Query from Darrel Emerson and John Payne: These are specs that need to be defined by the systems group, with input from the science group (presumably ASAC). The systems group then needs to use the science input to define specs as they apply to the front end IF output, to the IF stages and so on. 1. In single dish mode, for one antenna, what is the maximum acceptable level of narrow-band (CW) spurious signal that can be tolerated in the passband. The spurious signals might be generated either in the front end (for example from local oscillator impurities) or in back end electronics. This figure probably needs to be specified in terms of receiver noise, within a given bandwidth: e.g. "not to be more than 1 mK within a 10 kHz bandwidth at frequencies below 100 GHz, and not more than 2 mK within a 100 kHz bandwidth at frequencies above 100 GHz." The numbers are probably not correct, and do other parameters (e.g. variability of the spurious features?) need to be included as well? It may be better to specify "0.001% of rx noise" rather than "1 mK." What is this spec? 2. The same question as (1) above, but for interferometer mode. 3. Is it now firmly established that only 4 antennas will be used for single dish continuum observing, in the cases where the maximum receiver stability (1:10^4 in 1 second) is required? 4. Will these 4 special antennas sometimes be used for interferometric observations too? 5. Will these same 4 antennas ever be used for single dish spectral line observing? For example, one could imagine that some special receiver design aimed at high stability continuum observing might not be suitable for spectral line observing. Would it be acceptable for these 4 special antennas never to be used for spectral line observing, in single dish or interferometric mode. 6. Although only the 4 special antennas will be used in single dish continuum mode for the cases where high receiver stability is required, is it true that all of the remaining 60 antennas (or even all 64 antennas) will sometimes be used for single dish spectral line observations? For example in such cases, the 60 (64) SD spectra from the individual dishes might be averaged together. 7. Local oscillator purity. Depending on band, the local oscillator is likely to include some spurious component, for example at twice the frequency, or (depending on the LO multiplication scheme) at a frequency such as 7/5 (i.e. the ratio of small integers) of the fundamental. What is the maximum acceptable level of such LO impurities, which will likely generate comparable receiver response (depending on rx band and lots of other things)? There are implications for calibration precision here. [Better than -20 dB may be difficult to achieve in some cases; would this be an acceptable goal? Would -15 dB be acceptable?] Charles sought a clarification on the interplay between nutator and receiver stability on the four total power antennas. Darrel responded: "To first order, the existence of a chopping secondary has no impact on the needed stability spec on the 4 special antennas. The needed stability was derived from consideration of the on-the-fly observing mode, which does NOT in general use a chopping secondary. Other observing modes, in particular pointing, will benefit from a chopping secondary. [It is possible to use a chopping secondary with on-the-fly, but that has other important and undesirable implications which I believe we want to avoid - for example, it would severely limit the scanning speed and hence the size of map that can be covered in a reasonable time. With a 10 Hz chopping rate, as you need to get about 3 sample points per beamwidth, it would limit the scanning rate to about 300 ms per beamwidth. The OTF stability spec was derived assuming 100 ms per beamwidth. On top of that, beam-switched OTF distorts the spatial frequency response in an undesirable way. It's possible to recalibrate for that, but with a serious loss of s/n at certain spatial frequencies, which translates to more needed observing time.]" AW responded: "I agree with your response to Charles' question. Receiver stability will also help to achieve polarization science goals, which will not be promoted as effectively by the nutator. There have been a number of recent email discussions in the science and calibration groups concerning the design of the production nutator. The science group will be refining that design over the coming weeks. The default, at the moment, is that on the prototype antennas. However, that design was optimized for antenna testing, not ALMA science. Thanks for the list of questions. The way the Science IPT works is that we will formulate a list of responses to the questions. The Project Scientist may then decide to pass these to the ASAC for comment. I'm not sure we can do this by Monday but we can certainly begin the Science IPT discussion (I've already started that)." Hi Al, I’m concerned about the approach we are taking on the LO system for ALMA. In this note I attempt to explain my concerns and I recommend testing a complete ALMA Band 6 prototype receiver as soon as possible with the Green Bank spectrometer. The reason I am writing to you on this subject is that Skip seems to be misinterpreting some information that he attributes to you and I ask that you correct this. In the interests of simplicity and reliability we have specified that mechanical tuners are not permitted in the Front End design. This approach forces us into a position that I am fearful will result in performance that is unacceptable to our customers. For example the old 12m telescope was used to detect spectral lines of a few milli-Kelvin in amplitude. This is with a system temperature of around 100 Kelvin using the position switching mode and single ended mixers. In these telescope receivers the local oscillator used was inevitably a tunable Gunn oscillator made by John Carlstrom. As the device was tunable a high Q cavity was involved that resulted in a rejection of any noise on the LO at the signal frequency. In a single ended mixer this noise would be indistinguishable from noise at the signal frequency. At the higher frequencies a Gunn was used with tunable multipliers made adjustable by moving back-shorts. It is instructive to calculate both the intensity and stability of any spurious signals that may accompany the LO A few assumptions are needed. Make the following: the minimum bandwidth is 100 KHz: the position switching time is 30 secs: the LO power required is 0.5 micro Watts: the bandwidth of the LO signal is 1 Hz and any spurious signals in the bandpass are stable to better than one part in one thousand over one minute( the usual position switching cycle). This latter point establishes the residual amplitude of any spurious signals in the I.F. passband to be around one Kelvin. Of course a crucial point here is the bandwidth of the spurious signal. If the same as the LO its intensity will be lessened by a factor of 10E5 by the resolution of the spectrometer. However in what follows that is almost down in the noise! The strength of the LO signal in Kelvin is around 10E16 so any spurious signal for this very simple scenario should be around 160dB down on the LO signal. Its worse than this for ALMA as the IF is very broad-about ten times wider than the old receiver so the chance of seeing spurious signals in the bandpass is higher. The next point is one that we should talk over with Tony. I notice that the original Virginia Diode quintupler has some response at *7 as well as *5. No doubt this is due, at least in part to the broad band nature of the device. Unless the response of the receiver is known at this unwanted frequency it is hard to see how a goal of 1% accuracy can be pursued with any chance of success. Anyway as a result of all the above I feel that the prototype system should be tested as completely as possible. The spectrometer in Green Bank will become available some time in the very near future and it turns out to have almost ideal characteristics for this test. Darrel feels as I do that its important to get an answer as soon as possible and its my judgment that this test is so important that it should be done as soon as possible even if a delay in schedule results. I have talked this over with Bob Wilson who raised the same concerns with NRAO people several years ago and feels that these fears were dismissed too lightly. What do you think about all this? John.