Cable Calibration

In order to form the best possible result out of the LOFAR system, it is essential that we can calibrate for the signal travel-times. This means not only being able to know the direction from whence the signal came, but also how long it takes to travel through every cable and every circuit.

Each cable within the system is carefully manufactured and labelled. The serial number on the cable contains information about its performance and role. For example, the one in the photograph (CHY-0853425-1305090755) is an "C"able from the "H"igh-Band Array, "Y"-polarisation. It is "085" metres long and has a delay of "3425" tenths of a nanosecond. This means that, in laboratory conditions, it takes 342.5 nanoseconds for a signal to travel down this cable.

However, that is in the lab.

Although this is a pretty good measurement, to really get the best performance we need to know the signal delay in the field, now that the cable is in place. Things such as connectors, thermal variations and other deployment factors can contribute to subtle variations in the effective cable length. Calibrating these sorts of things is part of the vital commissioning work that is currently being carried out on the LOFAR Chilbolton station.

Fibre

Yesterday and Tuesday, the BT engineers were on site to install the network termination equipment (NTE) for the 10 gigabit link. This is the connection equipment that will transmit the data from the main Chilbolton building to the next major network point which is located in Southampton.

Caption 1 : The BT van arrives at the main building at Chilbolton Observatory

Caption 2 : Installed. This is what the connection looks like in the comms room at the observatory.

IT rack

On 27th August 2010, work started on installing the IT equipment in the main building. Here are the first two modules to be added. The top one is the optical fibre patch panel for the 8-core fibre that runs between the LOFAR field and the Main Building. Below it is the 10-Gbit/s network switch.

Data link

On 16th August 2010, the ISDN control line was established. Between the Chilbolton site and the supercomputing centre in Groningen, there is 10 Gbit/s link. However, this link is mono-directional. We can send data to the Netherlands, but there is no way to send data back. For the data itself, this is fine. However, it is also necessary to send back control signals. For the telescope a conventional Internet connection is used. However, for the control of the 10 Gbit/s transmission equipment a second ISDN line is required.

Caption 1 : The lines run through a tunnel on the site. This view is looking down the tunnel in the direction of the termination point, where the BT lines first come onto the site.



Caption 2 : BT engineer fitting the ISDN connection in the room where the terminations are.

Data link

Between the RF-container and the main building at Chilbolton Observatory, there is the LOFAR data link. This is a high-capacity optical fibre and the data rate is typically up to 10 gigabits per second. With the electronics mostly installed, it was time to start getting this fibre run through the ducts and conduits between the two locations.



Caption 1 : The contractor vehicles assemble in the main observatory car park.




Caption 2 : Preparing to start spooling the fibre through one of the draw pits.



Caption 3 : The drawing of the fibre lasted the entire day, through the rain too. However, by the end of the day, the line was in place ready for termination at each end.

Digital signal processing

While the HBA tiles are being completed, work continues on the installation of the digital signal processing electronics.

Caption 1 : After the signal has been digitised, there are a number of tasks that are done on the data. Firstly, there is the initial delay compensation which is applied for the lengths of the cables. Because the cables across the fields are different lengths, due to the physical distances, this needs to be corrected. Then the received signal is divided into sub-bands, a bit like radio channels. A small sub-set of these are selected and are then combined to final beam, thus selecting the part of the sky to be observed.



Caption 2 : Additionally, statistics are calculated and engineering parameters are monitored. These are used to not only analyse the data for interesting features or to apply calibration corrections, but also to ensure that the equipment is performing correctly and to detect potential errors or interference.

Last of the analogue signal path

With the sub-rack assemblies now in place, the signal cables can be connected. These come from the patch panels, under the false floor and up onto the F-type connectors on the front of the processing units. Until now, the signals have been analogue from the antennas, front-end amplifiers and electronics through to this point. Once they go into the receiver units they are subjected to some final amplification, switching and filtering, but are then digitised. From then on, the signals are digital and all subsequent processing happens in the digital domain.



Caption 1 : The first batch of cables are fed under the racks and are connected onto the front of the RCUs (Receiver Units)




Caption 2 : Sientje Hartman (Excel Assemblies) makes excellent progress in patching the signal cables into the RCUs. Soon many of the panels are covered in connections.

Sub-racks ready for cabling

Within the LOFAR RF-cabin, there are four, full-height 482.6mm (19") rack frames. The sub-rack assemblies go into the first three of these and the fourth is for the control computers, clocks and networking. Once the sub-rack assemblies are in place, they can be cabled up with the white signal cables that come from the back of the patch panels.



Caption 1 : Tonjes Jipping and Peter Gruppen lift the last sub-rack assembly into the RF-container and secure it into its rack frame.



Caption 2 : Ready for cabling! The signal processing electronics are now in place and are ready to be connected up.

The other side of the cable patch panels

For those of you who were following the work on the cables during the early days of the LOFAR-Chilbolton installation, you may recall that they went from the aerials and tiles, through the trenches under the ground before eventually surfacing at the RF-container. There, they were patched in to the outside of the RF shielding on a special patch panel (Ref: http://blog.lofar-uk.org/2010/05/all-patched-in.html). On 13th July 2010, the installation picks up where that work left off... on the other side of the patch panels.



Caption 1 : Ineke Mol (from Excel Assemblies) starts patching in cables for the LBA.




Caption 2 : The LBA cables are now complete. From the inside of the patch panel, the LBA cables snake their way under the false floor over to where they enter the signal processing sub-racks.



Caption 3 : Here, you can see some of the LBA cables that are now ready. They are the ones that lead off the top of the photograph. The coiled cables have not yet been fully deployed and are waiting for their installation.

Packing the electronics

With the HBA nearly complete, by the 12 July 2010 it was time to start
preparing the electronics for installation in the RF-container. These
boxes had arrived the previous November and had been patiently waiting
in storage ready for their final deployment. The following day, the
ASTRON installation team were due to arrive, so all the equipment was
readied for transportation out to the field.

Caption: Boxes of sub-rack assemblies, patch cables and network
electronics, all queued in the workshop ready for loading into the
observatory van for transport to the LOFAR field.

Checks and tests

At every stage of the installation work, the processes are being checked, tested and inspected.


Eric Petrusma from Autonational inspects a cable pair. It is important to ensure that the cables are always in good condition.



The 18-tonne digger does a test drive over some geotextile. As we will need to drive the digger over this fabric later, it is important to assess whether or not the tracks damage it in any way. (They don't!)

More work on the HBA cables

Before the HBA tiles can be placed, the cables need to be prepared to accept them. This means that the metal support stakes are removed, the surrounding rubble extracted and the support pipe taken off. They are then set to their final position ready for the next stage of the project. This work is being done by the engineers from Autonational.

Zabet Ahmadi opens up the cable egress points ready for separating and preparing the cables.


Stuart Keenan starts filling in the cables again having set the template to ensure that they are at the correct separation.

Another day on the LBA

Just another nice photograph of the LBA. Looking through a cable loop, you can see the laser leveller and some of the gravelled surface of the new LBA. In the background there is the 25m Chilbolton dish.

Cable Testing Revisited

As you may have seen earlier, every cable that we installed was tested before it went into the ground. This was to check that they worked according to the tight electrical specification. Well, now that they are all in place, they have to be tested again to ensure that they are still working.

During the installation process, the cables are exposed and vulnerable. As the backfilling work occurs, it is possible that a tool, rock or other object causes the cables to be crimped, abraded or even cut completely. And, of course, now that the burying has been completed any discovered fault constitutes a major problem.

To test the cables, the same vector network analyser is used. However, instead of testing them individually, they are tested in pairs. A signal is sent from the RF-container patch panel out one cable. At the field end, a small loop back cable is placed to direct the signal out of the first cable back down the second cable. Then at the RF-container, the other end of the vector network analyser is connected to the second cable and the round trip signal degradation is tested.

Harry Smith attaches the loopback to a pair of antenna cables on the HBA. As part of this process, the connectors and cables are visually checked and inspected and afterwards they are re-sealed and secured.


Meanwhile, back in the RF container, Jon Eastment looks at the results of the cable-pair under test.

LBA cabling nearly done

Here are some photographs of the work being done on the LBA cable grave.

Martin Bell pulls back some of the final cables across the LBA grave. This is the last time they will look all neat and tidy.


Now that the cables have been all patched in, the cables in the grave are "shuffled about" to make them sit better and reduce stress points. This does make them look rather messy, but it does help ensure cable integrity.


The next step is the covering of the cables with the protective sand layer.


Then the crew move their way up towards the RF container, following the feeder trenches.


Almost there.


Finally, the sand covers all the cables. Warning tape is placed over the area (should anyone excavate in this area in future). Then all the graves are backfilled.

Backfilling the LBA

Now that the cables have been run out for the LBA, it is necessary to cover them with a layer of protective sand and then backfill with the soil and rock that was originally dug out. This is not easy, as the LBA trenching pattern is complicated and it is difficult for the diggers to manoeuvre between any partially filled trenches.

The first step is to apply the sand. A narrow bucket is attached to the digger and builders sand is scooped from the dumptruck into the trench. This is then carefully spread by hand to ensure that the cables are properly protected.


Once the sand is in place, the wide bucket is attached to the digger and the piles of chalk and soil are used to back fill the trench.

Sentinels of the HBA

Like soldiers on parade, the HBA field is covered with pegs and cables, waiting for the next stage of testing. The white plastic pipe and the steel peg are removed when the tiles are put in place. However, for now they keep the covered connectors up off the ground and protect the cables against any movement that may have occurred during the leveling and compacting process.

Final run of the LBA cables

The LBA cable work has now been complete. All the cables from the LBA have been run from the field into the RF container.

Looking down a trench, there are numerous cables stacked up, waiting to be routed through the LBA cable grave.



Harry Smith working a new line through the LBA cable grave. Like the HBA, the LBA also has a cable grave to take up cable excess before it goes into the final RF container.


Brian Finegan prepares to feed the final cable pair through the duct and into the RF container.

Mind Your Language!

There's been a pretty diverse team at Chilbolton working on the LOFAR installation, with staff and students from many different institutes and different origins. During the patching of the cables in the RF container, it became necessary to segregate the panel into two parts, according to whether the upper or lower ingress duct was to be used. A marker pen and some insulation tape were used to add some temporary labels. But for some reason, someone added a second set of labels on the panel in a different language. Then the idea had caught on and everyone who was there on the day added a language that they knew. By the end of the day a total of seven languages accumulated. Obviously the sixth line down is English, but do you recognise the others? Answers can be submitted by posting a comment.

Sand People

Now that the HBA cables are in place and patched into the RF-container, the HBA cable grave can be filled in.

First a thick layer of protective sand is spread across the cables. The crew work the cables gently to ensure that the sand penetrates between all of them.


Here is the finished sand cover. This protects the cables against flint and other sharp rock as the spoil is back filled into the grave.


With all cables completely covered, the digger moves in to bulk fill with previously excavated material. And that is it! The HBA cables are fully covered and the crew can walk through the area without the risk of falling into the hole or collapsing a trench wall.


There have been many people involved in the laying of the HBA cables, and here are a just a few of them. From left to right: Dave King, Jon Eastment, Mike Willis, Alejo Martinez-Sansigre and Harry Smith. To these few, and everyone else involved in this stage of the project: thanks and well done!