- measuring the drift.
- changing the attenuator
- monitoring the drift
after attenuator changed.
Jumps,ripples in the xband power was
seen with the udc & mock spectrometers (only polB was
monitored). The ripples correlated with the dehumidifier
running in the pedestal. Some debugging showed an instability
in the rf filter, attenuator interface in the rf chassis in
the pedestal. Changing the attenuator caused the ripples and
jumps to decrease in amplitude.
01,03,05 mar 16: monitoring the xb
total power. (top)
mar16 simultaneous data was taken with the sb and xb receivers
on the 12 meter telescope. This was before the attenuator after
the the xb rf filter was replaced (on 10mar16).
The setup was:
- polA fiber rcvr (upper most)-> amp -> mixer (2270
to 325MHz) -> amp -> filter -> amp -> mock
mixer -> mock spectrometer
- cfr 2270 MHz, bw: 172 MHz
- polB fiber rcvr -> udc(1.5-2ghz output) -> mixer
(udc out to 325 MHz) -> mock mixer -> mock
- cfr 8605 MHz, bw: 172 MHz
- Mock spectrometer:
- 1 second dumps, 172.032 MHz bw, 8192 channels
- For each 1 second dump the total power was computed over
the entire 172 MHz. (no rfi excision was used)
- For each data set, the total power was divided by the
median power (to switch the units to Tsys).
- Temperature monitoring:
- on 03 and 05 mar16 the temperature in the wapp room was
- this room has the fiber receivers,udc, and mixers.
Plots of total power vs time.
- 01mar16 sb,xb total
power vs time (.ps) (.pdf)
- The telescope was sitting at az=20, el=45 deg.
- data acquisition started at 09:40 and lasted for about 2
- black is the sband total power, red is the xband total
- vertical lines were placed every 550 seconds. The period
is not constant.
- to room temperatures were recorded during this data set.
- the peak to peak value of the ripple is about 3% of Tsys
- the total power also varied by about 20% Tsys during the
- sband total power was relatively stable.
- 03mar16 sb,xb
total power with wapproom temperatures overplotted (.ps)
- The telescope was sitting at az=20 , el=45 degrees
- black is sband totPwr, red is xband totPwr, green is the
wapp room temp, blue shows a 1 degF temperature step.
- Top plot: data acquisition started around 08:30.
The sky was clear.
- The temperature sensor was on the ceiling above the
mock spectrometer. A fan was blowing directly on the
- the green temperature ripple is close to the period of
the xband ripple..
- But the temperature period drifts relative to the
- So the xband ripple is probably not coming from the
temperature variation in the wapp room.
- The sband totPwr is relatively stable.
- Bottom plot: data acquisition started around 15:30 (not
sure whether it was cloudy or not).
- the temperature sensor was sitting on top of the udc
(shelter from any fans).
- The xband ripple was not as evident.
- The drops in power toward the end (15% at the
end) may or may not be weather related?
- 05mar16: sb,xb
total power, turning dehumidifier on,off (.ps) (.pdf)
- The telescope was sitting at az=20, el=45 degrees.
- data acquisition started around 08:30.. the sky was
- There is a dehumidifier in the 12meter pedestal. It
blasts hot air into the pedestal when it is running.
- I saw it running when the relative humidity was 50% or
- Around second 3100 I opened the pedestal door and turned
off the dehumidifier (the blue line).
- This time stamp may be off by a few minutes...
- about 500 seconds later the xband oscillation stop..
but the linear ramp continued
- Near second 6400, the dehumidifier was turned back on
- at 6700 seconds, the signal jumped by about 30%.
processing: x101/ 160301/p12mripple.pro,
- The xband ripple is coming from inside the 12 meter
- the periodicity is coming from the dehumidifier in the
pedestal turning on off.
- Some xband components in the pedestal are very sensitive
to temperature variation.
- I wonder if this could be a loose connector that is almost
always making contact?