Draft Proposal for VLBI Standard Interface Specification
from Jpanese Group

Comments on "Draft Proposal for VLBI Standard Interface Specification, February 3, 1999" and Results of Japanese VSI Discussions

by
Noriyuki Kawaguchi
National Astronomical Observatory

and

CRL Technology Development Center
(Tetsuro Kondo, Hitoshi Kiuchi, Junichi Nakajima,
Yasuhiro Koyama, Mamoru Sekido and etc.)
Communications Research Laboratory

June 15, 1999: Draft release
June 17, 1999: "Summary" and "Appendix-A Abbreviations" were added.)
June 18, 1999: Error correction in Table A2-2 (tf at 64 MHz from 5 to 3))


Summary

Bottom lines of our proposal are as follows. We define signal connection at a different bit rate as,

Table. Signal connection at a different bit rate
Aggregated bit rate Number of Connectors Clock Rate Logical Number of Wires
64 Mbps 1 4 MHz 16
128 Mbps 1 8 MHz 16
256 Mbps 1 16 MHz 16
512 Mbps 1 32 MHz 16
1024 Mbps 2 32 MHz 32
2048 Mbps 2 64 MHz 32

And we introduce a time code (IRIG-B) for VSI. We also describe the software implementation and illustrate the expected views of the system configurations which will be seen in the various stages of the VLBI operations with the use of VLBI Standard Interface. Investigations of VSI signal definition and VSI connector pin assignment and cable wiring are also introduced.
Finaly, our conclusions for VSI are as follows.
(1) A standard VLBI recorder interface is urgently required to promote future world wide corporations in a VLBI observation and to encourage new developments for more advanced VLBI data acquisition system.
(2) The clear and transparent interface makes easy to operate different types of recorder system in parallel at an observatory and at a processing site. Easy translation also becomes possible.
(3) Physical specifications for input and output data, timing signals and time codes are important
(4) Realization of the VSI on different types of recorders is strongly requested.
(5) The VSI specifications do not exclude any ADDITVE functions but only defines MINIMUM functions.
(6) In the VSI realization, any internal design is acceptable.
(7) A standard software control protocol, a communication line and messages, should be defined in the final VSI documentation.


TABLE OF CONTENTS


Comments on "Draft Proposal for VLBI Standard Interface Specification, February 3, 1999"

by
Noriyuki Kawaguchi
National Astronomioal Observatory

(June 14, 1999)

General Comments

A VLBI standard interface has long been expected by many VLBI astronomers and geodesists who hope to use radio telescopes in foreign observatories being equipped with different types of recording system. It is very welcome for us to have the standard interface like as a proposal given in a document titled "Draft Proposal for VLBI Standard Interface Specification, February 3, 1999". In this short note we would like to make some technical comments on the draft. Also we would like to point out a fact that the standard interface does not give a perfect solution but leaves media incompatibility behind. We still need to use a "translator" or a "hybrid correlator". We should aware the importance of the media translation now partly performed in Mitaka VSOP correlator, National Astronomical Observatory, Japan. Easy translation should be taken into consideration on the standard recorder interface.

Scopes of targets

In considering the standard interface, following goals shall be achieved;
(1) Easy translation of a tape media The DOB (Digital Output Box) output shall be connected to the DIB (Digital Input Box) in the tape translation, not only to a correlator. The output connector of the DOB shall be mated to the input connector of the DIB.
(2) Tandem operation of two or more DTS(Data Transmission System)s Two or more DTSs will be involved not only in correlation but also in data acquisition. Wide band data recording at a rate beyond 1 Gbps would be made on two or more recorders. Synchronized operation of the DTS to others is a matter of concern as much in the parallel recording as in correlation processing.
(3) Real time correlation Real time correlation would be a future important function of a reliable long-baseline interferometer, in which the data on a tape is always checked for the auto-correlation spectrum, cross correlated for monitoring the phase stability while the recording is carrying on. Single operation of DIB at an observatory, (and of DOB at a correlator) will be less important in future, though it would keep economical benefit on the recording (or playing back) cost. My preference is not on the separation of DIB and DOB boxes.
(4) Sub network correlation Future operation of a VLBI network connecting a number of telescopes over the world will be made in sub networking. In the correlation, dynamical resetting and offsetting of ROT (Requested Observe Time) might be important.
(5) Unified software support A complicated software control might cause difficulty in the operation of different types of DTS. The DTS shall be controlled under the FS9 field system following the VEX schedule file.

The expected views of the system configurations which will be seen in the various stages of the VLBI operations with the use of VLBI Standard Interface (VSI) are illustrated in Appendix-1.

Basic DTS Configuration

Functionally or Physically the DTS is divided into two boxes as indicated in Figure 1. Usually the DTS of a "Data Recorder" has both functions of DIB and DOB, but by some economical reason it might be a good idea to separate the DTS to the DIB and the DOB for the use in VLBl data transport system.

The DIB and DOB is just the same ones as described in the draft proposal VSI specification except for adding timing signals I/O on both boxes. Even in case that the DTS is divided into the DIB and DOB, the both input and output functions of the timing signals is useful for various applications to be noted later.


Figure 1. Basic Configuration of the DTS (Data Transmission System)

Parallel recording of a high rate data

A high rate data transmission beyond 1-Gbps from an observatory site to a correlation center is a strong demand of a VLBI astronomer in their high sensitive observations. Parallel recording is one possible way to allow for such high rate data transmission. Figure 2 shows a connection of two or more DIBs. Timing signals from a master clock at an observatory is supplied to all DIBs with no time offset and multi-stream data from a data acquisition system are connected to each DIB in parallel. The same timing signals are probably used in a data acquisition terminal.

In this application synchronization of two or more DIBs is performed with a similar way commonly done in correlation processing.


Figure 2. Parallel recording of a high bit-rate data on two or more DIBs.

Synchronized reproducing in a correlation center

At a time of correlation processings, the data at a specific time on tapes is reproduced synchronously from two or more DOBs. Timing signals are successively to each DOBs. The signals are also transferred to a correlator for the use in synchronizing correlation parameters and the data reproduced.


Figure 3. Synchronized data reproducing at a correlation site.

Media Conversion

Media conversion is one possible way of cross correlating tapes of different type. If VSI standard is realized, the work can be done easily in a straight forward manner as shown in Figure 4. By manual operation or at the minimum software support, the media conversion or tape translation in other words can be done with simple connection of DOB and DIB.


Figure 4. Media Conversion

Time Code Transfer

Two different time codes are currently used in different types of VLBI recording systems, the Mark IV and the VLBA system. One is based on the UTC time system such as year, day in the year, hour, minute and second. Only one digit is assigned for the last digit of the year. The precision of the second is 1 milliseconds. Another is based on the MJD time system of Modified Julian Date followed by an integral number of seconds in a precision 0f 0.1 milliseconds. The Japanese K4 system uses UTC time codes similar to the MK-IV system except for the year codes, two digits of the last two year codes are assigned and the precision of the second is 0.1 second. In all these system the time codes are inserted in or replaced with observed scientific data. In the S2 system, the time codes are written on a magnetic tape in a manner completely independent from the scientific data. The time code of the UTC base can be set and recognized by computer communications. The variety of the time codes makes difficult to keep compatibility between different types of recording system. Here an unified time code and the time transfer system are proposed.

The IRIG-B code is widely used in a time transfer system in the world. The precision of one second is enough in presumable applications to VLBI observations and correlation processings. Higher precision to the sub millisecond will easily be obtained by counting a data clock, if necessary. The second epoch is defined by a start of the coding sequence but not enough in the accuracy. The combination of an 1-pps signal and the IRIG-B time code is proposed. In an original draft of the VSI proposal the time set is assumed to be done by a computer communication. But in this case, interchangeability of a control software is difficult to provide. The combination of the 1-pps signal and the IRIG-B code makes possible to set a time automatically without any supports by a control computer. An example of the realization of such a time code transfer system is shown in Figure 5. From a clock and an 1-pps signal, an internal clock is tuned to a master clock outside, then a time code is recognized by decoding an IRIG-B code. Internal timings and arbitrary time codes on a tape are generated in a digital data processor in a DIB. The time code on the tape is reproduced at a specific epoch of a clock, then the 1-pps signal, data are generated in a processor in a DOB. By referencing the time code on a tape the IRIG-B code is generated in the encoder.

"Year" is not defined in the IRIG-B code. It is, however, not a difficult work to generate any time code including the "year" on a tape of any type by referencing to time description written on a schedule file or a log file at a time of observation and correlation processing.

Any time codes are permissible on a tape. The realization is an internal matter of a DTS. Both insertion and replacement of a time code in a data stream is acceptable.


Figure 5. An example of a time transfer system to be realized in Japanese system

Data Transmission

Bit stream rates on a fixed number of logical wires should not exceed 64 Mbits/sec. Reliable cable connection limits a physical number of pins within 50, and a physical number of data connections within 32. For simplicity in the connection, it is better to limit a number of connectors within 2. The aggregate bit rate thus shall be less than 2048 Mbits/sec. The data transmission in much higher bit rates with metal wires is NOT realistic. Optical transmission should be considered. The bit rate less than 2048-Mbits/sec transmission will be technically feasible in the next 10 years with a single pair of the DIB and DOB. By considering above limitations, the data transmission listed in the Table 1 is reasonable on the current technical conditions.

Table 1. Signal connection at a different bit rate
Aggregated bit rate Number of Connectors Clock Rate Logical Number of Wires
64 Mbps 1 4 MHz 16
128 Mbps 1 8 MHz 16
256 Mbps 1 16 MHz 16
512 Mbps 1 32 MHz 16
1024 Mbps 2 32 MHz 32
2048 Mbps 2 64 MHz 32

Regarding more details about signal definitions and connectors, see Appendix-2 and Appendix-3.

Software Implementation

If hardware system could be realized as those mentioned above, very few software supports are required. The minimum required software controls are listed in Table 2.

Table 2. Software Commands
Operation Command Code Comments
Recording Start Command REC Record
Playback Start Command PLAY Play back
Synchronous Record SYNR Record synchronously
Synchronous Playback SYNP Play back synchronously
Stop Command STOP Recorder stops and waits for a command.
Bit Rate Specification BR=nnnn Unit: Mbps
Examples
"nnnn=0064" : 64 Mbps
"nnnn=0128": 128 Mbps
"nnnn=0256": 256 Mbps
"nnnn=0512": 512 Mbps
"nnnn=1024": 1024 Mbps
"nnnn=2048": 2048 Mbps
Fast Forward Command FF Go to a tape end
Fast Rewind Command RW Go to a tape head
Tape Loading LOAD Load a tape
Tape Unloading UNLD Unload a tape

The control software can sense the recorder status on the items listed in Table 3.

Table 3. Recorder staus
Items of status Responce Code(*)Comments
Transport Status
OPR=REC
=PLAY
=SYNR
=SYNP
=FF
=RW
=STOP
In recording
In playing
In synchronously recording
In synchronously playing
In fast forwarding
In fast rewinding
Stop and waiting for command
Bit Rate Sensing BR=nnnn Unit: Mbps
Examples
"nnnn=0064" : 64 Mbps
"nnnn=0128": 128 Mbps
"nnnn=0256": 256 Mbps
"nnnn=0512": 512 Mbps
"nnnn=1024": 1024 Mbps
"nnnn=2048": 2048 Mbps
Tape Staus
TAPE=LOAD
=UNLD
=EMPTY
Tape is mounted and ready to start.
Tape is mounted but not ready.
Tape is not mounted.
Tape Position
TPOS=BOT
=EOT
Tape is at the beginning.
Tape is at the end.
Tape Remains BEST=nn Percent of tape rest
Data Status
DATA=VALID
=INVALID
Data is available.
Data is not available.
(*) For query, a character "=" is replaced with a character "?" followed by no status codes.

Messages listed in the Table 2 and Table 3 are minimum requested codes in a software communication. Other special commands and status items can be defined for each recording system for their own convenience. Regarding an actual communication link to be used, e.g., Ethernet, RS-232C, IEEE-488, USB, etc., and the details should be defined later.

Conclusions

(1) A standard VLBI recorder interface is urgently required to promote future world wide corporations in a VLBI observation and to encourage new developments for more advanced VLBI data acquisition system.
(2) The clear and transparent interface makes easy to operate different types of recorder system in parallel at an observatory and at a processing site. Easy translation also becomes possible.
(3) Physical specifications for input and output data, timing signals and time codes are important
(4) Realization of the VSI on different types of recorders is strongly requested.
(5) The VSI specifications do not exclude any ADDITVE functions but only defines MINIMUM functions.
(6) In the VSI realization, any internal design is acceptable.
(7) A standard software control protocol, a communication line and messages, should be defined in the final VSI documentation.


Appendix-1. Illustrative prospects with the VLBI Standard Interface

This appendix is intended to illustrate the expected views of the system configurations which will be seen in the various stages of the VLBI operations with the use of VLBI Standard Interface (VSI).

A1-1. Observations

With the use of the VSI, various types of the data recorders can be used with any formatter/interface which samples the analog data to be recorded on a tape by the data recorders. In this way, an operator at an observation site can choose a data recorder according to the requirements of the VLBI session to be performed.


Figure A1-1. Three different configurations at an observation site expected to become possible with the developments of the VSI.

A1-2. Tape Copy (Dubbing)

With the use of the VSI interfaces, VLBI observation data recorded on any media can be copied to different tape media for dubbing operation as shown in the Figure A2.


Figure A1-2. The expected configurations of the system with which observation tapes are copied to the other tapes in a different tape media using the VSI.

A1-3. Correlator

At a correlator station, multiple data recorder systems should become possible to coexist with the use of the Output Interfaces of the VSI. The correlation processing will be started by first setting the time of the wall clock system. After each data output box is commanded to synchronize the data to the time code generated by the wall clock system, the wall clock is commanded to increment. Then each recorder is controlled by the data output box and synchronized to the time code. The correlator system correlate the observation data and the results will be written to output files with the auxiliary data including the flags which indicate whether the data are valid or not. When the data is synchronized to the time code from the wall clock, the flag will indicate the results are valid and vise-versa. In an actual implementation, much more complicated procedures will be necessary to position each tape position and such a detail must be considered in the course of designing the correlator system.


Figure A1-3. An example of the system configuration for a correlator using VSI.


Appendix-2. VSI ECL-signal definition (revised, June'99)

A2-1. Target

We will define the paralleled wire and the data clock rate combination which is used in the VSI data transfer. The VSI data rate will vary from 64,128,256,1024 to 2048Mbps(maximum). These are due to connected instruments or observation mode. With the document which defined the cable and connector, real VLBI instrument under multi-vendor environment are connected electrically. Sampled channel allocation in this digital stream is not included for further discussion.

A2-2. Proposal

Table A2-1. VSI data rate and connector, parallel line expansion (essentially same as Table 1 in the Kawaguchi's draft)
Data rate
(Mbps)
Data Clock
Frequency
(MHz)
Number
of
Wire
Connector
primary
Connector
Secondary
64 4 16 DATA0-DATA15 Not used
128 8 16 DATA0-DATA15 Not used
256 16 16 DATA0-DATA15 Not used
512 32 16 DATA0-DATA15 Not used
1024 32 32 DATA0-DATA15 DATA16-DATA31
2048 64 32 DATA0-DATA15 DATA16-DATA31

Proposed VSI data rate and parallel data transfer is presented in Table1. The bus width is varying two steps 16 and 32 wire. Within the width, data clock is selected to accomplish its total data rate. In Japanese case our K-4 system is 8 wired system up to 256 Mbps. For the VSI fruit, we will make an effort to change it. As for 1024 Mbps mode, the data bus width is doubled. Finally, 64 MHz clock is used to achieve 2048Mbps. We hope other VSI group will also take this simple idea.

A2-3. Electrical definition


Figure A2-1. VSI input signal chracteristic definition.

The VSI electrical characteristics is mainly based on differential- ECL connection rules. Although other low voltage differential line drivers are appeared recently, Still the ECL is very popular among our VLBI engineer, technician and manufactures around. World-wide parts availability is another important reason to choose this. Among several ECL definition, The "ECL10K" is recommended. The "ECL10KH" is also recommended in high speed 64MHz operationfor tr and tf improvement. The timing specification of the VSI is summarized Table 2. Our basic rule to the multiple connector expansion is as follows. The clock in a connector will define a timing to determine the data in the connector. This means the primary connector clock is not used for the secondary connector vise versa. The other lines are referred from the primary connector. We have three special lines attribute the clock and data. VALID is the line which tell a DOB data validity to latter stage. This line must activate at the moment of a recorder DOB synchronize. Any DOB asked to use this line to indicate its data status. 1PPS in ECL determine precise data relation to 1-PPS. The 1PPS width tw_1pps is a Hydrogen-maser dependent. Usually this is several micro sec to hundreds micro-sec duration and we use the rising point only. Since the 1-PPS can become signal to check the system directly, conventional 1PPS using a BNC connector output is recommended in DOB. IRIG-B is an idea introduced by Kawaguchi. The IRIG-B is the only TTL signal in this VSI electrical definition. The IRIG-B is used to mark the 1-PPS tick with certain UTC. The IRIG-B is used to label each 1-PPS and issued at coarse timing itself. This serial transfer is very slow. Thus it is possible to remain it in TTL besides high speed ECL. The IRIG-B marker delay to the 1PPS will be around 2-ms. (Please refer our main draft proposal.)

Table A2-2. Electrical time characteristics definition
CLOCK
(MHz)
tc(ns) td(ns)
transmit side
Tw(ns)tr(ns)
20-80%
tf(ns)
80-20%
4 250 125 125+/-3 5 5
8 125 62.5 62.5+/-3 5 5
16 62.5 31.25 31.25+/-3 5 5
32 31.25 15.625 15.625+/-3 5 5
64 15.625 7.8125 7.8125+/-3 3 3

A2-4. Electrical measurement

The VSI signal quality defined in previous section should be monitored at the VSI cable with appropriate terminator in front of a DIB or a correlator as in Figure A2-2. Detailed measurement chart which is used by engineer or technician using oscilloscope on site should be prepared to identify problem on the way to successful connection.


Figure A2-2. VSI input signal chracteristic measurement.

A2-5. Standard ECL circuit recommendations (Kiuchi et al.)

The VSI -TDC should show an example of ECL driver and differential receiver which qualify the characteristics in Table A2-2. This will useful for the future design and new VLBI developing countries. The Figure A2-3 shows an example of driver / receiver combination proposed by Kiuchi. We may knows the parameters, but Ideally the resistors Ra and Rt should be determined after the VSI connector / cable selection with our experiment. They will be adjusted to the twisted pair line impedance around 100 ohm. This will minimize the reflection and enable long distance transmission. Several capacitor and diodes not shown in fig.3 also increase its performance. These interface part tend to occupy large area on printed board. Kiuchi recommend to use SMT(Surface Mount Parts). It is thought the protection diode is important.


Figure A2-2. ECL line driver and line receiver.

A2-6. Additional remarks

Current Japanese 1024 Mbps VLBI is accomplished under stable 32 MHz by 32 paralleled line. We see less problem in this clock frequency. 64 MHz clock will be the key to achieve 2048 Mbps VLBI. But as you see in Figure A2-2, the VSI input parameters becomes very severe in this frequency. Especially, slight timing difference occurred between the data and clock will bring vast increase of error rate and faulty data synchronize. The TDC experiment using actual VSI connector and VSI cable should be carried out before final determination of VSI Electrical definition.


Appendix-3. VSI connector pin assignment and cable wiring (revised, June'99)

A3-1. Target

We define the VSI connector and associated metal cable assembly which enable to handle 64MHz data with other fundamental data attribution lines. Either in a telescope site or in a correlation site, digital signals are transferred maximum length of 20 meters between DAS, DIB, DOB and correlator.

A3-2. Concept

Table A3-1. VSI required data line
Name Number Purpose
CLK 1 Data clock
DATA 0-31 32 Data
VALID 1 inform data validity
1PPS 1 1PPS tick
IRIG-B 1 Separate coarse time code
Total 37 …need more than 74 metal lines

Currently the required lines for the VSI data transfer based on Whitney and Japanese group documents are summarized in Table A3-1. Since cable / connectors more than 50 wire is not common, and even in our VSI application VLBI mode below 256 Mbps will remain certain period. To implement the idea to the cable and connectors available, a compromise with multiple cable and connector is our choice. The concept which uses two cable in high speed data rate are reasonable. The cable is 20-set of twisted pair cable assembly. On the other hand there is no compromise with cable electrical structure. Due to the cross talk and external effect, 64 MHz clock data transfer without the appropriate shield will be a troublesome part. For the high speed data transfer, each twisted pair are shielded as in Figure A3-1. In Table A3-2, reduced lines to a connector is shown. When we separate data to connectors, data clock should supplied separately to eliminate cable uncertainty among the cables.


Figure A3-1. Cable assembly.

Table A3-2. VSI data line into a 50-pin D-sub connector
Name Number Level Purpose
CLK 1 ECL Clock rise timing determine data valid point In each connector
DATA 0-15
or
DATA 16-31
16ECL Data separated 16 lines each. Low speed data need only one connector.
VALID 1 ECL Flag which used to check stream data validity to latter stage. For an example correlator will use this flag as a DOB synchronization. Recorder DOB will use this flag as servo lock.
1PPS 1 ECL Tick rise point define a certain data point k to station H-UTC. (From a view of In/Output transparency, An additional 1PPS output using BNC connectors TTL level is recommended in case of DOB.)
IRIG-B 1 TTL Coarse time code give the 1 PPS tick to certain H-UTC. Please refer Japnese group draft poroposal.
Total20 40-pins are used for signal. Remaining 10 are used for shielded grounds


Figure A3-2. D-sub 50 female panel adapter.

Table A3-3. VSI pin assignment to Dsub-50
Pin name Pin name Pin name
1 CLK+ 18 GND 34 CLK+
2 DATA0(16)+ 19 GND 35 DATA0(16)-
3 DATA1(17)+ 20 VALID+ 36 DATA1(17)-
4 DATA2(18)+ 21 VALID- 37 DATA2(18)-
5 DATA3(19)+ 22 RESERVED 38 DATA3(19)-
6 DATA4(20)+ 23 RESERVED 39 DATA4(20)-
7 DATA5(21)+ 24 GND 40 DATA5(21)-
8 DATA6(22)+ 25 GND 41 DATA6(22)-
9 DATA7(23)+ 26 GND 42 DATA7(23)-
10 DATA8(24)+ 27 GND 43 DATA8(24)-
11 DATA9(25)+ 28 IRIG-B 44 DATA9(25)-
12 DATA10(26)+ 29 GND(IRIG) 45 DATA10(26)-
13 DATA11(27)+ 30 1PPS+ 46 DATA11(27)-
14 DATA12(28)+ 31 1PPS- 47 DATA12(28)-
15 DATA13(29)+ 32 GND 48 DATA13(29)-
16 DATA14(30)+ 33 GND 49 DATA14(30)-
17 DATA15(31)+

50 DATA15(31)-
*() for secondary connector name

The pin assignment to the D-sub 50 pin connectors is presented in Table-A3-3. The pin assignment is already used by Philips and Panasonic in their digital equipment. This means the connector assembly with the cable and panel-adapter mount to a printed circuit board is established. Only the CLOCK, DATA_n, VALID, 1PPS and IRIG-B are commonly used signal in the VSI. At this moment one reserved line are remained for furthur requirement. But other data line depends particular DAS, DIB, DOB or correlator hardware are not allowed to appear in the VSI connection. Using carefully shielded cables and enough ground pin remaining in the 50pin, the multi-vendor environment connection will become stable and successful. Currently experiment using the cables are prepared at CRL.

The D-sub connector shell should use ISO bolts and nuts for mechanical compatibility.

A3-3. Additional comments on cables

Figure A3-3. Flat cable.Figure A3-4. ID1 cableFigure A3-5. ID1 & VSI

VLBI data transfer had been using differential ECL signals and twisted pair cables. Figure A3-3 shows flat cable with and without shield. Figure A3-4 shows D-sub25 connector and cable used in the Japanese K-4 VLBI. The assembled cable is also consist from twisted pairs and an assembly shield. These cables are enough performance below 32 MHz data clock rate. But above the 32 MHz and multi vendor environment expected in the VSI, exact ECL signal shape must be maintained under higher clock. The IVS-TDC CRL group experienced parallel 128 MHz ECL data transfer will cause terrible data degrade with slight cable difference. We strongly propose to employ twisted pairs with shielded each for 64 MHz expansion. Although this will increase the weight of cables, the de-factory standard cable for the D-sub 50 like in Figure A3-5(left) exists, the assembly employed AWG27 equivalent conductor is diameter of around 15 mm and reasonable price. If a special cable is designed for VSI, it will drastically increase its cost and they are difficult to obtain replacement one.

As for a connector, The connector shell and connector latch mechanizm expected be have enough stiffness to support the cable weight. The D-sub with screw bolt is preferred. VLBI equipment are connected and disconnected often on site. Also it important the on site engineer can inspect signals directly. Others new connectors are too small to handle and often weak. It is better to employ D-sub 50 for the VLBI purpose.


Figure A3-6. Current 1 G system uses four 256 Mbps cables.

Figure A3-6 shows succeeded Japanese 1024 Mbps VLBI system interface which uses four ID1(256 Mbps) cables currently . This is very confusing in cable arrangement. We will adapt our interface to the VSI standard when they are determined.


Figure A3-7. ECL data transmit (left) and receive (right).

An example of ECL clock / data quality at transmission and reception point is shown in Figure A3-7. This is the worst case established connection at 64 MHz. The amplitude of signals are reduced. The shape of signals are far from ideal condition at the reception point (right) We should not think little of cable in VSI.

Table A3-4. Investigated VSI Connector Candidate
Connector
Common
Name*
(pin)
Figure Approx size
Panel cross section
(mm)
Combination assembly
with shielded long
cable (20meters)
Price in Akihabara Application,
reputation and
history.
D-sub
(50 pins)
65x14YES$12-$20 Good reputation and long history. Large price dynamic range. D-sub 25 is very popular. Easy to check pin signal on site.
D-sub
half pitch (50)
52x11Unknown,Latch seems poor to support shielded cable weight of 15 mm diameter$10 SCSI, no long history. A CRL VLBI Giga-bit correlator use the connector at 32MHz with cable less than 2-meters. From our experience, less reliability than D-sub.
AMP
Centronics
(50)
80x15YES$15 Printer, computer, Share decreased. due to huge connector size and repeatability.
AMP
Centronics
Half pitch(50)
40x10Unknown,Latch seems poor to support shielded cable weight$18 SCSI
DIN VME
(64)
95x11Possible, huge connector size. No connector latch. Friction only. $15 VME, cable connector available. Occurs poor contact in certain manufactures connector combination.
MIL
Flat cable
68x6Plastic latch poor to support shielded cable weight$5 Computers, very popular cheap.
*Commonly used connector name. May be differ by country.


Appendix-A. Abbreviations

DAS Data Acquisition System
DIB Digital Input Box
DTS Data Transmission System
DOB Digital Output Box
DOT Data Observe Time
ROT Requested Observe Time
VSI VLBI Standard Interface


June 18, 1999