LRD 1/26/98 (v1)
09:00 MST (16:00 UTC), January 21st 1998
Present (via teleconference):
This meeting focussed entirely on cryogenics. A list of issues to be discussed was circulated in advance, and the meeting followed that list. The following summarizes the discussion of each issue. Also circulated in advance were two background documents; those are reproduced at the end of these minutes.
A straw-man set of specs requiring 0.3W of cooling at 4.0K with 10mK stability, all at the SIS mixer block mounting surface, was considered (see Attachment A). No one thought that this would not be satisfactory, except that Tony Kerr and S-K Pan indicated that colder would be better; we do not have enough information to evaluate the tradeoff of receiver performance against cryogenics construction and operating cost. It was pointed out that achieving 4.0K at the mixer will require somewhat lower temperatures at the expander. Dick Plambeck recommended that we allow for active temperature stabilization in order to achieve the 10mK (or better) stability, regardless of whether the final stage uses continuous-flow (J-T) or cyclic (e.g., GM) refrigeration.
To minimize the 4K load, most radiation and conduction from outside must be intercepted at warmer stages. Estimates of the resulting load are presently very rough, as are estimates of the residual load at 4K due to imperfect interception. Guidance based on experience at other observatories was solicited; John Payne suggested that Pat Sikes at ATNF has useful data, and after the meeting Charles Cunningham sent some JCMT data (LHe system) via email.
A possible refrigerator arrangement (although currently not the favored one) would involve only two stages of cooling and hence only two cold heat sinks, with the first being near 40 or 50K and the second at 4K. Since the second stage capacity must be relatively small, most dissipating loads (especially HEMT amplifiers) would need to be on the first stage. It was Marian Pospieszalski's feeling that the noise temperature penalty at 40K (vs. 15K) would be unacceptable for the HEMT receivers, although no quantitative study of the tradeoffs has been done. One option is to implement only one or two stages of amplification at 4K, with the rest at the warm stage (see Attachment B); Marian pointed out the extra cost of construction due to additional amplifier blocks, and possible difficulties in interstage coupling if 30-50 cm of separation is needed. Note that the SIS receivers are already expected to have split IF amplifiers, with 1-2 stages integrated with the mixers, so there should be no impact on those receivers. In addition, it is assumed that all HEMT amplifiers not in use will have zero bias applied, resulting in (typically) four simultaneously active channels (out of 8*4+2*2=36 channels total). No one questioned these assumptions.
My conclusion from the discussion is that this configuration (where there is no 15-20K stage) is considered undesirable but perhaps acceptable if it yields a significant advantage in power consumption or reliability.
John Payne quoted JPL's experience with CTI refrigerators as showing that operation close to rated capacity results in shortened lifetime. Larry D'Addario agreed to seek quantitative information directly from JPL. Dick Plambeck speculated that this may be just an artifact of bookkeeping, in that small degradations at capacity may be much more noticable, and thus more likely to be counted as "failures," than at light loading. There seems to be no reason to suspect a relation between loading and such maintenance problems as seal wear or regenerator deterioration.
This was strongly favored by Marian, on the grounds that it allows each type of receiver to be freed from design constraints imposed by the presence of the other. There was some discussion of whether there would be any reliability improvement and of whether servicing of each dewar would be easier. It was clear that the total power consumption will be higher for any 2-dewar arrangement than for a 1-dewar arrangement, and that more hardware will be required. No clear conclusion could be reached in the absence of a more detailed design, including plans for receiver servicing during operation.
LRD 971121, rev 980121
Total of 10 separate receivers on different bands: 3 HEMT amplifier + 7 SIS mixer, or 2 HEMT amplifier + 8 SIS mixer One feed horn per receiver, cold. HEMT receivers: Two channels (both polarizations); about 5 stages/chan. SIS mixer receivers (only one active at a time): Two mixers (both polarizations) Four IF channels (both polarizations, both sidebands) Two IF amplifier stages per channel integrated w/ mixers REFRIGERATION STAGES: 4.0K nominal with good stability: 10mK p-p in 1 minute 100mK p-p in 1 day 4.5K maximum before maintenance; 1 year desired. "Warm" stages: one or two stages at temperatures TBD, somewhere between 15K and 80K. REFRIGERATION LOADS: Very rough guesses, from experience Heat Source Warm Stage 4K Stage ------------------------------------- ---------- --------- Radiation shields, intercepting 300K 10 W 0 Radiation, 80K shields to 4K 16 mW Windows: leakage past IR blocks *50 mW Electrical dissipation: HEMT amplifiers - 2 rcvrs x 2 chans 2 stages 4K @5mW 40 mW 3 stages warm @10mW 0.12 W SIS IF amplifiers - 1 rcvr, 4 chans 2 stages 4K @5mW 40 mW 3 stages warm @10mW 0.12 W LO multipliers or photo mixers 0.5 W 0 Conduction: Waveguides, coax *2.0 W 50 mW Conduction: JT circuit parts, if any 1.5 W 10 mW Miscellaneous, margin 1.0 W 50 mW -------- ------ TOTALS 15.24 W 256 mW ======== ====== * These estimates may be subject to especially large errors. All numbers are rather uncertain at this time.
TWO STAGES 1st stage ~50K ~20W Radiation shields Cable and WG heat sinks HEMT amplifiers, RF and IF, stages 3--5 (x36) LO multipliers or photodiodes HEMT receiver mixers 2nd stage 4.0K ~0.3W All feedhorns (10 ea.) SIS mixers (7 or 8 ea.) HEMT amplifiers, RF and IF, stages 1 and 2 only (x36). THREE STAGES 1st stage ~75K ~20W Radiation shields Cable and WG heat sinks LO multipliers or photodiodes HEMT receiver mixers 2nd stage ~15K ~ 1W Cable and WG heat sinks HEMT receiver feed horns (2ea) HEMT amplifiers, RF, 5 stages (x4) HEMT amplifiers, IF, 3 stages (x32) 3rd stage 4.0K ~0.3W SIS mixers (8 ea.) SIS receiver feed horns (8 ea.) HEMT amplfiers, IF, stages 1 and 2 only (x32)
Larry D'Addario, 1/26/98