Corrupt Data Suspected

Use the SPD and VIS displays to look at your data. If something seems wrong, consider the possibility of:

a display error:

do SPD and VIS concur? If SPD is very strange, try restarting it:

	In SPD type:
	$>$ exit		Or + .
	CACCC$>$ spd		From a CACCC window on FURIOUS.

interference affecting short baselines worst:

Sinusoidal amplitude visibilities at 20cm during the day are a fact of life.

a correlator block problem affecting all baselines using that block:

reconfigure the correlator.

operator error:

your source really looks like that!

Example: The following discussion considers a typical problem, and presents a diagnosis strategy that might give you some ideas.

Problem: The amplitude visibility display on VIS shows that one baseline is unlike the others. You should:

CHECK ASSISTANCE

Does it give any useful advice or errors?

Use the color-coding in VIS to identify the baseline of interest. If needed you can zoom in on a particular plot using:

VIS$>$ zoom [top|centre|bottom]

Where [top|centre|bottom] refers to the plot you wish to zoom into.

Select a rectangle in the highlighted plot by clicking the left-mouse-button on the lower left corner and the upper right area of your region of interest. It should then be possible to identify the baseline. To exit the zoom mode:
VIS$>$ scale

Alternatively you can eliminate selective antennas to show all baselines without those antennas, and determine the baseline that way. For e.g.,

VIS$>$ select aa,bb,-5

Shows all baselines excluding antenna 5.

For example if your identified baseline is 34AA, this corresponds to the XX polarisation of Frequency 1 for the baseline between antennas 3 and 4. (If needed see definitions of the VIS and engineering (used by CAMON) codes for different polarisations.)

	To view that baseline in greater detail:
	VIS$>$ select 34		This will show all polarisations on both frequencies.
	To view the spectra use SPD:		Runs from CACCC, and is usually displayed on Furious.
	CHECK the lag spectrum		This is the correlation function or Fourier transform of the frequency spectrum.
	SPD$>$ l Go to a strong calibrator		Flux $\, \buildrel > \over \sim \,$2 Jy.

Once the antennas have arrived at the source you should see a single, centered spike. (To see example lag spectra of bright and weak calibrators ) If you do not see this, then the interferometer is not generating fringes. This is probably a correlator block failure

You should also CHECK the power spectrum:

SPD$>$ a

For amplitude.

If the spectrum is not fairly flat in the central half, then you have problems. A sloping spectrum suggests a correlator problem (consult the DA), whereas narrow features are common due to the inevitable interference.

Also you should CHECK it isn't an ACC failure. Check ATDRIVEMON. To fix, reset the problem ACC

If the problem affects only one polarisation on all baselines for a single antenna, suspect some aspect of the receiver-IF chain. Use CAMON to investigate the receivers.

E.g., to inspect Antenna 3:

	$>$ /rx/3		This shows the receiver hardware status of antenna 3.
	For a menu of other receiver pages:
	$>$ /0