CCB Test Results
15sep10 in Lab - with the new 90 degree hybrids in place.
Data are in /home/simdata/TCCB15SEP10
scan |
notes |
3 |
hot with cal |
4 |
blank (HEMTS debiased): zero level |
5 |
hot w/cal |
6 |
cold w/cal |
7 |
cold w/cal |
8 |
long cold, no cals |
9 |
long cold, no cals |
10 |
cold w/cal |
11 |
hot w/cal |
12 |
long hot, no cal |
Everything functions as predicted & to spec. Gains are in the range 31 - 36 counts/Kelvin, which puts them saturating just over 400K (aside from offsets), which is right. Taking into account the data zero levels (HEMT bias OFF) using the first HOT and COLD scans, here are the Tsys's and cal diode equivalent temperatures I get --
channel |
freq. |
Tsys |
Tcal(A) |
Tcal(B) |
9 |
38.25 |
23.5 |
5.47 |
5.35 |
10 |
35.75 |
18.3 |
7.99 |
7.01 |
11 |
32.25 |
17.7 |
10.03 |
8.04 |
12 |
28.75 |
22.2 |
7.03 |
4.79 |
13 |
38.25 |
20.8 |
5.45 |
5.33 |
14 |
35.75 |
16.4 |
7.86 |
6.98 |
15 |
32.25 |
15.3 |
9.89 |
7.91 |
16 |
28.75 |
20.0 |
6.95 |
4.69 |
Below is the power spectrum of the data from one of the long cold load scans. Notes & explanation:
- the white line is the power spectrum of the total power data
- magenta is the power spectrum of the differenced data
- the dashed white line is the expected theoretical noise given the measured system temperature (overestimated by 5-10% due DC offsets in the data, but still indicative)
- Overall the noise performance is excellent (i.e., consistent with what is expected for the differenced data). There is some 1/f and 60 Hz in the total power data but it is switched out.
- the first 8 channels are not connected to the receiver, hence the very low noise levels (showing post-receiver electronic noise).
- if you were to make the same plots for the hot load scans: a couple integrations (because of the noise scatter in integrations) saturate for ports J9, J10, and J14. This is not a performance issue since such integrations get flagged automatically. It does however throw off my power spectrum software, which requires regular samples.
--
BrianMason - 2010-09-15