This is the view of the Smithsonian Submillimeter Array on Maunakea while in the subcompact configuration. Note that 7 of the 8 antennas are visible in this photo with the 8th out of view to the left. The 3 other common configurations are compact, extended, and very extended where the antenna spacings extend to 500 meters to provide higher angular resolution at the cost of lower sensitivity. The altitude of the array is approximately 13,400 feet (4,085 meters). This photo was taken with my iPhone 6 in 2018.
I was actually an SAO employee between 1999 and 2000 and later supported the SMA full-time as an ASIAA employee from around 2004 - 2009. I'm still involved with the SMA today but at a lower level of effort.
2020-Mar - This integration effort was organized by Paul Y. with Ryan and myself lending a helping hand for a few days. It was a culmination of a number of new designs and upgrades described in this section which increased the total analog processing bandwidth from our initial 4 GHz per antenna to our present 6-quadrant 24 GHz per antenna (P. Yamaguchi, R. Chilson, D. Kubo, et al).
2020-Mar - This photo shows two different versions of Block Down Converter (BDC) units developed by ASIAA for the SMA bandwidth expansion. I provided the design and parts procurement for the upper 8-10 GHz unit and most of the assembly and test performed by Ryan and John K. (D. Kubo, R. Chilson, J. Kuroda).
The lower BDC unit supports 10-12 GHz and the design, assembly, and test was performed by Ryan (R. Chilson).
2019-Sep - There are a total of 4 BDC unit types, 4-6 & 6-8 GHz developed by SAO, 8-10 GHz, 10-12 GHz by by ASIAA, and 12-14 & 14-16 GHz subcontracted to Norden Millimeter.
These 4 IF Distribution Plates were developed in our ASIAA Hilo office and perform the function of distributing each of the 16 IF signals (2 per antenna) to the 4 BDC unit types. Each IF path supports 2-18 GHz leaving room for an additional 4 GHz expansion (D. Kubo, R. Chilson).
Simplified system diagram of IF signal path:
2016 - Development of 10-12 GHz BDC units were largely copy build of the 8-10 GHz units deployed ~2 years earlier. Ryan performed 100% of this work from parts procurement, mechanical redesign, to assembly and test. There were some minor improvements included in this version. A big thanks to Ryan for single handedly taking on this task and delivering these two units (R. Chilson).
2016 - Photo of the internal IF processing brackets for the two 10-12 GHz BDC units developed for the project (R. Chilson).
Interesting fact - unbeknownst to both Ryan and I, we both worked at Northrop Grumman Corporation in Redondo Beach, CA, in the same R6 building no less. I was a contractor at the time and he was a NGC employee.
2014 - All of the BDC units require software monitor and control for converging the AGCs to maintain optimal SNR, linearity, and drive levels into the backend ADCs. Here is a photo of Ryan working on the 8-10 GHz BDC at Maunakea with Ranjani, remote from Hilo, who was/is responsible for generating and integrating the software servo control.
Note the temporary rack in the rear which we used to house the first pair of 8-10 GHz BDC units to support the initial deployment of SWARM's first quadrant (R. Srinivasan, R. Chilson, D. Kubo).
2013 - First of two completed 8-10 GHz BDC assemblies for the SMA Wideband effort. Two units were constructed to support the16 IFs from the antennas. Each IF is signal conditioned with AGC and filtering, down converted to baseband, followed by a second level of AGC and Nyquist filtering to provide an optimal signal to the ADCs for digitization (D. Kubo, R. Chilson).
Description of the unit on the ASIAA website:
2012 - Here's a photo of the very early stage of the BDC unit development in my office. Mechanical packaging was based on earlier work I did for the ALMA ALS project two years earlier. I started with a 5U Par-Metal EMI chassis along with a custom 1/4" thick aluminum deck plate with threaded M3 holes spaced by 20 mm. DC supplies consist of low noise Daitron units that can supply 15V at 10A with 10 mV noise ripple. Both top and bottom were populated with electronics to utilize the volume (D. Kubo, R. Chilson).
Electrical design package:
2016 - This unit provides 2-18 GHz of noise and has the option of switching in an external tone source. The wide-band noise is distributed via a 4-way power divider to four Noise Distribution units
I started this effort several years earlier with help from John K. but deferred its completion to support other projects. The SMA was in need of this unit for calibration of the digital correlator so Paul Y. (CFA) stepped up and completed it in short order. You'll notice that the unit in this photo has a rear panel with fans but I ended up using it on the LO Photonics Receiver unit for the GLT project. (J. Kuroda, P. Yamaguchi, D. Kubo)
Diagram of noise calibration system:
2014 - These 4 noise distribution plates were designed and assembled by John K. who now works for EAO across the street from us. He did a beautiful job which allowed me to removed the existing legacy hardware which supported 4-6 GHz and drop this assembly in and use the same mounting holes. With this modification, the entire noise calibration path from Noise/Tone Generator to the C1DC units process a noise bandwidth of 2-18 GHz (J. Kuroda, D. Kubo)
Photo of 1 of 4 chassis with plate installed:
Swept response of existing noise cables:
2013 - The Correlator 1st Down Converter (C1DC) unit was originally designed to received the optical 4-6 GHz IF signal from the antenna and channelize then down convert to six separate 1 GHz +/- 164 MHz blocks.
16 (+ 2 spare) units were modified to support 2-16 GHz and is visible as the aluminum plate with the black fan near the bottom of the photo. The modifications were performed serially one at a time replacing an unmodified unit with a modified one over a period of two years. The down conversion portion is obsolete and is replaced by the BDC units. Electrical design/parts procurement - myself; mechanical design - John K.; assembly and test - Solomon/John K.
Test data for S/N 01 and 17:
2012 - Each of the 8 antennas has an IF/LO Hoffman enclosure that contains 5 subassemblies that were cooled primarily though conduction and radiation. The 2 IF subassemblies were modified for increased bandwidth and encountered thermal problems. This thermal problem was solved by adding forced air convection cooling with the upper fan drawing in the stabilized cabin air (typically 16 +/-1 C) and the lower fan exhausting the warm air out near the floor. We added 4 addition internal fans to circulate the air in a CCW direction to reduce local hot spots. Note the metal baffle over the top fan prevents direct convection cooling of the upper subassembly modules. All 8 antennas were retrofitted with these modified covers (D. Kubo, P. Yamaguchi, B. Koga).
2006 - This fiber switch was developed to support extended SMA (eSMA) which replaced SMA antennas 5 and 6 (receiver-B) IF/LO fiber-B with JCMT and CSO antennas, respectively. The rationale being that replacing two SMA 6m antennas with 15m and 10.4m would increase the array sensitivity. The design was based around a pair of Newport single-mode fiber switches shown in black at bottom of photo, and were controlled by an ADAM TCP/IP controller with auxiliary circuitry. eSMA is now defunct, however, the JCMT occasionally uses the fiber switch to phase up with the SMA for VLBI fringe verification (D. Kubo, T. Hunter).
Description on the ASIAA website:
Design and operation memo:
2006 - I took this photo with a Canon PowerShot ELPH from Puu Poliahu with Paul Y. and Tomas K. (now at Keck). It was just after a heavy snow fall of about 2 feet and we used snow shoes to hike up to the side of Poliahu. This photo has been used in a number of ASIAA publications. I've been meaning to go up and take another photo after a snow storm but haven't done it.
Link to photo (if you use it please credit the source... and um that would be me):
2020 - Gratuitous selfie to fill space. I share this lab space with Paul Y. and am very thankful that he is a neat and organized person. And this feeling is probably not reciprocal because I'm more often than not in a rush and don't put things away at the end of the day. But those sideways glances are getting to me and has motivated me to be a better lab partner.