NATIONAL RADIO ASTRONOMY OBSERVATORY

Transcription

1 NATIONAL RADIO ASTRONOMY OBSERVATORY September 18, 1964 Scientific Staff A. M. Shallo ay Subject: Format for Magnetic Tape at NRAO A standard for magnetic tape formats has been agreed upon. This memo describes the standardized portion of the format. The data portion of the format varies, depending upon the observation and type of equipment being used. The variable portion of the format for several present systems are described in the addendum of this memo.. As each new system is developed an additional addendum will be written to describe it. j.,_ahigiilatk.219:1k A. Observation: An observation consists of all the data (logical record) from one source recorded during one interval of time. B. SID i : Standard Identification Data One. A record consisting of 14 words (36 bits each) at the beginning of each observation. C. SID 2. Standard Identification Data Two. A record, similar end of an observation. os1d,, at the D. Character: One row on tape (perpendicular to direction ofmotion of tape). Consists of seven binary bits, one of which is odd parity on NRAO tapes. The bits are designated CBA8421 where!f i n generally is the least significant bit and C is parity. E. Word: A group of characters; six in the case of NRAO. F. Record: A group of words between inter-record gaps. G. MG: Inter-record gap, a physical spacing of records consisting of three or four blank characters, a longitudinal even parity character, followed by 3/4" of blank tape. H. Logical Record (Block): A group of records associated with one even the case of NRAO a logical record (block) is one observation. File: A group of logical records (blocks) generally one roll of tape.

2 DE FINITIQ IL.FORMAT:. ued J. EOF: End of File. An indication of the end of a file. Consists of 3 1/2 w of blank tape following the last longitudinal parity of an MG, followed by the number 15 (001111) with even parity, followed by an MG which treats the 15 as a record. The format for one observation ERG DATA IRG DATA SID records contain 14 ea. 36 bit words. IRG DATA record lengths depend on receiving system. ETC. III. SID FORMAT 3: DATA ERG SID 2 A. The format for an SID is as follows: Word Number W1 Fortran Code W2 Scan and Record Number W3 Observer's Code W4 Date W5 Time W6 Frequency W7 Declination W8 Right Ascension Telescope 1 SID W9 Hour Angle WiSY Declination W11 Right Ascension Telescope 2 W12 Hour Angle W13 Receiver Control Positions W14 Reserved for future use All SID words except Record Number are in BCD. Record Number is in iinary. However, all parity bits are odd throughout all NRAO tapes.

3 RMAT 3 Continued B. In detail the SM words are as follows where an X a 0 or I. WI Fortran Code I I I W2 - Scan and Record Number C B A I X I X X x x X o Scan (Observation) Number Record Number The 1 in the first character, column B, can be used to indicate that this is a SID record as opposed to a data record. The record number is an actual count of data records. It is zero in SID and its value in SID 2 is the same as in the last data record. Thus SID 2 contains a record number indicating how many data records there are in the observation. The SID 1 can be differentiated from SID 2 by the fact that the record number is zero in SID and other than zero in SW 2. 2 W3 - Observer's Code X X > X 0 0 1E- 103 X X X >

4 - SID FORM T 3 Continued W4 - Date C B A tens of years units of years-- x 0 0 tens of months- 0 units of months tens of days units of days W5 Time C B A S tens of hours x 0 0 units of hours X 0 0 tens of minutes- X 0 0 was of minutes- X 0 0 tens of seconds- X 0 0 units of seconds- S = 0 for sidereal time; s = 1 for solar time. W6 - Frequency X 0 0 4t X 0 0 ii."""1".""*". ' 106 S 4, L 1O4 xx [ b.} 11L IMMMOIWY <r MININIMO Frequency in cycles per second. W7 - Declination (Telescope 1) X NS 0 tens of degrees - X 0 0 units of degreesx 0 0 tens of minutes - X 0 0 units of minutesx 0 0 tens of seconds - X 0 0 units of seconds- NS indicates north = 0 or south

5 III. SE, FORMAT 3 (Continued) W8 Right Ascension (Telescope 1) C B X 0 0 units of hours Stens of tens of minutes lhours X 0 0 units of minutes- -tens of seconds units of secondstenths of 0 seconds W9 Hour Angie (Telescope 1) C B A EW units of hours -- X 0 0 tens of minutes X units of minutes- X 0 0 tens of seconds - X 0 0 units of seconds- X 0 0 tenths of seconds EW indicates east = 1 or west = O. w10 - Declination (Telescope 2) Same as W7. W11 - Right Ascension (Telescope 2) Same as W8. W12 - Hour Angle (Telescope 2) Same as W9. In the case of the interferometer, for example telescope is 85-1 and telescope 2 is WI3 Receiver Controls Position X 0 0 BCD Bit X 0 0 *BCD Bit B..). X 0 0 *BCD Bit C-0. X 0 0 *BCD Bit D* BCD Bit E-6, 4PBCD Bit F*

6 III SW FORMAT 3 Contipliesk One or more BCD bits above may be used for automatically indicating the position of a receiver control. The addendum describes which BCD bits are associated with a particular control in each receiver. W14 Reserved for Future Use. zero word. Each word in the SID is reserved for the specific item as indicated above. If a particular item is not required for a system (such as right ascension and hour angle on the 300 foot telescope) that word Is an all zero word. IV. SID FORMAT 2: Prior to September 16, 1964 the SID for all observations except the autoco elation receiver was different from that described above, as followss W2 - Scan and Record Number The 1 in the first character under column B was a O. W6 - Frequency The bit in character two, column A indicated the local oscillator being used in the receiving system. 0 = 1395 Mc 1 = 1390 Mc. W7 - Declination N and S were indicated in character 2 instead of character 1. Tape produced at the 300-foot scope used the code N = 10 and S = 01 for bits B and A in character 2. Tape produced at the 85-2 scope used the code N = XO and S = X1 for bits B and A in character 2. Tape recorded at the 85-2 scope contained in character 6 under columns BA a 10 to indicate 85-2 scope. W8 - Right Ascension Tape recorded at the 85-2 scope contained in character under columns BA a 10 to indicate 85-2 scope.

7 IV SID FOB MAT 2L,C... 1 ;ont1nued W9 - Hour Angle E and W were indicated in character 2 instead of character I. Tape produced at the 85-2 scope used the code E = 1 and W = 0 for bit A of character 2. Tape recorded at the 85-2 scope contained in character 6 under columns BA a 10 to indicate 85-2 scope. V. DATA: The data format is standardized only in the first two words. The remaining words are as dictated by the system and are described in the addendum. The data format is as follows: W1 - Fortran Code Same as WI. in Sm. W2 - Record Number X X The record number is an actual count of data records. It can have a maximum count of 12 binary bits. In the normal system where the SID and data are started automatically and synchronously with the clock, the record number can be used to indicate the exact time of reference of the data words. AMS

8 ADDENDUM 1 Se tember (To Memo on Format for Magnetic Tape at NRAO dated September ) AUTOCORRELATION RECEIVER as used Ma I to Se 1964: This is the only system which had an SID completely different than previously described. The system was used during the dates shown above and will never be used again. The words were 24 bits long instead of the usual 36 bits. SID FORMAT 1: W1 - Code C A X 0 Z <,103 _0 Z = 0 for SID i and Z = 1 for SID 2. W2 - Date X tens of years units of X years I tens of monthsunits of months I tens X of days I units of days W3 - Time --- tens of hours -±-- units X of hourss-t--- tens of minutes X units of minutes-4- tens of X seconds -I units of seconds W4 - Declination B A NS; N = 10, 5=01-> tens of degrees I units of degrees tens of minutes -. 4-units of minutes I tens of seconds

9 ADDENID UED W5 Frequency C A X -tens of megacycles -units of X megacycles- hundreds of kilocycles X -tens of kilocycles- -units of X kilocycles- hundreds of cycles - DATA: W Up Count C B A 8 4 X Y X X X W2 - Reversible Counter C B A 84 X X X X Y W3 - Channel 1 Counter C B A 84 X X X X Words W4 through W102 are the same as W3 and contain charmel counters 2 through 100. Y = 0 for noise tube off and Y = 1 for noise tube on. The 1 following the Y in W1 differentiates data from SID.

10 10- ADDENDUM 2 Se p tember (To Memo On Format for Magnetic Tape at NRAO dated September ) AUTOCORRELATION RECEIVER as used a_ j b e r SID contains information in words WI, W2 W3 W4, W5 W6 W7 (W8 and W9 when used on a two axis antenna) and W13. WI3 Receiver Controls Positions BCD Bit A indicates position of operate-calibrate switch: = operate (signal switched) = calibrate (comparison non-switched). BCD Bit B indicates bandwidth in uses 1 =2.5 Mc 2 = 625 ke, 3 = 250 kc = 62.5kc. DATA: W1 - Fortran Code As described in above referenced memo. W2 - Record Number As described in above referenced memo. W3 - Up Counter X Y X X X W4 - Reversible Counter C B A 8 4 y I X

11 - 11 W5 Channel 1 Counter C B A 8 4 Y I X X Words W6 through W104 are the same as W5 and contain charm]. counters 2 through 100. Y 0 for noise tube off and Y = 1 for noise tube on. Arthur M. Shalloway

12 ADDENDUM 3 September 18, 1964 Magnetic Tape at NRAO dated September I 1964) SID contains information in words Wl, W2 W4, W5, W6, W7 8 and W9 when used on a two axis antenna) d W13. DATA: W13 Receiver Controls Positions BCD Bit A indicates which fixed local oscillator is in Use, 0 = 1395 Mc and = 1390 Mc. W1- Fortran Code As described in above referenced memo. W2 - Record Number As described in above referenced memo. W3 - Channels 1 through 3 Outputs X Sign Y Z X X Sign Y Z X X Sign Y Z X Channel 1 Channel 2 Channel 3 Sign is 0 =+ and 1 = = 0 for noise tube off and Y = 1 for noise tube on. = 0 for normal data and Z = I for overflow data. Seven more words contain channels 4 through 23 and a battery (reference) channel in the same format as above.

13 ADDENDUM 4 September (To Memo on Format for Magnetic Tape at NRAO dated September 8, 1964) OCCULTATION RECEIVER. SID Occultation tapes after September 16, 1964 will use standard SID Format 3. Tapes before this time used SID Format 2. Very early tapes were made without a DIT. These very early tapes contained garbled data in the SID locations. Future tapes will contain information in SID locations: We, W 3, 0 We W 51 W 7 W8 and W 9. The date (W 4 ) is not available at the present time. DATA: - Fortran Code As described in above referenced memo. W2 - Record Number As described in above referenced memo. W3 - Receiver Outputs C B A 8 4 y X X X X X i 261, Receiver 1 Receiver 2 Receiver 3 W4 - Time C B y units of hours X -tens of hours - tens of minutes X 0 0 units of minutes- X 0 0 -tens of seconds- X 0 0 -units of seconds- X 0 0 -tenths of seconds- - 0 for noise tube off. Y or oise tube on.

14 DpEsi_DtalArpN$IME121. The format for one record is as follows: W1 -- Fortran Code W2 - Record Number W3 - Receiver Outputs W4 - Time W3 - Receiver Outputs W4 - Time etc. W4 - Time W3 - Receiver Outputs IRG W3 - Receiver Outputs -- words are recorded at the beginning of each odd tenth of a second, e.g seconds. W4 - Time -- words are recorded at the beginning of each even tenth of a second e. g., 20.2 seconds. A record, including IRG, is 20 seconds long. The IRG causes the system to miss one time word, so that each record starts with W3 - Receiver Outputs at the following three times in each minute: 00.1 second, 20.1 seconds, and 40.1 seconds. The first record after SID occurs asynchronously with respect to time; therefore, it could have a W4 Time word following directly after the W2 - Record Number. This would require a test of the number of words (odd or even) to determine which is first (W3 or W4); thus it is recommended that the first record after SID be ignored. The next MG will occur at the proper time to make the record appear as indicated above e. in the order W1, W2, W3, W4, etc. A few tapes made during June, July and August of 1964 were made with an old magnetic recorder that required a longer time to record the ERG' s. In addition to the W4 - Time word that is lost during IRG's, one W3 - Receiver Outputs word was also lost. These tapes can be detected by counting the words per record. Claude C. Bare

15 15 - ADDENDUM 5 _16±. (To Memo on Format for Magnetic Tape at NRAO dated September ) INTERFEROMETER RECEIVER SID. The interferometer SID format is the same as described in Addendum 4 dated September , in which the occultation receivers format is described. DATA: The interferometer data format is the same as described in Adden uxn with the following exceptions: 1. An ERG occurs every minute instead of every 20 seconds. 2. The 12 bits associated with receiver I data in occultation are used to record fringe data for the even tenths of a second. Those bits associated with receiver 2 record fringe data for the odd tenths of a second. Receiver 3 bits are not used and are zero. Thus, in this arrangement, receiver 2 i 1ata location represents the fringe measurement for the first j second and ieceiver I represents the fringe measurement for the second ro- second. 3. Y, the calibration bit, is always zero.

16 ADDENDUM 6 (Revised) March (To Memo on "Format for Magnetic Tape at NRAO" dated September ) INTERFEROMETER RECEIVER after January 1965: SID FORMAT: DATA: Standard SID format 3 with telescope 1 referring to 85-1 will be used. In earlier interferometer data the 85-2 was recorded in telescope 1 location in the format data will be added later in words W 1 9 W 8 and W The beginning time in the SID W will always be 50 or seconds. Each data record will be 601 words. Word 1 Fortran code Word 2 Record Number: The record number added to the hours and minutes specified in W5 of the beginning SID indicates the beginning time of the record. Words 3 through 601-Data: One word is recorded e..?ch 1/10 second. The word representing the last 1/10 second of the minute is omitted. B A Parity X 0 0 X Parity X X X X Parity Y 0 YY Parity Y Y 0 Y Y Parity Z 0Z 0 Z Parity Z Z Z Z X = fringe output data. Adjust as follows: No signal decimal Max fringe = 768 decimal Min fringe = 256 decimal X o = usually zero octal 1400 octal 0400 oct

17 2 Words 3 through 60 1-Data (continued): Y = AGC output data. Data recorded during even seconds represents 85-2 AGC and data recorded during odd seconds represents 85-1 AGC data. Z = Delay number. The most significant Z o is a one when the 10 db attenuator is in use. The least significant Zo is not used at the present time. The delay number is recorded in the remaining 10 bits. Two readouts indicating the same delay should be detected before a new delay number is assumed to be correct. The first integration of data actually begins 10 milliseconds before the minute and ends 90 milliseconds after the minute. The centers of integr tion intervals occur at the minute milliseconds N x 100 milliseconds, where N is the number of readouts after the inter-record gap. Any additional delay occurring in the analog portion of the system must be added to the 40 milliseconds. Claude C. Bare

18 ADDENDUM 8 May 8, 1965 (This addendum replaces Addendum 8 dated February 11, to Memo on "Format for M. :netic Tape at NRAO" dated September 18, 1964.) 20-CHANNEL MULTIFILTER RECEIVER as was used durin February March 1965 Addendum 3 covers the basic 20-channel tape format. Word 14 is all zeros. The 24 data channels were used to record the following data: Channel (more pos = true) 8 9 if 10 It 11 I? 12 VT t? V? 20 if 21 I? 22 V? Data Receiver Ground Receiver (70 off north to west) Ground 750 Receiver Ground 20 - declination 10 - declination 8 - declination 4 - declination 2 - declination 1 - declination 40' - declination 20' - declination 10' - declination 8' - declination 4' - declination 2' - declination - declination 40" - declination 20" - declination 10" - declination Ground Calibration voltage (BAT) Timing: The system can sample the RC (the length of the RC is not recorded) at 2 second of 10 second intervals. The first sample in the 10 second mode occurs at 10 seconds after the start time (T2) and the last sample occurs at the stop time (T3).

19 2 Timing (continue If the 2 second mode is used, the number of samples should be a multiple of 5, since the start and stop times are specified in 10 second increments. The 2 or 10 second mode is determined by counting the readouts between T2 and T3. The first sample in the 2 second mode should be at 2 seconds after the beginning time (T2) and the last sample occurs at (T3). A logic error caused the first sample to be missing in the two second sample mode until the error was corrected. The error was corrected in the late afternoon of March 21, LST before tape number 279 was begun. NOTE: Before March 2 1, the first sample was omitted when operating in the two second sample mode. Claude C. Bare

20 ADDENDUM 9 March 25, (To Memo on Format for Magnetic Tape at NRAO" dated September 18, 1964) 140' TELESCOPE OUTPUT SYSTEM SID contains information in words Wl, W2, W3, W4, W5 and W13. DATA: W13 Receiver Controls Positions BCD bits A, B and C is a BCD number indicating the number of integrations between IRG (interrecord gaps). A is the most significant BCD bit. BCD bits D and E indicate the channels in use. Channels axe indicated as follows: El Channel 1 E2 Channel 2 E4 Channel 3 E8 Channel 4 D1 Channel 5 D2 Channel 6 D4 Channel 7 D8 Channel 8 A one in a channel position indicates the channel is being used. A zero indicates the channel is not being used. BCD bit F is a BCD number indicating the integration and sampling period as follows: 1 = 0.2 sec 2= 1 sec 3= 2 sec 4= 5 sec 5= 10 sec 6= 20 sec 7= 60 sec 8 = 120 sec W1 - Fortran Code As described in above referenced memo. W2 - Record Number As described in above referenced memo.

21 2 ADDENDUM 9 (CONTINUED) DATA (continued): W3, 4, 5, etc. - Channel Data X X J > }, A/D converter output Integration Counter Number (3 --BCD _bits) Channel indicator See note below. A/D converter output is the digital data for the particular channel. It is a 12-bit twos complement number. Integration Counter Number is a 3 bit BCD number indicating how many integration (and sampling) periods have passed when the particular data is recorded. Y is a one if the noise tube was on for that particular integration, and is a zero if the noise tube was off. Channel Indicator is a binary number from one to eight indicating which channel the data in that particular word refers to. Arthur M. Shalloway

22 ADDENDUM 10 May 26, (To Memo on Format for Magnetic Tape at NRAO dated September 18, 1964) OCCULTATION RECEIVER as will be used on the 300-foot scope during June-4y 1965: SID: Occultation tapes will use standard SID Format 3. Tapes will contain information in SID locations: W2, W3, W4, W5, W7. DATA: W1 - Fortran Code As described in above referenced memo. W2 - Record Number As described in above referenced memo. W3 - Receiver Outputs X y ' X f X ', X i X X ) Receiver 1 Receiver 2 Receiver 3 W4 - Time-Position X y 0 units of degrees X 0 0 tens of minutes of arc X 0 0 units of minutes of arc- X 0 0 tens of seconds of arc X 0 0 units of seconds X 0 0 tenths of seconds Position Time Y = 0 for noise tube off. Y = 1 for noise tube on. Continued --

23 The format for one record is as follows: W1 - Fortran Code W2 - Record Number W3 - Receiver Outputs W4 - Time-Position W3 - Receiver Outputs W4 - Time-Position ADDENDUM 10 (CONTINUED) to etc. CAL:- W4 - Time-Position W3 - Receiver Outputs IRG W3 - Receiver Outputs -- words are recorded at the beginning of each odd tenth of a second, e. g., 20.1 seconds. W4 - Time-Position -- words are recorded at the beginning of each even tenth of a second, e. go, 20.2 seconds. A record, including IRG, is 30 seconds long. The IRG causes the system to miss one time word, so that each record starts with W3 - Receiver Outputs at the following two times in each minute: 00.1 second and 30.1 seconds. The first record after SID occurs asynchronously with respect to time; therefore, it could have a W4 - Time-Position word following directly after the W2 - Record Number. This would require a test of the number of words (odd or even) to determine which is first (W3 or W4); thus it is recommended that the first record after SID i be ignored. The next IRG will occur at the proper time to make the next record appear as indicated above, 1. e., in the order W1, W2, W3, W4, etc. A starting time of 20 to 27 or 50 to 57 seconds is recommended. A continuously switched cal is planned. The rate will be 1Hertz. The cal will be on during odd seconds and off during even seconds. The cal will follow the BCD n one second" bit of the NRAO sidereal clock. The cal on condition will be indicated by a true bit in the most significant bit location of the data and time-position words. Claude C. Bare