VLB TEST MEMO N ( L ^. Phae Repone of VLB 16-MHz IF Band J. S. Ulvetad 6 February 1997 1 Introduction Thi note report the reult of an invetigation of the phae repone of the 16-MHz VLBI IF bandpae of the VLB and VL. The meaurement made ue of exiting data from a cientific experiment proceed by the VLB correlator. The invetigation wa motivated by the launch of the VSOP atellite, but it i relevant for any experiment uing 16-MHz bandpae. The et of VLB experiment uing 16-MHz bandpae i rather mall. lthough thi i a tandard mode, it allow no peedup in correlation, o almot all VLB experiment ue IF bandpae of MHz or maller. Several poible experiment for the bandpa phae meaurement were identified by Jon Romney; the mot uitable of thee took place on 3 ugut 1996, and included a 22-GHz obervation of the calibrator 4C 39.25. Thi obervation ued one 16-MHz IF with 2-bit ampling, and wa correlated with 512 channel. Relevant characteritic of the experiment are ummarized in Table 1. It good point are that it contained a trong calibrator, ha high pectral reolution, and alo contain a ingle VL antenna (antenna 27 at pad N). It uffer the diadvantage of including only one IF per tation and of containing no uable data from MK, due to poor recording quality. Tab! e 1: Experiment Propertie Date Freq. Nit Sampling Chan/IF Calibrator ntenna 3ug96 22 GHz 1 2-bit 512 4C 39.25 VLB+Y1 2 Bandpa Meaurement 2.1 Procedure The phae characteritic of the 16-MHz band for the VLB antenna and for Y1 were meaured in a multi-tep proce, uing FD a the reference antenna, ince FD howed the mallet catter in phae ao the bandpa during initial tet. Firt, the data on the calibrator 4C 39.25 were averaged to 64 pectral channel to reduce the computational load, then fringe-fitted uing the IPS tak FRING. The four channel 1
on either end of the band (1-4 and 61-64) were not ued in the FRING olution. Second, the reulting SN (olution) table wa applied to the CL (calibration) table for the data having the full pectral reolution, uing CLCL. Third, BPSS wa ued to compute the phae characteritic of the bandpa for the can on 4C 39.25, with the olution from the fringe-fit applied to remove the ource-dependent phae term. BPSS olution were computed both with and without pectral averaging applied, and the reult were written in BP (bandpa) table. Fourth, POSSM wa ued to plot the bandpa for each tation and to compute the final tatitic. 2.2 Reult The pecification in [1] i for a phae ripple of le than 5 ao the lower 0% of VLB bandpae, and le than 10 ao the upper 20% of the band. Table 2 give the mean and RMS phae for each tation, for the entire bandpa, the lower 0%, and the upper 20%. Table 3 give value for the ame quantitie with Hanning moothing applied in BPSS. Smoothing with a width of four channel wa found to reduce the RMS phae variation ao the band by about 10-30%. In all cae, channel 1-10 and 503-512 were not ued, in order to eliminate the effect of the band edge. The phae characteritic of the band between channel 11 and 502 (0.3 MHz to 15.7 MHz) are hown in Figure 1 through 5. The table how that the RMS phae variation ao the 16-MHz band are le than 2 for all VLB antenna except HN and SC. t both of thee tation, the ignal/noie of the fringe fit i coniderably lower than for the other tation, apparently becaue of reolution of 4C 39.25. Therefore, the intrinic phae characteritic of the 16-MHz band at thee tation are better than indicated; a higher ignal/noie obervation would be neceary to iolate the intrinic propertie. The upper 20% of the band for each tation how little ineae in phae noie when compared to the lower 0%. t about half of the VLB tation (notably HN, KP, PT, and SC), there i evidence for a light roll-off of ~ 5 in the phae in the upper ~ 10% of the band (ee figure a well a mean value of lower and upper p u t of the band in Table 2). Even taking thi into conideration, it appear that the 16-MHz channel for all VLB tation meet the phae pecification given above. The VL (Yl) data were acquired with the VL filter et to their full bandwidth of 50 MHz. The performance of the 16-MHz VLBI bandpa from Yl i omewhat wore than for the VLB tation, and doe not meet the VLB pecification. Figure 5 how, there i a phae ripple ao the band, with almot two cycle ao 16 MHz, and coniderable phae noie at the upper end of the band. Thi may caue a light degradation in the VLBI data on ground baeline due to the mimatched bandpae. For pace VLBI obervation, the phae characteritic are at leat a good a the VSOP paceaft, o the VL i unlikely to caue ignificant degradation on the pace-ground baeline. 3 cknowledgment I thank Jon Romney for identifying the poible experiment for thee meaurement, and Phil Diamond for communicating the proper input for BPSS. Reference [1] VLB Project Book, Verion 7, October 1, 19. 2
Tal >le 2: Phae Characteritic of 16-MHz Bandpa No Smoothing ntenna 100% of Band Lower 0% Upper 20% Mean RMS Mean RMS Mean RMS (deg.) (deg.) (cleg.) (deg.) (deg.) (deg.) BR 0.06 1.2 0.02 1.1 0.22 1.6 FD 0.00 0.00 0.00 0.00 0.00 0.00 HN 0.02 3.4 0.76 3.04-3.16 3.49 KP -1.06 2.22-0.46 1.75-3.59 2.23 L 0.02 1.51 0.10 1.42-0.34 1.79 NL 1.01 1.63 1.17 1.57 0.31 1.70 OV -1.27 1.47-1.1 1.3-1.63 1.75 PT -0.37 2.19-0.04 1.76-1.1 3.06 SC -2.35 5.90-1.94 5.76-4.12 6.15 Y27-2.77 13.7-1.22 14.0-9.41 10.1 3: Phae C laracteritic of! 6-MHz Bandpa Hanning Smoo ntenna 100% of Band Lower 0% Upper 20% Mean RMS Mean RMS Mean RMS (deg.) (deg.) (deg.) (deg.) (deg.) (deg.) BR 0.06 1.40 0.02 1.42 0.23 1.34 FD 0.00 0.00 0.00 0.00 0.00 0.00 HN 0.02 2.97 0.76 2.53-3.15 2.5 KP -1.05 1.9-0.46 1.47-3.5 1.9 L 0.02 1.07 0.10 0.99-0.33 1.2 NL 1.01 1.20 1.17 1.17 0.31 1.09 OV -1.26 1.10-1.1 1.04-1.62 1.30 PT -0.37 1.91-0.04 1.44-1.1 2.1 SC -2.32 4.16-1.90 4.05-4.15 4.15 Y27-2.77 13.4-1.21 14.0-9.49 7.29 3
«5 JS * hfl Plot file verion 31 eated 06-FEB-199715:20:37 4C39.25 JOB5240.UVDT.1 Freq = 22.226 GHz, Bw = 16.000 MHz Bandpa table # 1 C* g. ^ w ff e & 0 O a. <p S w N T3 > > V V * * M t-l # * * * * *** V\ * * 4 W a XI c a -1501 100 200 300 Channel/IF 400 500 Lower frame:bp phae Tlmerange: 00/19:20:25 to 00/19:25:20 ntenna: * Stoke: LL
^ (t 3 to - H J 5- Plot file verion 32 eated 06-FEB-199715:20:37 4C39.25 JOB5240.UVDT.1 Fraq 22.226 GHz, Bw = 16.000 MHz Bandpa table» 1 V v V** > n' * * + * * *. w i oi W M cl TJ n 55 a- W? o a. pi -10-2 -6-10 *p 1 1 4. J * * * * i ~~ *7v * / * * * + % V * % i I 1 i i Channel/IF Lower frame:bp phae Timerange: 00/19:20:25 to 00/19:25:20 ntenna: * Stoke: LL 4 -a
X a a g» 3 o 2 e-t- P* S" a. <S B- S et-» a- -2 Plot file verion 33 eated 06-FEB-199715:20:3 4C39.25 JOB5240.UVDT.1 Freq = 22.226 GHz, Bw = 16.000 MHz Bandpa table # 1 L. \ * * * v** v * / 4 t* OS w N o* o -6 4 NL 1 '! i -t «* ' 4 * 4 * \ \ f + \ 2 * * " ** * + * ^ ^ #.. 25 r<? 0 O o- ef- I P O Lower frame:bp phae * i i 'i 1 ' 'I Channel/IF Timerange: 00/19:20:25 to 00/19:25:20 ntenna: * Stoke: LL * i 5"
k a of to Plot file verion 34 eated 06-FEB-199715:20:3 4C39.25 JOB5240.UVDT.1 Freq = 22.22 GHz, Bw = 16.000 MHz Bandpa table t 1 - g &. t '» N a. t> H I S r> 2-2 -6-10 -14 Lower frame:bp phae Channel/IF Timerange: 00/19:20:25 to 00/19:25:20 ntenna: * Stoke: LL <yi
g 2 WQ3 * g * 5 3 Cn et- P" E O. $ fr R S" Plot file verion 35 eated 06-FEB-199715:20:39 4C39.25 JOB5240.UVDT.1 Freq= 22.226 GHz, Bw = 16.000 MHz Bandpa table # 1 > t» w N a. O O- P ft I p o 100 200 300 Channel/IF 400 500 Timerange: 00/19:20:25 to 00/19:25:20 ntenna: * Stoke: LL Lower frame:bp phae cn