z[m] += sec_u[m]; sec_u[m] = 0.0; } } void sec_logger (orig, mess) char *orig; char *mess; { void vxcall(); vxcall ("_logger_it", orig, mess, 0,0,0,0,0,0,0,0); } int sec_telslewing() { int telslewing = 0; void vxcall(); vxcall ("_telx_sec_telslewing", &telslewing, 0,0,0,0,0,0,0,0,0); return (telslewing); } int sec_working; double sec_azimuth; double sec_derotator; double sec_elevation; /* current altitude in radians */ double sec_temperature; /* current temperature in degrees C */ static int sec_autofocus; static int sec_autocoll; void sec_update_tl () /* update temperature and elevation */ { void vxcall(); vxcall ("_telx_sec_temperature", &sec_temperature, 0,0,0,0,0,0,0,0,0); vxcall ("_telx_sec_elevation" , &sec_elevation , 0,0,0,0,0,0,0,0,0); vxcall ("_telx_sec_azimuth" , &sec_azimuth , 0,0,0,0,0,0,0,0,0); vxcall ("_telx_sec_derotator" , &sec_derotator , 0,0,0,0,0,0,0,0,0); } void sec_autoreset () { } void sec_autofocus_set (status) int status; { sec_autofocus = status; } int sec_autofocus_get () { return (sec_autofocus); } void sec_autocoll_set (status) int status; { sec_autocoll = status; } int sec_autocoll_get () { return (sec_autocoll); } void sec_update_corr () { int m, c; double sc; double temp, cosel; double d[C]; double cos(); sec_update_tl (); temp = sec_temperature; cosel = cos(sec_elevation); /* sanity checks */ if (sec_min_temp < temp && temp < sec_max_temp) { sec_lastgood_temp = temp; } else { temp = sec_lastgood_temp; } if (sec_min_cosel < cosel && cosel < sec_max_cosel) { sec_lastgood_cosel = cosel; } else { cosel = sec_lastgood_cosel; } d[FTC] = sec_autofocus ? temp : 0.0; d[FLC] = sec_autofocus ? cosel : 0.0; d[CTC] = sec_autocoll ? temp : 0.0; d[CLC] = sec_autocoll ? cosel : 0.0; for (m = 0; m < M; m++) { sc = 0.0; for (c = 0; c < C; c++) { sc += sec_c2m[m][c] * d[c]; } sec_c[m] = sc; } } void sec_calct () /* calculate targets */ { int m, v; double t; for (v = 0; v < V; v++) { t = 0.0; for (m = 0; m < M; m++) { sec_m[m] = sec_u[m]+sec_c[m]+sec_z[m]; t += sec_m2v[v][m] * sec_m[m]; } sec_t[v] = t; } } void sec_calcp () /* calculate true positions */ { int m, v; double p; for (m = 0; m < M; m++) { p = 0.0; for (v = 0; v < V; v++) { p += sec_v2m[m][v] * sec_v[v]; } sec_p[m] = p - sec_z[m]; } } int sec_checkMove() { int v; int mstatus = 0; int rstatus = 0; double residual; char errormessage[100]; int strlen(); errormessage[0] = 0; for (v = 0; v < V; v++) { residual = sec_t[v]-sec_v[v]; sprintf (errormessage+strlen(errormessage), " %.4f", residual); if (abs(residual) > sec_dv) mstatus = 1; if (abs(residual) > sec_dvmax) rstatus = 1; } if (mstatus) sprintf (sec_errormessage, "residuals =%s", errormessage); return (rstatus); } void sec_voltsNotOK () { int v; char errormessage[100]; int strlen(); errormessage[0] = 0; for (v = 0; v < V; v++) { sprintf (errormessage+strlen(errormessage), " %.4f", sec_t[v]); } sprintf (sec_errormessage, "bad target =%s", errormessage); } void sec_sprintf (c, array) char *c; double array[]; { int m; int strlen(); *c = 0; for (m = 0; m < M; m++) { sprintf (c+strlen(c), " %12.6f", array[m]); } } void sec_sprintfu (c, array, array0) char *c; double array[]; double array0[]; { int m; int strlen(); *c = 0; for (m = 0; m < M; m++) { sprintf (c+strlen(c), " %d", array[m] != array0[m]); array0[m] = array[m]; } } void sec_log_servc () { char c[100]; double readmotorcurrent (); sprintf (c, "%12.6f %12.6f %12.6f", readmotorcurrent (4), readmotorcurrent (2), readmotorcurrent (3)); sec_logger ("serv_c", c); } void sec_log () { char c[100]; int strlen(); double cos(); sprintf (c, "%12.6f %12.6f %12.6f %12.6f %12.6f", sec_elevation*57.29577951, cos(sec_elevation), sec_temperature, sec_azimuth*57.29577951, sec_derotator*57.2957795); sec_logger ("sec_el", c); sec_sprintf (c, sec_u); sec_sprintfu (c+strlen(c), sec_u, sec_u0); sec_logger ("sec_u", c); sec_sprintf (c, sec_c); sec_logger ("sec_c", c); sec_sprintf (c, sec_z); sec_logger ("sec_z", c); sec_sprintf (c, sec_m); sec_logger ("sec_m", c); sec_sprintf (c, sec_t); sec_logger ("sec_t", c); sec_sprintf (c, sec_v); sec_logger ("sec_v", c); sec_sprintf (c, sec_p); sec_logger ("sec_p", c); sec_log_servc (); } void sec_loop_init () { void getSensors(); int m; getSensors (); sec_calcp (); for (m = 0; m < M; m++) { sec_u[m] = sec_p[m]; sec_u0[m] = sec_u[m]; } sec_log (); } void sec_loop (timedout) int timedout; /* !0 if triggered by timer */ { int n; int m; int big; void strcpy(); void getSensors(); if (!timedout || ((sec_autofocus||sec_autocoll) && !sec_telslewing())) { sec_working = 1; sec_update_corr (); sec_calct (); if (sec_checkVoltsOK (sec_t)) { getSensors (); sec_calcp (); for (n = 0; n < sec_max_mloops; n++) { big = 0; for (m = 0; m < M; m++) { big |= abs(sec_m[m]-sec_z[m]-sec_p[m]) > sec_threshold[m]; } if (!big) break; strcpy (sec_errormessage, "moving"); movev (sec_t); strcpy (sec_errormessage, ""); getSensors (); sec_calcp (); } if (sec_checkMove ()) { vme8_kill (); } } else { sec_voltsNotOK (); sec_calcp (); } sec_log (); sec_working = 0; } } void sec_printf (id, array) char *id; double array[]; { int m; printf ("%s:", id); for (m = 0; m < M; m++) { printf (" %12.6f", array[m]); } printf ("\n"); } void sec_printd (id, array) char *id; int array[]; { int m; printf ("%s:", id); for (m = 0; m < M; m++) { printf (" %12d", array[m]); } printf ("\n"); } extern int sec_s[]; void sec_print () { printf ("%s\n", sec_errormessage); sec_printf ("z", sec_z); sec_printf ("u", sec_u); sec_printf ("c", sec_c); sec_printf ("m", sec_m); sec_printf ("t", sec_t); sec_printd ("s", sec_s); sec_printf ("v", sec_v); sec_printf ("p", sec_p); sec_errormessage[0] = 0; } void telx_log_servc () { sec_log_servc (); } void telx_lvdts () { int v; char errormessage[100]; int strlen(); errormessage[0] = 0; for (v = 0; v < V; v++) { sprintf (errormessage+strlen(errormessage), " %.4f", sec_v[v]); } sprintf (sec_errormessage, "lvdts =%s", errormessage); } void telx_rlvdts () { void getSensors(); getSensors (); telx_lvdts (); sec_calcp (); sec_log (); } void plvdts () { void getSensors(); int v; getSensors (); telx_lvdts (); sec_calcp (); printf ("lvdts ="); for (v = 0; v < V; v++) { printf (" %.4f", sec_v[v]); } printf ("\n"); } void pmcurr () { double readmotorcurrent (); printf ("%12.6f %12.6f %12.6f\n", readmotorcurrent (4), readmotorcurrent (2), readmotorcurrent (3)); } int telvatt_telfocustimeout = 70; int telvatt_telfocuscmdlevel; void telvatt_telfocus (c) char *c; { void sleep(); void secondary_trigger(); int sscanf(); double focus; int timeout = telvatt_telfocustimeout; if ((telvatt_telfocuscmdlevel > 0) && (sscanf(c,"%lf",&focus) == 1)) { sec_u[FOCUS] = focus; sec_working = 1; secondary_trigger(); while (sec_working && timeout--) sleep (1); } sprintf (c, "%.2f", sec_u[FOCUS]); } #ifdef OUTOFDATE /* This is an arbitrary zero point for the user motions. * * It was derived using Tom T.'s goodplace lvdt values * and multiplying by the v2m matrix. * { 4.348, -6.021, 5.151, -5.925, 3.126, -5.148 } double sec_z[M] = {97.0968554,-46.2452714,1763.9614446, 136.6180699,20.2980418,-633.5096204}; * new values 3/25/97 via cromwell->nelson->harvey double sec_z[M] = {-103.1331446,38.2247286,1763.9614446, 150.2880699,-0.5719582,-658.7596204}; double sec_z0[M] = {-103.1331446,38.2247286,1763.9614446, 150.2880699,-0.5719582,-658.7596204}; * new values 5/30/97 */ double sec_z[M] = {-273.5316558,168.0022732,1962.0991461, 26.3139608,-168.5847471,-427.0699187}; double sec_z0[M] = {-273.5316558,168.0022732,1962.0991461, 26.3139608,-168.5847471,-427.0699187}; /* Steve West's original matrix {0.0029680, 0.0051070, -0.0001555, -0.0000931, -0.0001510, -0.0069780}, {0.0016000, -0.0007957, -0.0033810, -0.0033320, -0.0058000, -0.0003111}, {0.0029300, -0.0050660, -0.0000175, -0.0000407, 0.0000389, -0.0069420}, {-0.001475, -0.0010590, -0.0032270, -0.0031640, 0.0057680, -0.0004334}, {-0.005346, -0.0000617, 0.0000176, 0.0000674, 0.0001066, -0.0063050}, {-0.0001785, 0.0017840, -0.0034270, 0.0065690, 0.0000873, -0.0002048} */ /* The following matrix of user motions (across) to lvdt voltages (down) * already has the following correction applied. */ /* There is correction in dtx and dty when applying focus. There are * 2 reasons: the positioner is mounted slightly tilted wrt to * the primary axis, and the primary comes up on its supports * slightly differently each time causing a small but variable * tilt term. For now, this correction approximates the constant * term for mounting misalignment and the tilt for which the primary supports * usually produce. * dtxSlope = -0.1166667; * dtySlope = 0.0166667; */ double sec_m2v[V][M] = { { 0.0029680, 0.0051070, -0.0001555, -0.0000931, -0.0001510, -0.0072391, }, { 0.0016000, -0.0007957, -0.0033810, -0.0033320, -0.0058000, -0.0005110, }, { 0.0029300, -0.0050660, -0.0000175, -0.0000407, 0.0000389, -0.0073683, }, {-0.0014750, -0.0010590, -0.0032270, -0.0031640, 0.0057680, -0.0002790, }, {-0.0053460, -0.0000617, 0.0000176, 0.0000674, 0.0001066, -0.0056823, }, {-0.0001785, 0.0017840, -0.0034270, 0.0065690, 0.0000873, -0.0001542, }, }; /* The following matrix of lvdt voltages (across) to user motions (down) * is the inversion of the above. */ double sec_v2m[M][V] = { { 49.8854752, -3.5532007, 52.6613884, -0.1987483,-131.5227203, 0.4847288, }, { 98.9533844, -3.2159302,-97.4382248, 0.0154418, 0.5947783, -0.8311595, }, { 3.4616475,-100.5291367, 5.1929212,-99.6347046, 5.4746842,-98.9561081, }, {-25.0618954, -50.5131378, 29.3149300,-53.1427727, -1.6702307,100.8553619, }, { 16.7931862, -85.4085312, 12.2465229, 88.4728546, -33.9353409, -0.0462707, }, {-47.9788094, 0.8650309,-47.8929062, 0.9076155, -52.8920441, 0.4418987, }, }; /* The following matrix of (T,L) corrections (across) * {focus_delta_temperature, focus_delta_sin_elevation, * collimation_delta_temperature, collimation_delta_sin_elevation} * to user motion corrections (down) * * These were derived from Cromwell's original (June 10, 1996) coefficients * for ELCdv and TCdv by multiplying by v2m. * 0.2230000 0.0914000 -0.1024000 -0.0426000 0.0031000 0.1702000 * 0.2148000 0.5129000 0.0363000 0.1717000 -0.0712000 0.4409000 */ double sec_c2m[M][C] = { { 0.0, 0.0, 20.3485814, 5.0904193, }, { 0.0, 0.0, 15.6625737, 31.6100656, }, { 0.0, 0.0,-111.7561536,-21.5290905, }, { 0.0, 0.0, 5.2340839, 6.2167347, }, { 0.0, 0.0, -22.1677258, -9.1975216, }, { 15.0, -120.0, -7.4841021, -5.8433957, }, }; /* values as of 10/6/97 */ double sec_c2m[M][C] = { { 0.0, 0.0, 21.2217299, 5.7721492, }, { 0.0, 0.0, 15.5378383, 31.5126765, }, { 0.0, 0.0,-111.7561535,-21.5290909, }, { 0.0, 0.0, 5.2340894, 6.2167361, }, { 0.0, 0.0, -22.1677345, -9.1975223, }, { 15.0, -120.0, -7.4840994, -5.8433935, }, }; void sec_update_cromwellB (v0, v1, v2, v3, v4, v5) double v0; double v1; double v2; double v3; double v4; double v5; { int v, m; double vv[V]; double mm[M]; vv[0] = v0; vv[1] = v1; vv[2] = v2; vv[3] = v3; vv[4] = v4; vv[5] = v5; for (m = 0; m < M; m++) { mm[m] = 0.0; for (v = 0; v < V; v++) { mm[m] += sec_v2m[m][v]*vv[v]; } sec_c2m[m][CLC] = mm[m]; printf ("%.7f\n", mm[m]); } } void sec_update_cromwellB0 () { sec_update_cromwellB (0.2230000,0.0914000,-0.1024000, -0.0426000,0.0031000,0.1702000); } void sec_update_cromwellC (v0, v1, v2, v3, v4, v5) double v0; double v1; double v2; double v3; double v4; double v5; { int v, m; double vv[V]; double mm[M]; vv[0] = v0; vv[1] = v1; vv[2] = v2; vv[3] = v3; vv[4] = v4; vv[5] = v5; for (m = 0; m < M; m++) { mm[m] = 0.0; for (v = 0; v < V; v++) { mm[m] += sec_v2m[m][v]*vv[v]; } sec_c2m[m][CTC] = mm[m]; printf ("%.7f\n", mm[m]); } } void sec_update_cromwellC0 () { sec_update_cromwellC (0.2148000,0.5129000,0.0363000, 0.1717000,-0.0712000,0.44090000); } void sec_update_cromwellZ (v0, v1, v2, v3, v4, v5) double v0; double v1; double v2; double v3; double v4; double v5; { int v, m; double vv[V]; double mm[M]; vv[0] = v0; vv[1] = v1; vv[2] = v2; vv[3] = v3; vv[4] = v4; vv[5] = v5; for (m = 0; m < M; m++) { mm[m] = 0.0; for (v = 0; v < V; v++) { mm[m] += sec_v2m[m][v]*vv[v]; } sec_z[m] = mm[m]; sec_z0[m] = mm[m]; printf ("%.7f\n", mm[m]); } } void sec_update_cromwellZ0 () { sec_update_cromwellZ (0.0,0.0,0.0, 0.0,0.0,0.0); } /* auto variables */ /* focus versus temperature */ double sec_ft0; /* starting point */ double sec_ft1; /* last measurement */ double sec_ft2; /* measurement at last adjustment */ double sec_fdt0; /* accumulated delta */ double sec_fdt; /* delta */ /* focus versus elevation */ double sec_fl0; double sec_fl1; double sec_fl2; double sec_fdsl0; double sec_fdsl; /* collimation versus temperature */ double sec_ct0; double sec_ct1; double sec_ct2; double sec_cdt0; double sec_cdt; /* collimation versus elevation */ double sec_cl0; double sec_cl1; double sec_cl2; double sec_cdsl0; double sec_cdsl; #endif