[evlatests] WIDAR wobbles vs. time, contd.
Michael Rupen
mrupen at nrao.edu
Thu Jul 16 11:24:30 EDT 2009
Hi Vivek --
> b. There are a variety of periods, a few Hz, quite a bit faster then the
> fringe rate.
We've seen, and you've mentioned, phase periods that are half those seen
in amplitude. The periods below I believe are from FRPLT, which
takes the Fourier transform of the (complex) visibility time sequence,
thus mixing the amp & phase together. I do not have an intuitive feeling
for what this does I'm afraid.
> c. The periods do change over an hour, but I see no way to relate them to
> the corresponding fringe rate changes.
By how much do they change? 10s of %, factors of 2, factors of 100? Do
wobbles vanish or just move around?
> d. There is no differential in the wobble across the band = it is _not_
> a delay change = there is no delay jitter associated with the wobbles.
> Plot 1: red= phase wobbles, averaged over band. blue= phase jump at
> delay change, differential between low/high half of band. time span =
> 10 sec, 40ms steps.
I'm confused by this plot.
It looks to me like the blue shows a smaller version of the wobble seen in the
red.
If blue= phase(channels 129-256)-phase(channels 1-128), then this
means that there *is* a difference between the low and high channels.
Help!
> e. The noise-diode switching at 9.6Hz is seen at ~few % in the autocor
> and is at about 0.1-1% of the cross-correlation amplitude on most
> baselines. Assuming 30K Tsys, 3K noise injection, that is ~30Jy, so
> the f_shift is suppressing the noise diode by a factor of >100 (1000
> on most baselines), else it would be comparable to 3C273.
I'm confused by this argument (have you noticed that I'm confused a lot? ;).
Should the noise tube correlated between antennas? or are you thinking
leakage of an antenna's noise tube through the sidelobes of its neighbor?
In either case how do you know how much of any suppression is due to fshift?
> h. Details of which baselines have what: (most have 9.6Hz at 0.2% or so).
> I think this says that the antennas with even slot IDs are worse -
> right Michael?
>
> 1-* clean
> 2-18 ~1.9Hz 1% of cross corr peak on 3C273
> 2-19 3.7Hz 5%
> 2-28 ~1Hz 2%
> 3-* clean
> 9-* 5Hz weak ~0.2%
> 9-18 6.5Hz 1%
> 18-19 6.5Hz 2%
> 19-* 9.6Hz 1-2% noise diode
> 19-24 2Hz 2%
> 19-28 3Hz 8%
> 23-* clean (23-3, 23-19, 23-25 have 9.6Hz 1%)
I believe the StB layout at the time of this observation (using AIPS'
labels, i.e., NOT accounting for the mis-labeling of 25 and 28) was:
Slot Rack 1 Rack 2 Rack 3
0 sb0006 = ea19-A = W04 sb000a = ea28-A*= N08* sb0004 = ea18-A = N16
1 sb0010 = ea01-A = W16 sb0009 = ea25-A*= N02* sb0005 = ea23-A = E18
2 sb000c = ea02-A = E02 sb000b = ea24-A = W14 (sb0025 = ea04-A = W10)
3 sb0002 = ea09-A = E06 sb001f = ea03-A = E16 (sb0018 = ea08-A = N12)
(The * is a reminder that what AIPS calls ea28 on pad N08 is actually
ea25 on pad N02, and vice versa. Antennas 4 and 8 were not included in
this observation.)
1- ea19 is the worst offender. Is it significant that its pad (W04) is
also closest to the barn?
2- 9.6 Hz wobblers
ea23 is at E18; {ea03,ea19,ea25*} are at {E16,W04,N08}
ea19 is the closest antenna on the west arm, and the one nearest the barn
No obvious pattern here for me!
3- Non-9.6 Hz wobblers vs. antenna/StB location
Vivek's list shows wobbles for antennas
{ea02 ea09 ea18 ea19 ea24 ea28}
vs. the even-numbered slots
{ea02 ea18 ea19 ea24 ea28}
So all the bad baselines are between antennas in even-numbered slots,
*except* antenna 9, which shows weak oscillations to all other antennas
and a stronger oscillation to ea18.
The bad antennas are at
{ea02 ea09 ea18 ea19 ea24 ea28*} ("ea28" is actually ea25)
{ E02 E06 N16 W04 W14 N08*} ("N08" is actually N02)
The good antennas are at
{ea01 ea03 ea23 ea25*} ("ea25" is actually ea28)
{ W16 E16 E18 N02*} ("N02" is actually N08)
So the good antennas are those furthest out on the arms, *except* "ea25",
which is in between the antennas at N02 and N16.
So either
- ea09 is the fly in the ointment, being in an odd-numbered slot; or
- ea18 is the one that ought to be stable, being a distant antenna; or
- ea28 (called "ea25") is the problem child, being clean but
located between two problem antennas.
AARGH!
To put the odd/even slot question to rest we might try
(1) taking 12-antenna data (if it's even slots this should give wobbles
to ea04), and then
(2) trading fibers for ea23 & ea04 (if it tracks slots ea23 will
then wobble, even though it's our most distant antenna).
4- Comparison with earlier observations
I re-examined the fast-dump C-band data from 29jun09 (observing 3C84)
using FRPLT over 2.5 minutes, to allow a direct comparison with
Vivek's result (1 min FRPLTs gave the same results but noisier).
The data below are for sb4, whereas Vivek's are for sb1; I will do
subband comparisons in a future e-mail, on the off chance those
will be instructive.
* ea01 is very noisy -- can't see much
* ea03-ea24 shows fringes in the FRPLTs over much of band
In the tables below I've put * by the baselines where Vivek also
sees a wobble.
Basic conclusion: I see more wobbles, except that I do not
see Vivek's ea09 wobbles. The worst offenders are the same in both
data sets.
a. 9.6 Hz: amplitude (odd units) from FRPLT:
r1s1 r1s2 r2s3 r1s3 r3s0 r1s0 r3s1 r2s2 r2s0 r2s1
W16 E02 E16 E06 N16 W04 E18 W14 N02 N08
ea01 ea02 ea03 ea09 ea18 ea19 ea23 ea24 ea25 ea28
ea01 ... ... ... ... ... ... ... ... ... ea01 W16 r1s1
ea02 1 - - *1 - 0.4 - 1 ea02 E02 r1s2
ea03 0.8 1 *1.5 *1 1 1 1 ea03 E16 r2s3
ea09 - *1 - - - 1 ea09 E06 r1s3
ea18 *1 - 0.5 - 1 ea18 N16 r3s0
ea19 *1 *1.5 *1 *1.5 ea19 W04 r1s0
ea23 - - *1 ea23 E18 r3s1
ea24 0.5 1 ea24 W14 r2s2
ea25 1 ea25 N02 r2s0
ea28 ea28 N08 r2s1
b. Other wobble frequencies:
ea01-ea28 3 Hz, amp: 4.5
ea02-ea09 1 Hz, amp: 0.3
ea02-ea18 6 Hz, amp: 5
ea02-ea19 1 Hz, amp: 0.5
ea02-ea23 1 Hz, amp: 0.5 (only neg. freq.)
ea02-ea24 1 Hz, amp: 0.5
ea02-ea25 1 Hz, amp: 1
ea02-ea28 1 Hz, amp: 0.3
ea02-ea28 4 Hz, amp: 0.3
ea03-ea25 1 Hz, amp: 0.7
ea18-ea19 3 Hz, amp: 2.5
ea18-ea25 1 Hz, amp: 0.5
ea19-ea25 0.5 Hz?, amp: 4.5
ea23-ea25 1.5 Hz, amp: 0.5, pos.freq. only
ea24-ea25 1 Hz, amp: 0.5
ea25-ea28 1 Hz, amp: 0.5?
Approximate wobble frequencies from FRPLT, in Hz; () --> amp <1.
Again * --> seen by Vivek, maybe not at the same frequency
r1s1 r1s2 r2s3 r1s3 r3s0 r1s0 r3s1 r2s2 r2s0 r2s1
W16 E02 E16 E06 N16 W04 E18 W14 N02 N08
ea01 ea02 ea03 ea09 ea18 ea19 ea23 ea24 ea25 ea28
ea01 ... ... (*) ... ... ... ... ... 3 ea01 W16 r1s1
ea02 - *(1) *6 *(1) (1) (1) 1 *(1,4) ea02 E02 r1s2
ea03 (*) - - - - (1) - ea03 E16 r2s3
ea09 * (*) (*) (*) (*) (*) ea09 E06 r1s3
ea18 *3 - - (1) - ea18 N16 r3s0
ea19 - * *0.5 - ea19 W04 r1s0
ea23 - (1.5) - ea23 E18 r3s1
ea24 (1) - ea24 W14 r2s2
ea25 (1) ea25 N02 r2s0
ea28 ea28 N08 r2s1
- Note that my "problem baselines" is ea01-ea28, which is the
only strong wobble which involves an StB in an odd-numbered slot.
I don't see any obvious consistent pattern here. But maybe someone
else will...
Michael
More information about the evlatests
mailing list