SD Forum #2 - minutes
Wednesday, 10 july, 1996
A. Conversion chain at Parkes .. G.Graves
The plan is to rationalise the conversion system at Parkes. There are two major components:
- The Focus Cabin Unit. A unit to which all receivers installed on the translator will be connected. This will have a set of RF switches so that the observer can quickly switch from one receiver to another, connecting the active receiver to the conversion chain. This unit will provide a universal conversion system translating the RF band centre to a standard IF frequency. Four IFs (two independent frequency channels will be supported, much as in the standard AT system). This will provide for dual-beam or dual frequency operation.
This unit is needed if we wish to exploit the flexibility offered by the new focus cabin translator.
- A modified AT conversion rack. A dual-frequency, four IF conversion system, based on the AT model. It will be modified to accomodate the Parkes-specific requirements (RF switches at the input; a new control interface; an upconversion scheme for the UHF IFs; possibly additional signal conditioners at the output).
The is a full discussion in Appendix A of the current design.
A number of points/problems were noted:
- The AT rack expects the IF frequencies to be segregated by cable - ie, separate cables for L-, S-, C- and X-band. But the Parkes design intends to provide multi-purpose cables - they may carry either 300 MHz, or L- or S-band. Thus an RF switching system is required, as well as some up-conversion scheme.
- The AT rack has very specific outputs - sampled data streams of standard bandwidths. It is not clear how non-standard users (such as the pulsar group) are to be accomodated.
- How would the PTI requirements be met - could they be incorporated into the system?
- Polarimetry will be discussed at a later meeting, but one may note here that the Bonn polarimeter should be accomodated with this scheme, since four IFs will be provided. Equally, the AT correlator could be used.
- The Focus cabin unit is a dual frequency unit; since many observers will be operating in a single-frequency mode, the unused IF pairs may be available to check-out the off-axis receivers.
- Setup and monitoring. A revised interface is proposed; this will allow manual setup, and better (compared to the AT units) monitoring of the signal path settings. The computer interface will not follow the AT model.
- The AT system provides for precise phase shifting between the two IFs; this is used in the linear-to-circular polarisation conversions of the VLBI systems. How will this be managed?
- Interference is always a potential problem. Some difficulties have been detected with the National GPIB unit, as well as leakage from the present focus cabin unit.
Timescales:
- The Focus Cabin unit - several months
- The modifications to the input to the rack - 1 month
- The interface - several months
- The upconverter system - ?
- Completion of the rack wiring (the RF wiring) - several months
B. Parkes Utilities .. P. te Lintel
Three categories were identified:
1. General purpose utilities, such as Coco; Coords.
These represent a modest investment and will be maintained, revised and supported as required.
2. Processing tasks, such as SPC; SLAP.
Debate on this issue was postponed to the next meeting.
3. Observing tasks, such as Spectra; Spot; Multi(beam); PTI; Scan.
The difficulty is to reconcile the basic needs with the long term (ATOMS). SCAN currently does not work, yet it is needed for some experiments which have been given time by the TAC. (There are both hardware and software problems to be resolved). SPOT is usable, but far from convenient when used in conjunction with the FCC. MULTI has been developed to cater for problems not amenable to SPECTRA.
The ATOMS design brief is to provide the software to control the observing sub-systems, along with a suite of tools which allow the observer to construct his observing task. The obvious utilities (eg, SPOT, SPECTRA) will be provided, but the observer will be encouraged to recast variants to meet his specific needs.
Alan Wright is in the process of compiling a list of the current utilities, identifying their functions, usefulness and ideosyncracies. This will help define the bridging effort required to keep the observatory functioning, as well as clarifying the scope of the tools which ATOMS will need to provide.
4. Data formats for spectral line processing. FITS and other formats. The native Spectra format can be read by few processing packages. A translation operation is needed if packages other than SPC and SLAP are preferred. Several utilities are available at Parkes; users should consult P. te Lintel, or check the URL:
http://www.parkes.atnf.csiro.au/library/
and select the Reduction of Data cupboard.
In the longer term a universal SDFITS will be defined, and the requirements will change.
C. FCC calibration at 22 GHz - J. Reynolds (This material was not presented at the meeting, but is relevant to current operations, so is included here rather than holding it over till the next meeting). Data on the axial and lateral focussing at 22 GHz were obtained in June. fig. 1 , fig. 2 and fig. 3 show the data along with least squares fits. The axial focus appears to show a night/day difference of about 6 mm. The focus cabin auto-trackingt task, TRKFCC currently uses a SIN(elevation) dependance - the data here suggest that a second order polynomial may be prefered.
Of more interest to the observer is the 22 GHz gain-elevation performance, shown in fig. 4 and fig. 5 . Fig. 4 shows what can be achieved with TRKFCC in operation, while fig. 5 shows the likely penalty if it is not used. The bottom line is that TRKFCC, once calibrated, works well.
Appendix A -- G.Moorey, G.Graves
NEW CONVERSION AND LOCAL OSCILLATOR SYSTEMS FOR PARKES
(discussion paper)
1. Introduction.
The existing RF Conversion System at Parkes differs from that at Culgoora on the Compact Array and at Mopra, largely for historic reasons. However there is an increasing move to build similar Front Ends for the AT sites. It would greatly reduce future effort and improve ease of operation, flexibility and maintenance if the Parkes Conversion System was similar to that at Culgoora and Mopra. Now is an appropriate time to make this investment.
To complement the new Focus Cabin and Front End Translator, where up to four Front End frames can be mounted simultaneously, it is proposed to update the RF Conversion system and to install a versatile and frequency agile Local Oscillator system.
During the construction of the AT Array sufficient Conversion modules were built to equip Parkes with a four channel Conversion system. However due to labour shortages the Conversion rack to hold the modules was never built. Times have changed, and now we can build such a system
We therefore propose the installation of a four channel AT style Conversion system using commercial frequency synthesisers as Local Oscillators.
2. Parkes Front Ends.
Table 1 is a list of existing Front Ends that can be mounted on the Focus Cabin Translator and details the RF input and output bands of those Front Ends. All Front Ends are dual polarisation unless otherwise stated.
3. Control Room to Focus Cabin RF Cabling
At present the only media for transmission from the Focus Cabin to the control room is low loss coaxial cables. The 10 half inch superflex type FSJ4-50B cables installed from the cabin to the control room have 17.5 dB loss per 100 metres at 2 GHz and at 4 GHz the loss is 26 dB per 100 metres. If we set the useful broad-band range to say 50 to 2300 MHz this would then define the required conversion system in the Focus Cabin to best utilise these cables. It is also noted that half inch superflex is over moded above 10.2 GHz.
4. Old Conversion System
The existing dual channel conversion system known as the Batchelor Box, as used in the old aerial cabin has broad band mixers capable of converting UHF , L , S , C or X band down to a 0 to 700 MHz base band. fig1
To maintain operation while a new conversion system is installed we would retain the Batchelor Box conversion capability for the existing narrow band frequency translators located in the control room.
However, by inserting a power splitter and coax cable drive amplifier after the mixers and before the 700 MHz low pass filter it is also possible to get a broad band 50 to 2300 MHz IF output.
5. New Conversion System
Because the AT conversion system has L , S , C and X band inputs it is proposed that all Front Ends higher in frequency than 10 GHz should contain a built in conversion to either C or X band. Referring to the list of Front Ends an example would be the K-Ku Band Front End which converts down to 4.5 to 6.0 GHz. (see fig2 and fig3 )
If the outputs of the UHF , S , C or X band Front Ends were converted to L band and provision made for the L band Front End output to be directly patched in after the power splitter so as to use the cable drive amplifier in a modified Batchelor Box type system then only the L band, 1 to 2 GHz , input of a standard AT conversion rack would be required at present. fig3
A modified Batchelor Box type conversion system would be built in standard AT modules and mounted in a bin in a Back End rack in the Focus Cabin. This would be a four channel conversion system with multi-pole RF switches at the input to select the required Front End and output cable drivers for the 50 to 2300 MHz band. The new Focus Cabin conversion modules will be computer controlled and monitored via a Data-Set or locally controlled from the front panels. fig3
The Parkes conversion rack will be similar to those used in the 22 metere antennas at Narrabri and Mopra. It will provide four independent IF channels allowing input frequencies from 1.2 to 12 GHz. If a UHF/VHF Splitter module was constructed to up-convert to L band , input frequencies in the range 300 to 800 MHz could be accomodated.
The Conversion rack will provide four channels of 256 , 128 or 64 MHz bandwidth direct to the correlator samplers. It is also planned to take the signals out at the splitter modules to provide 1 GHz broad band RF and detected outputs for each channel. fig2
By installing an AT Digital Narrow Band system, bandwidths of 32 , 16 , 8 , 4 , 2 , 1 down to 0.0625 MHz can also be obtained. fig4
Both Digital Synch Demod and Analogue Synch Demod modules will be provided.
The standard AT conversion system would be built into a 39U rack with integral power supplies and control and monitoring data-sets. There will be a spare bin in the rack to hold an analogue fibre optical transfer system when available. The conversion system can be either computer configured, controlled and monitored or set up locally from the front panel. The local oscillators can also be computer controlled via the GPIB or setup from their front panels.
The AT conversion system and its associated Local Oscillator sources would be resident in the Control Room for easy access, cabling and maintenance.
6. Local Oscillators
To accomplish the frequency conversions, stable frequency agile Local Oscillator, phase locked to the station 10 MHz standard are required in the Focus Cabin fig4 . The most efficient way of supplying such a source is to purchase commercial GPIB controlled frequency synthesisers covering 1 to 20 GHz.
One synthesiser would be used as the first conversion local oscillator in the K-Ku Front End or any Front End with a higher input frequency than 10 GHz. This first conversion would be either to C band (4.5 to 6 GHz) as in the K-Ku Front End or possible X band (7 to 9 GHz) for higher frequency Front Ends.
The second synthesiser would be used as a tunable local oscillator for the mixers in the modified Batchelor Box system and for simultaneous dual frequency operation such as S and X band VLBI in conjunction with the afore mentioned synthesiser. These synthesisers should cover the 1 to 20 GHz range for maximum flexibility.
The AT conversion rack in the control room requires three synthesisers. fig5 -one for converting the C and X band inputs to L band, one to convert the L and S bands to UHF and one for the UHF to baseband conversion. The UHF to baseband synthesisers have already been purchased. Thus we require a total of four microwave synthesisers of which one exists covering the 1 to 20 GHz range.
It is therefore recommended that we buy three more 1 to 20 GHz synthesisers to complement the existing unit in the sense of spares , installation and operation.
7. New Cabling required.
Excluding the H-Line multibeam receiver which is a stand alone system the maximum number of Front Ends that can be mounted on the translator at any one time is four for example the C-X Multiband, K-Ku band , Galileo and the Methanol C-Ku. These four have a possible requirement for 14 RF/IF output cables and 4 input LO cables.
Thus we require a minimum of 18 superflex coaxial cables running from a bulkhead termination panel on the translator through a cable snake up to a bulkhead panel in one of the equipment racks.
From the new Focus Cabin conversion system built in standard AT modules there will be four output cables running to the Focus Cabin wall bulkhead. Because the outputs are 50 to 2300 MHz they will be run in the lower loss half inch Superflex cable.
A Front End DC supply system capable of running 5 standard AT Front Ends or the Multibeam Front End together with 2 other standard AT Front Ends will be installed in an equipment rack complete with monitoring and control electronics and a Data-Set. This DC system is contained in two bins. Shielded cabling from the DC supply will be run via the snake to a distribution boxes mounted on the translator. Due to the long cable run from the power supply system extra RFI/noise filtering will be installed in the DC distribution boxes.
A distribution box on each side of the translator will be fitted with 3 fourteen pin Cannon connectors wired in parallel.
A Back End DC supply system identical to the Front End DC supply system will also be mounted in the equipment rack to power any associated back end and general electronics situated in the Focus Cabin. Distribution will be via terminal strips mounted on the bin extenders within the rack.
To control and monitor the Front Ends it is proposed to mount a bin of 6 Data-Sets in the equipment rack. Four of these would each have their bus signals and 2 analogue channels routed via the existing snake to 4 individual bulk head connectors mounted on the translator. The other two Data-Sets can be used to control and monitor back end and general electronics in the Focus Cabin.
The above method of terminating cables at bulkheads or distribution boxes on the translator will make it quick and painless to install and remove Front Ends and then for rotating Front Ends a simple system of coiled cables will suffice.
8. Future Developments.
In the future it would be desirable to purchase a multi channel analogue optical fibre system capable of transmitting L , S , C , or X band directly to the standard AT conversion rack in the control room. This could replace the Batchelor Box and give immunity to RFI. We would still require the synthesisers in the Focus Cabin for the higher frequency front end first conversions.
9. Summary
The following is a list of benefits of the proposed conversion and local oscillator systems for Parkes.
- Standardise the Back Ends at all AT sites.
- Standardise AT Front End design resulting in savings in labour and cost
- Provide four independent IF channels
- Improve the ease of operation leading to a labour saving
- Increase the overall flexibility and frequency agility of the back end
- Improve the user friendliness of the system
- Increase the reliability through modular construction
- The modular system will allow easy updating and future expansion
- Enable computer control and monitoring of the complete conversion and LO systems
Table 1 - PARKES FRONT ENDS NAME OBSERVING BAND RF/IF OUTPUT BAND 70 cm 436 MHz 436 MHz 50 cm 660 MHz 660 MHz HOH 1.2 to 1.7 GHz 1.2 to 1.7 GHz H-Line Multibeam 1.25 to 1.55 GHz 1.25 to 1.55 GHz GALILEO 2.2 to 2.4 GHz 2.2 to 2.4 GHz MULTIBAND S 2.2 to 2.5 GHz 2.2 to 2.5 GHz MULTIBAND C 4.5 to 5.1 GHz 4.5 to 5.1 GHz MULTIBAND X 8.0 to 8.9 GHz 8.0 to 8.9 GHz METHANOL C 6.0 to 6.8 GHz 1.25 to 1.75 GHz METHANOL Ku 12.0 to 12.4 GHz 300 +/- 150 MHz K-Ku Band 13 to 15 GHz 4.5 to 6.0 GHz K-Ku Band 21 to 24 GHz 4.5 to 6.0 GHz Q - Band Maser 43 GHz (single pol) 300 +/- 30 MHz