History does not repeat itself, but it rhymes.

Image may be NSFW.
Clik here to view.© Astron

As system engineer you sometimes stare at a picture or plot and spot something unusual. Then the thinking process kick’s in “what could cause this?”. You try to exclude all kind of possibilities. Is it a wrong calibration table? Could it be that the settings are applied incorrectly? You discuss it amongst colleagues and try to get grip on the issue. You know you have seen it before and has to do with wrong delay settings. Why would you else get a strong radio source split-up in two parts? This means that parts of the array disagree on the source position.

This time it was very subtle. On RS307 we had two different cable lengths (115m and 85m) in HBA array. The longest cable does not need extra delay but the shortest cable does. We applied the delays correction to all 85m tiles. After a quick calibration check by Cees Bassa and David McKenna they found 4-6 nS delay error on the 85 meter coax cables.

The difference in delay was however, caused by rounding! For LOFAR2 the calculated value of 24.6 was truncated to 24. In LOFAR1 it was however rounded to the closest value, being 25. After correction we got perfect “calibrated” all-sky HBA image, pointing at CasA.

In LOFAR1 we had something similar. That time it was that 80m of coax cable could have a difference of 0..8nS in delay (spec +/- 1nS). This has been corrected in the configuration table of LBA array of station DE605. In this way are eager to hunt down, as a team, all open issues in LOFAR2 by using the new PTS stations.