2010 June GBT KFPA Call for Proposals

The K-band Focal Plan Array Receiver (KFPA) was successfully commissioned in March and April 2010. Initial results of these tests are encouraging, and in this call, we invite KFPA shared risk proposals. During the commissioning observations, the KFPA was integrated with the GBT observing system and we performed a series of pointing, focus, gain calibration and mapping observations. The KFPA receiver specification is below 25 K for 75% of the band and below 35 K for the entire band, 18 to 26.5 GHz. During the commissioning tests, system performance looked good, with receiver temperatures for the different beams between 16 and 34 K, and system temperatures between 43.9 and 68.9 K at 66.7 degrees elevation (see below).

During summer 2010, the receiver will be reconfigured to yield better than 25 K receiver temperatures in the band center for all beams. We are hopeful that the receiver temperature in the center of the band with be better than 20K for all beams.

Observing modes

The KFPA is capable of operating in the same modes as current 2 beam K-band, except the widest separation in frequency of spectral windows (spectral bands) is 1800 MHz. We performed point and focus tests with all beams. The spectral line and mapping modes include:

  • 4 beams, two spectral windows, 2 polarizations, each with 12.5 or 50 MHz bandwidth
  • 4 beams, 1 spectral window, 2 polarizations, 200 or 800 MHz bandwidth
  • 4 beams, 2 spectral windows, 1 polarization, 200 or 800 MHz bandwidth
  • 7 Beams, one spectral window, 2 polarizations, each with 12.5 or 50 MHz bandwidth
    • Plus one additional (different) 12.5 or 50 MHz spectral window for one beam, 2 polarizations, within 1.8 GHz of the first spectral window.
  • 7 Beams, 1 spectral window, 1 polarization, 200 or 800 MHz bandwidth

tmcHc3nKfpa.png tmcHc5nKfpa.png tmcHnc3Kfpa.png
First line image, HC3N 18196 MHz, from the 4 beams, two spectral bands mode observations, with 1 hour total mapping duration.

This map was made scanning along lines of constant galactic latitude.

Second line image, HC5N 18639 MHz, from the 4 beams, two spectral bands mode observations.

The angular size of the GBT and NRAO 140ft telescope beams are shown for this observing frequency.

HNC3 18673 MHz image from 4 beams mode. This isomer of HC3N is detected in On-Off observations with 60 second duration, but is very weakly detected in the few second mapping observations. The GBT was configured so that the HC5N and HNC3 lines fell within the same spectral window.

Images in two modes: 7 Beams (+1) and 4 Beams, Two Spectral bands

  • 7 Beam, dual polarization, 50 MHz bandwidth, 4096 channel spectral line mode. Additionally 1 beam was configured for observations at an additional 50 MHz dual polarization, 4096 channel spectral line band. Ie NH3 1-1 and 2-2 lines for all 7 beams, plus NH34-4 lines using beam 1.
  • 4 Beam, dual polarization, dual 50 MHz bandwidth, 4096 channel spectral line mode. Ie HC7N J=16-15 (18048 MHz) and HC5N J=7-6 (18639 MHz) images produced simultaneously.

At the rest frequency of the NH3 1-1 line (23694 MHz), the channel resolution, 12200 Hz, corresponds to a velocity resolution of 0.15 km/sec. In both of these modes, we have tested a spectral integration time of 1 second per set of all beams, polarizations and spectral bands. We anticipate that for frequency switched observations, which generate twice as much data as position switched observations, the minimum integration time is 2 seconds.

Other spectrometer data recording modes are possible, but untested. It should be possible to observe in the 4 beam, dual polarization, 200 MHz or 800 MHz bandwidth spectrometer configuration.

We have checked the RMS noise level reached in a mapping observation. For the NH3 (1-1) line maps, we observed in the 4 beam mode, and had approximately 8 spectral integrations per beam. Averaging both polarizations and 2 channels and assuming a 45K system temperature, we predicted the RMS noise in the images to be 0.036K. In the actual average image, with some smoothing to reduce the map RMS noise due to striping in the gridded images, we measured an RMS noise of 0.048K. For the purpose of the proposal call, it might be reasonable to expect to reach 125% of the theoretical RMS noise level, for a given observing configuration and integration time (ie for a required sensitivity, increase the observing time by about 50% over theoretical).

moon1Beam.png moon7Beams.png moon7Smooth.png
Moon daisy map with 1 beam. These data were obtained in the 7 beams plus one mode. This 24138 MHz data was obtained only with beam 1.

The daisy pattern had only a 10 minute duration. The Moon's brightness was coarsely sampled. This image has a 40 arc minute diameter, and a full moon observation with 1 beam would take more than 2 hours.

Moon daisy map with all 7 beams at frequency 23706 +/- 25 MHz. The duration of the daisy pattern was too short to obtain a fully sampled map of the moon, but increasing the observing time to 25 minutes would have allowed making a complete lunar map. Moon daisy map with 7 beams and a larger, smoother convolving function. The dark patches to the north are where no data were obtained.

The north and south poles of the moon appear cooler than the equator. Note that the average temperature of the equator of the Moon, 200+/-5 K, which is close to the predicted value indicating the KFPA beams are reasonably well calibrated.

We have also tested observations made with the Daisy mapping mode, observing the Moon, and correcting for smearing caused by motion of the Moon during the observations.

In order to assure optimum GBT performance, typically an observer will schedule point and focus observations hourly. Following the point/focus observations, the observer is encouraged to perform an off -source (map reference) spectral line observation before and after all mapping scans. If a known bright spectral line source is near the target region, performing On/Off observations toward this source is useful for checking the spectrometer setup and confirming the Doppler tracking of the source is satisfactory.

The mapping observations are made on-the-fly, taking data while continually moving the antenna pointing center in a desired region. To avoid overly smearing the target source, the integration times should correspond to between 1/4 and 1/2 the time it takes the telescope to move 1 beam width. For observations at 23694 MHz, the beam size is 32" and the minimum integration time is 1 second, so the telescope maximum rate is 8 arc seconds/sec (or 8 arc minutes/minute or 0.13 degree/minute). The mapping scans can be separated by 1/2 a beam-width. With the current observing system, there is a 30 second setup time between mapping scans, so longer duration scans are preferred. Mapping a 0.25x0.25 degree region would require 56 scans, each with 2.5 minutes total duration. Since this duration is longer than the recommended 1 hour between point and focus observations, the map would be broken into two 0.25x0.125 degree images, with point/focus and calibration observations before observing each sub-image.

These observations can be processed with the KFPA calibration and imaging pipeline, which greatly speeds the data reduction process.

Mapping Example: 5' diameter circular map

When making small maps (< 10') it is more efficient to use the Daisy pattern mapping mode that allows mapping in a circular region, without stopping and restarting telescope motion. A suggested observing sequence is below:
  • On/Off Position switched observation (30 seconds at each location) of the target source and the reference position for the image.
  • 11 Minute Daisy pattern with 2' radius, 20 second duration to complete a single crossing of the region and a total duration of 660 seconds.
  • Second On/Off position switched observation (30 seconds at each location) of source and image reference position.

This observing procedure should yield approximately 0.1 K RMS noise in the 7' diameter region imaged (assuming good weather and source elevation '> 45 degrees). The outer portion of the image will have higher RMS noise than the central region. This sequence will require 15-20 minutes of
clock time. Three such regions could be imaged in an hour. Point and focus observations would require about 10 minutes clock time each hour.

orion1-122.png Spectrum of Orion A in NH3 1-1 and 2-2 lines,
produced from a 10 minute Daisy pattern map of the region (Scan 20 of Project TKFPA_16).
The left image is 2-2 line, the center image is the continuum image map
and right image is 1-1 emission.
The continuum image shows the image region, 20' in diameter, was not finely
enough sampled in this 10 minute observation. A 20 minute daisy scan could have fully sampled the central region of the image.

Performance of individual Beams of 7 Beam Receiver

The KFPA beams are configured in a hexagon shaped pattern, with number 1 the central beam. The receiver refrigerator placed to the side of the hexigon, and beams 2, 3 and 4 are closest to the refrigerator, and are coldest. Beams 1,2,3 and 4 are used in the four beam, dual frequency band mapping mode, as these have the lowest receiver temperature. The beam spacing is 94.88" and the maximum beam separation is twice this, 189.66".

Note that the KFPA can be placed in one of four orientations (see detailed description) , and if one of the other orientations is critical to the proposed science case, this could be requested during the call for proposals.
KFPA Orientation C, used for Commissioning tests.

Beam Polar- Trs Trx Tsys
Num. ization Specification Measured Measured
  (Left/Right) (K) (K) (K)
1 L 25.0 25.4 43.9
1 R 25.0 26.7 49.9
2 L 25.0 23.7 50.1
2 R 25.0 26.8 51.1
3 L 25.0 31.5 60.2
3 R 25.0 29.2 58.2
4 L 25.0 16.2 42.5
4 R 25.0 19.1 43.9
5 L 25.0 20.7 44.3
5 R 25.0 * *
6 L 25.0 34.2 65.2
6 R 25.0 33.3 68.9
7 L 25.0 29.7 61.2
7 R 25.0 26.1 50.1

Table of specified and measured receiver temperatures for each of the 7 beams and polarizations. A list of measured system temperatures is given for an observation at 66.7 degrees elevation. The values are taken at 23.5 GHz, near the NH3, 1-1 and 2-2 lines. The beam 5, polarization R amplifier was not functioning during these tests. The differences in receiver temperatures are due to differences in the thermal connections between the amplifiers and the refrigerator. In summer 2010, Beam 5 amplifier will be replaced and the thermal connections improved. Beam 4 is close to the refrigerator, and has the coldest amplifiers. We are hopeful that by winter 2010/2011, the performance of all beams will be similar to the measured beam 4 performance.

The system temperature measurements were tabulated during calibration observations of 3C48. The observations were made at 66.7 degrees elevation, on a day when the zenith opacity (tau) was 0.048. The expected atmospheric contribution to the system temperature is 15.8 K, approximately what was observed.

KFPA beam 4 Receiver temperatures (Trx) for both polarizations as a function of frequency (top two curves). The bottom two curves are measures of the equivalent temperatures of the two calibration noise diodes, as a function of frequency. kfpaBeam4Trx.png

Please contact Glen Langston (glangsto@nrao.edu, 304-456-2224) if you have questions about the KFPA configurations.

-- GlenLangston - 2010-05-10
Topic revision: r16 - 2016-06-08, PatrickMurphy
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback