1.3mm Receiver Status Report
The 1.3mm JT receiver has had a few changes since it was removed from the mountain this past observing season, so George and I thought we’d give an update on the status of the receiver for the new observing season. We have made some changes which should enhance the performance of receiver as well as solved a few problems that were observed with the receiver last observing season.
The measured DSB & SSB receiver noise temperatures have been measured both in the lab and on the telescope since it was installed last week:
The graphs show that the performance in channel A is better than in channel B. It is suspected that the problem is related to the mixer because several tests were done with the optics, and the poor performance always traces back to the insert. Unfortunately, time ran out in the lab to allow us to try another mixer. Also, note that these mixers are in the opposite channels from the spring 2005 observing season. Last season, mixer 373-3 was in channel A, and mixer 373-5 was in channel B. Now, mixer 373-5 is in channel A, and mixer 373-3 is in channel B.
2. LO Frequency Coverage
A more balanced 3 dB power splitter off of the Gunn oscillator was installed. It was hoped that this coupler would improve the coupling of LO power to each channel, but unfortunately, this did not turn out to be the case. Channel B was found to lack sufficient LO power below 210 GHz. Therefore, channel B can only be tuned at LO frequencies from about 215 GHz to 260 GHz. This should allow for a sky frequency coverage from around 210 GHz (LSB) to 265 GHz (USB). However, Channel A will be able to cover from 200 GHz (LSB) to 265 GHz (USB
3. LO Modulator
Ferrite modulators were installed for both channels. This allows for the control of LO power to set the mixer bias current at the console of the computer instead of having to go to the receiver room to adjust the current using the old level-set attenuators.
4. Electromagnets for the Mixers
The permanent magnets have been replaced with electromagnets. This allows for a much better control of the magnetic field used to suppress Josephson noise. Also, this should help with the stability of the receiver. Past total power measurements have shown kinks where the mixer was biased. These kinks in the total power can be removed by adjusting the current supplied to the electromagnet. This should allow for much more stable operation of the mixer which will no longer require “de-tuning” of the bias point to achieve high sensitivity and stable operation.
5. Receiver Stability
Last observing season, it was found that the total power level on the IF attenuator display fluctuated more than normal. This effect was observed with a FFT analyzer by monitoring the detected IF signal of the receiver. David Forbes found the source of the instability to be related to the LED displays of the IF attenuator box. Each time a digit would change, the LED display would draw enough current to inductively couple into the HFET IF amplifier display. This would cause the gain of the amplifier to change the output power of the IF thus, causing the LED readout to change. This would cause an unstable feedback effect which would adversely affect the stability of the receiver. We found that if we powered the LED displays with an external power supply, the inductive coupling to the HFET bias supply no longer occurred thus, significantly improving the receiver stability.
6. Rooftop Mirrors of the Martin-Puplett Interferometer
This past summer it was learned that the height of the rooftop mirrors when using the MPI was important to achieve the lowest receiver noise temperature. Measurements were carried out to best optimize the mirror height to yield the lowest receiver noise temperature. Many of the positions used last season were found not to be at the optimal mirror height. A table was generated for the operator so that the mirror can now be positioned at its optimal height.
The channel B rooftop mirror drive motor was found to oscillate at the frequency of the cold head drive motor. The bearings in the drive shaft housing were worn causing a considerable amount of slop in the rooftop mirror drive. The motor was replaced, and this solved the problem. The rooftop mirror servos now run very smoothly and are much more precise.
7. Insert wiring
One of the inserts had an intermittent connection on the 7-pin microtech connector used for the mixer bias. This connector was replaced with a MDM 15-pin connector which has proved to be much more reliable at cryogenic temperatures than the microtech connector. Also, heat sink bobbins were installed for the phosphor-bronze wiring harness.
8. Cold dumps
The orientation of the cold dumps was found to be incorrect. This stemmed from the decision to use the NRAO warm optics plate instead of the ARO warm optics plate when the receiver was installed last spring. The NRAO warm optics plate required the cold dumps to be moved to a different location from the original location using the ARO warm optics plate. The change in orientation caused the cold dumps to be misaligned with the polarization of the image beam. Brackets were made which corrected the orientation of the cold dumps, but no significant change in receiver performance was observed.
Receiver in DSB configuration, tuned to 225.658 GHz
Trec Ch1: 138 K
Trec Ch2: 122 K from a cold cal.
Started cool down. Vacuum only at 3.5 mTorr. This is higher than usual, but we wanted to see if the Channel 2 IF amplifier would start to oscillate at 200 K. If so, we would have more time to work on the problem. If the receiver worked, this level of vacuum should keep the receiver cold for the duration of the season.
There were some problems with the no. 2 sensor which is connected to the channel 1 mixer. It instantaneously jumped from about 60 K to 15 K and jumped back and forth a few times before returning to normal. It is currently operating, but will need to be looked into when the receiver comes back into the lab (the wires to the sensors will be cleaned up during the shut down any way).
Receiver reaching equilibrium temperature of 4 K.
Mark and Gene reconfigured the warm optics plate for DSB operation for both channels. This was done in the following manner (Instructions given by Tom Folkers):
The first spectra were taken and both channels seem to be operating. Receiver temperatures are around 170 & 150 Kelvin for channels 1 & 2 respectively.
Receiver “vital signs”:
Mark Metcalf and Gene Lauria arrived back on site. Receiver only at 200 K, so more time is needed for it to warm up before opening (it took another 3˝ hours to warm up). Channel 2 amplifier was turned back on and found to be still oscillating. Channel 1 is still operating.
Bob Stupak and Bill Peters found the baseline problem on Channel 2 to originate from LN2 spilling on the cold dump window. Bob now puts a piece of absorber underneath the cold load to prevent the LN2 from spilling on the window. This is a dangerous situation because the LN2 can embrittle the foam window and crack thus losing the vacuum integrity of the Dewar.
Receiver warm enough to open (coldest part of the receiver at 45° F). Mark and Gene broke the vacuum and started repairs.
Opened up Dewar and installed second IF amplifier with internal bias-“T”. The Ch. 2 IF amplifier was no longer oscillating. Found 2 cracked solder joints between the coax and connector on the stainless steel output cable, and on the input side of the IF amplifier with the cable between the isolator and IF amplifier. Mark also found that the input spark plug connector of the IF amplifier was loose. Upon removing the spark plug, found that the spark plug was never drilled to accommodate the set screw to lock the spark plug into place. Started the vacuum pump.
Channel 2 was having stability problems and found that the bias to the Channel 2 IF amplifier was unstable causing the IF amplifier to oscillate. Found to be inside the Dewar, and a warm-up will be required to inspect the bias wires. Gene and Mark will be returning on the 7th of June to inspect the wiring and install the other IF amplifier on loan from NRAO which has the internal bias-T.
During the power outage on Friday, the UPS failed when the observatory was put back onto commercial power. Smoke started to come out of the UPS when it was being bypassed, and was then shut down. The smoke detectors failed to trip during the event. This caused the receiver to warm up to 80 K. Power was restored to the receiver by finding receptacles that were not connected to the UPS. The Dewar was evacuated and cooled down again. The receiver got cold at around 20:30. The vacuum and temperatures upon our (Gene L. & George R.) departure were as follows:
5/27/05, 19:15 Operator: Bob Moulton
The compressor tripped off and the receiver warmed up today to about 68 Kelvin. The compressor tripped off to protect itself from overheating. (The outside doors were closed due to the weather). The compressor was off for at least 45 minutes, and the first and second stages warmed up to 80 and 40 Kelvin, respectively. The vacuum went up to as high as 19 milliTorr but recovered when the receiver cooled back down to 4 K. The receiver recovered, and observations started again that night. This event did not impact any observations because of the weather.
Performed DSB and some SSB measurements of the receiver. DSB measurements are as follows:
Verified mixer bias points, although Vj seems to want 0.1 mV higher than recorded in the lab. Warm optics adds a considerable amount of noise. The following temperatures were measured:
Have to investigate why the optics add so much noise to the receiver.
George, Martin and Gene left the observatory.
Cool down started at 7:45.
Made a circular plate out of 1/8" aluminum that fit into the opening of the cross grid. This facilitated the adjustment of the height of the receiver with the 12m optics plate. A laser was aimed from the other side of the elevation axis and the height of the receiver was adjusted so that the beam was centered on the plate.
Mixers got cold. Mixer 1 @ 4.3 K Mixer 2 @ 4.4 K
Martin adjusted the JT return pressure to lower the mixer temperatures. Temperatures now read:
Removed chirp transform spectrometer from Channel 1 backend IF cable.
Problems encountered with the Martin-Puplett servos:
Brought up to Mr. Graham. Support crew were: Tom Folkers, Martin McColl, George Reiland and Gene Lauria.
Since it was decided to use the 12m warm optics plate, the Dewar the cold dumps were moved over to the cold straps adjacent to each insert. The IR filters and vacuum windows were moved over as well.
Receiver was installed in the late afternoon, evacuation of the Dewar started at about 19:15 MST. Upon Martin's arrival 45 minutes later it was found that the turbo overheated and tripped off. Pump down started again at about 20:00.
Copyright Arizona Radio Observatory.