[evlatests] Receiver Status on Antenna 13 & 14
Robert Hayward
rhayward at aoc.nrao.edu
Tue Apr 4 19:53:22 EDT 2006
Brent, Chuck, John and I visited Antenna 14 and 13 today to investigate
several receiver issues:
Antenna 14:
----------
There was a complaint about no switched power being seen at C-Band. We
found that the receiver was warm. The Q-Band receiver was also warm. It
appears that Compressor B had died sometime between March 29 and 31
(when I first noticed that the Q-Band was no longer cool). Alas I have
been unable to access any EVLA Archive temperature data between these dates.
While the Q-Band dewar was pumping down quite well, the C-Band hadn't.
We found that the breaker for the 120 VAC used by the vacuum pump had
been tripped. No idea why that happened. The Cryo Group replaced the
compressor today and swapped out the Q-Band fridge since it had been
running for many days without proper helium pressure. Both receivers are
now cooling.
Rick had complained about the sensitivity of the RCP side of the X-Band
receiver being much worse than the LCP side. At the Rx output, a
Hot/Sky Y-factor indicated the LCP & RCP receiver temperatures were 24K
and 29K respectively. So the imbalance is not from the receiver. The
LO/IF Rack was not functioning because the L305 module had been pulled.
This meant we couldn't do a Y-factor at the output of the T304
downconverter. However, we did find that the RCP cable between the Rx
and the Rack had about 5 dB more loss than the LCP side, suggesting that
the sensitivity could be getting degraded here. We will replace the
suspect cable tomorrow.
Antenna 13:
----------
Rob Long had noticed that the detected Total Power of the L-Band RCP
channel measured at the T5 looked really awful. It exhibited a huge gain
oscillation with a period of several milliseconds. Thinking this may be
from microphonics in the coaxial probes on the OMT, we went out with a
crystal detector and an oscilloscope to measure it in the antenna
(general purpose power meters are to slow to measure this type of fast
gain instability). We found that turning the fridge off had no effect,
so it was not a mechanically induced problem. Tweaking the bias settings
on the RCP side did cure the periodic gain fluctuations. As this
receiver uses prototype balanced amplifiers that had shown a tendency to
break into oscillation, we checked the Rx output with a spectrum
analyzer. It was generating a nice birdy at 11.8 GHz.
As we had done the first time we found this problem in the lab, we
tweaked the drain voltages to make it happy (this time, VD1 & VD2 were
adjusted from 1.00V to 0.90V). I am a little uncomfortable doing this
since I have no way of knowing how this affects the input match of the
LNA's. This may cause the axial ratio of the polarizer to change. So
George Moellenbrock, if you see a big degradation in polarization
performance in your next run, this may be why.
I'm also curious as to whether the receiver had been acting like this
for some time or whether the oscillation was induced by the giant 48
volt sparks that occurred on this antenna last week.
There was also a complaint of no switched power seen at K-Band (has this
receiver not be tested out before?). A Hot/Sky Y-factor indicated both
sides had a Trx of about 33K (assuming a Tsky of 25K). We found that the
receiver produced a 0.2 dB increase in power when the noise cal was
turned on, which seem about right.
We did notice that the IF power at the input to the LO/IF Rack was about
-20 dBm on cold sky, which seems a bit high for the T304 downconverter.
We suspect that the RF amp on the T304 is likely being saturated. We
added 10 dB pads on the Rx outputs (i.e., before the UX Converter) to
help ensure that neither the T303 or T304 were running into compression.
However, we might need to add more. Perhaps someone in the Systems Group
might want to check on what the optimum power levels should be.
More information about the evlatests
mailing list