1. Cartridges / IF selection switch
a. Frequency range
(1) Band 1: 31.3 – 45 GHz
ii. Band 2: 67 – 90 GHz
iii. Band 3: 84 – 116 GHz
iv. Band 4: 125 – 163 GHz
v. Band 5: 163 – 211 GHz
vi. Band 6: 211 – 275 GHz
vii. Band 7: 275 – 370 Ghz request for 372 GHz
viii. Band 8: 385 – 500 GHz
ix. Band 9: 602 – 720 GHz
x. Band 10: 787 – 950 GHz
b. Noise temperature of the complete receiver average across full IF band
i. Band 1
(1) T(SSB) over 80%: 15 K (specification), 10 K (goal)
(2) T(SSB) at any freq.: 23 K (specification), 14 K (goal)
ii. Band 2
(1) T(SSB) over 80%: 28 K (specification), 16 K (goal)
(2) T(SSB) at any freq.: 43 K (specification), 24 K (goal)
iii. Band 3
(1) T(SSB) over 80%: 34 K (specification), 19 K (goal)
(2) T(SSB) at any freq.: 54 K (specification), 29 K (goal)
iv. Band 4
(1) T(SSB) over 80%: 47 K (specification), 26 K (goal)
(2) T(SSB) at any freq.: 76 K (specification), 40 K (goal)
v. Band 5
(1) T(SSB) over 80%: 60 K (specification), 32 K (goal)
(2) T(SSB) at any freq.: 98 K (specification), 51 K (goal)
vi. Band 6
(1) T(SSB) over 80%: 77 K (specification), 40 K (goal)
(2) T(SSB) at any freq.: 126 K (specification), 65 K (goal)
vii. Band 7
(1) T(SSB) over 80%: 133 K (specification), 69 K (goal)
(2) T(SSB) at any freq.: 198 K (specification), 133 K (goal)
viii. Band 8
(1) T(SSB) over 80%: 181 K (specification), 93 K (goal)
(2) T(SSB) at any freq.: 270 K (specification), 181 K (goal)
ix. Band 9
(1) T(SSB) over 80%: 335 K (specification), 202 K (goal)
(2) T(SSB) at any freq.: 500 K (specification), 301 K (goal)
x. Band 10
(1) T(SSB) over 80%: 438 K (specification), 351 K (goal)
(2) T(SSB) at any freq.: 655 K (specification), 525 K (goal)
xi. T(DSB) = 0.5 * T(SSB)
xii. Values are calculated according to T(SSB) = A*h*freq/k + 4K
(1) Bands 1-6: A = 6/10 (spec) 3/5 (goal) Tony Kerr argued for increasing this for bands 3 and 6. But DSB now on 9 rx on PdBI is 25-35 K DSB, this seems unreasonable.
(2) Bands 7-8: A = 8/12 (spec) 4/8 (goal)
(3) Band 9: A = 10/15 (spec) 6/9 (goal)
(4) Band 10: A = 10/15 (spec) 8/12 (goal)
c. FE passband amplitude and group delay
i. Flatness
ii. Ripple
iii. Slope
iv. Stability
(1) 1e-4 rms over 0.1 sec interval
(2) 1e-4 rms over 1 sec interval
i) Singledish 1 10-4 1s
ii) Interferometer 3 10-3 1min
iii) Diff 5 10-4 1 min meaning between two mixers same cartridge.
d. Sideband ratio/gain
i. Image suppression for 2SB and SSB mixers: 10 dB or better
ii. Gain accuracy should be 2-3dB within balance. More at edges.
e. IF frequency range for each polarization: 4 – 12 GHz
f. IF bandwidth per mixer
(1) SSB: 8 GHz (USB or LSB)
ii. DSB: 8 Ghz AB cannot commit to this for the time being
iii. 2SB: 4 GHz (USB and LSB)
g. IF power spectral density -40 +/- 3 dBm in any 2 Ghz
h. Phase stability. LD: 15 microns from antenna over what timescale? Band 10: 6.9 femtosec for (2 microns) all electronics; LO wants most of that. Everything at rf needs dimensional stability of on the order of a micron over tens of minutes calibration interval, possibly including tipping to calibrator. Suggest allocating 1 micron at Band 9. AB expects this to be a problem. This is an excess over anything occuring in all antennas. SG: I really don’t understand this 1 micron. LD: This comes from .1 rad in visibility, which gets divided between atmosphere, antenna, LO, and rx. CC: perhaps a subgroup should discuss this. MI: we measure this in ASTE frontend by vibration on front window.
i. Gain compression (saturation) per band Tony Kerr: measurement 0.5% going from LN to room temperature for four junctions at 230 Ghz.. .25% was in backend. Mixer preamp itself is .5%. This fits with his ALMA Memo on saturation. Lazareff: Band 7 presented to technical review board. They thought it a non-issue. Bernard thought 3% for his proposal. Kerr: points out that this is important if room temperature load used. SG: semitransparent vane can get around this. At higher freqs room temperature is similar to sky temperature. What is it at band 3? CC: What spec do we need to put on this? LD: What could we account for at the highest temperature we will see? Kerr: This can be measured in the lab, and we know the law we expect it to follow, so one number should be good enough. Effland: will write up memo; this can be measured to a precision or about .075%. AB: I did not say that. JE: Not sensitive to LO power. Not sensitive to LO freq. Other parameters to be determined. SG: can this system be implemented on the telescope? JE: It could. One just injects a large and small signal and examines the variation of the small one in the presence of the large one as it increases. <1% below band 6 and <3% band 7 and above.
j. Stabilization time after switch-on 15 minutes in Spec2.0
k. Monitor and control requirements. This falls more under ICD? Decided this is motherhood and housekeeping. Circulate list to other observatories for comment.
(1) IF switch attenuation
ii. IF power levels
iii. Cartridge temperature
iv. Mixer voltage and heater current if applicable.
v. HEMT gate/drain parameters
vi. Magnet currents
vii. Mixer-cycle to remove trapped flux