[daip] Re: SDA in VLA datasets

[douglas bock]

> 1. What was your observing frequency?

4.8851 and 4.8351 MHz

> 2. Several years ago we did a project to obtain single pulses and
> average profiles from Vela at many frequencies using the phased array
> mode. Vela is so strong that we could never get the off-pulse levels
> correct; the Automatic level control started to operate at each pulse.
> Its time constant is longer than the pulse width (actually somewhere
> between 0.5 and 3 seconds depending upon the antenna, as I recall from
> some measurements by Bill Coles about 10 years ago.) Could it be that
> the AGC recovery during your off-pulse sampling window could cause
> artifacts?

I think you imply that the gating takes place before the power measurement
for the AGC. If so, then I think I can rule out this cause, since some
heavily flagged data (in which the first and last ~minute of each "off
pulse"  scan were flagged) still shows a pulsar residual.

On the other hand, if the gating takes place after the AGC, then could
gain variations due to the longer term variability of the pulsar cause
phase-center artifacts? The pulsar is 0.37 Jy averaged over the pulse. I
calculate SEFD=18K for Tsys=60. The pulsar can vary by nearly a factor of
two over minutes. I have data (not immediately to hand) from ~10 min scans
on the pulsar that I could use to look for variation on timescales of our
integration time (5s).

(almost all the pulsar comes in just a millisecond, but as you say the
period, 89ms, is quite a bit shorter than the AGC time constant).

Douglas