Reviews Proposal 517 Complex ADC for Increased Bandwidth and Backend Simplification
Title: Proposal 517, Yeste, Complex ADC for Increased Bandwidth and Backend Simplification, Omar Yeste Ojeda
Grade: 2.5 (0.1 to 9.9, 0.1 is best)
This proposal seeks to evalulate high bandwidth ADC's for ALMA, and test a pair of ADC's in
quadrature sampling mode, utilizing a 90 degree hybrid (complex sampling instead of real).
They propose to develop and test a 16 GHz bandwidth sampler (16 GHz per polarization and per sideband)
using a pair of ADC's and an off the shelf 90 degree hybrid made by Marki.
They proposers have a strong team, with an excellent track record,
but I'm concerned that a lot of their proposed work has or is replicated and already funded:
Jonathan Weintroub and Bob Wilson have a new NSF MRI grant to test and evaluate several different high bandwidth ADC's for SMA and ALMA.
Weintroub and Wilson plan to deploy these high bandwidth ADC's on SMA, but I expect their tests will be useful for ALMA as well.
They have already tested the analog devices (formerly Hittite) ADC that the proposers want to test -
this ADC has several problems and analog devices has no plans to revise the chip.
I don't much purpose of another group testing the same ADC.
Weintroub and Wilson will also be testing Vadatech and Adsantec ADC's, which the proposers don't mention.
One relatively new concept the proposers want to do is to implement a pair of ADC's with a 90 degree hybrid,
and then digitally calibrate out the phase and amplitude imbalance induced by the hybrid.
This is probably different than what Weintroub and Wilson will do for their MRI grant and SMA.
But this digital calibration technique has already been implemented and well documented. (Glenn Jones Thesis, Ricardo Finger et al,...).
Both Jones and Finger have shown that digital correction of phase and amplitude imbalance works well.
I don't see much need to test this again.
1. Alignment with NA ALMA Partnership strategic goals: good alignment
3. Quality of the upgrade conceptual design: fine
4. Readiness for production in the context of the ALMA Development Plan: ready
6. Qualifications of the key personnel of the Study: excellent
7. Technical expertise, past experience: strong expertise and past
8. Assessment of the level of risk inherent in the design: low risk
9. Strength of the Scientific Team supporting the Study:
this is not science - it's engineering.
the engineering team is excellent.
11. Budgeted cost of the Study;
The budget seems a bit high for the limited work they propose,
especially high considering some of the work has been done by others,
and most will be done by others whether this proposal is funded or not.
Complex ADCs for Increased Bandwidth and Backend simplification
1. Alignment with NA ALMA Partnership strategic goals;
Proposal is well aligned with ALMA strategic goals, for increasing the bandwidth.
2. Strength of the scientific case for the proposed ALMA upgrade concept; Comment on the relevance to the ALMA 2030 development documents.
Proposal is to study the quadrature hybrid sampling method for high speed ADCs to achieve high observing bandwidth. Single high
speed ADCs often have multiple cores, each clocked with phase offsets. Here the idea is to have a quadrature hybrid and two
3. Quality of the upgrade conceptual design;
Proposal lacks details about the ADCs- only cursory reference is made to an Analog Devices ADC and There are potential candidates
in the market
with price quotes, on page 5, but no specific devices are listed. To the knowledge of this reviewer- there are perhaps
about half a dozen high speed (5 to 20 Gsps) ADCs available- most of these not yet ready to use- i.e., they are just chips as
experimental devices not yet released in the market.
4. Readiness for production in the context of the ALMA Development
Plan (the aim is to support a range of upgrades including both those
which can be implemented rapidly and those requiring longer-term
research and development);
Implementation for ALMA is discussed in section 6.0 - there is
readiness with minimal changes, but the main item is the ADCs
themselves (see Risks below).
5. Strength of the consortium organization (if applicable);
All investigators are from NRAO
6. Qualifications of the key personnel of the Study;
The PI seems well qualified for this project, but no CVs provided
for any of the co-Is
7. Technical expertise, past experience (also in series production,
if relevant) and technical facilities in the Institutes taking part
in the Study;
Sufficient technical expertise on the team, and at the institute
8. Assessment of the level of risk inherent in the design;
Availability of suitable hardware for this study, particularly the
ADCs. Also, actual performance of the ADC may turn out to be far
9. Strength of the Scientific Team supporting the Study;
10. Level of support guaranteed by the Institutes;
NRAO (although, signed by the PI!)
11. Budgeted cost of the Study;
Cost of $397,574 for a one year study seems high for this project.
Proposal # 517: Complex Analog-to-Digital Coverters for Increased Bandwidth and Backend Simplification
This proposal seeks to develop electronics that will form the basis of a future ALMA correlator, in conjunction with the ALMA suite of receivers, that will have 16 GHz of usable bandwidth. The concept lies in using a quadrature hybrid that will sample both the real and imaginary parts of the analog signal from a receiver and then pass them on to a digitizer platform, which will be purchased from industry. This architecture simplifies the electronics and minimizes the sampling rate, thus effectively increasing the usable bandwidth.
The science rationale behind this proposal is well-justified. Being able to observe 16 GHz of bandwidth simultaneously has many advantages, including flexibility of line selection and increased speed of spectral-line surveys. However, the project relies on the availability of a high speed digitizer with a sampling rate of 40 Gs/sec. Although such a digitizer exists in the oscilloscope industry, it is not clear how readily available they are in reality. Note that the firm e2v has a prototype working at 16 Gs/sec, and Curtiss-Wright has one operating at 12 Gs/sec; these are about the best there are at the moment that are commercially available. The proposal also states that an FPGA is a plus in the digitizer platform, but that is an understatement. An FPGA is an absolute necessity if the system is to function as a spectrometer.
There is some discussion of how other observatories have spectrometers with 8 GHz of band width. However, the SMA stitches together 8 GHz with 2 GHz segments. The NOEMA spectrometer uses a polyphase filter to divide the super-fast data stream into 64 slower data streams, each of which can be processed with much simpler, non-parallel FFTs. These 64 smaller FFTs then must be combined together. Furthermore, it is not clear when any ALMA receiver will have 16 GHz of usable IF bandwidth per sideband. As of now, the receivers have 4 GHz of IF per sideband per polarization. Increasing the bandwidth to 16 GHz, as the proposal describes, is not trivial. Currently, there are not suitable low noise IF amplifiers with 8 GHz bandwidth, let alone 16 GHz, and mixer performance with these bandwidths is uncertain. In this sense the proposed work seems premature.
Reviews not received.