Gopal Narayanan University of Massachusetts, Amherst

1 of 43

Overview of the Large Millimeter Overview of the Large Millimeter TelescopeTelescope

Gopal NarayananGopal NarayananUniversity of Massachusetts, AmherstUniversity of Massachusetts, Amherst

2 of 43

Overview

● LMT Status● Brief Outline of Scientific Results● Summary of Instruments

– 3mm Redshift Search Receiver (RSR)

– 1mm AzTEC Continuum Camera

– SEQUOIA 3mm Focal-Plane Array

– Phased Focal Plane Array (PFPA) for GBT & LMT

– 1mm VLBI Receiver & One Millimeter Array Receiver

3 of 43

About the LMT....● 50m diameter millimeter-wave telescope

– 75 microns RMS surface accuracy

– 1 arcsec relative pointing

– Active Primary Surface

– Focal Plane Array Instrumentation

● Binational Collaboration with Mexico

Site in state of Puebla, MX 15,000' Seasonal site: Submm

winter; 3mm summer.

4 of 43

LMT Employs Microwave Holography to Set SurfaceUse 12 GHz beacons from geo-stationary satellites to construct complex far-field map. Fourier transform to get aperture plane phase → maps to surface errors!

5 of 43

The LMT ScoreboardSpecification Desired Achieved (2014)

Diameter 50m 32m

Surface Accuracy Active Surface RMS (microns) Elevation Range (deg)

Yes75

15 – 85

Yes60

25 – 80

Pointing Accuracy Absolute (arcsec) Relative (arcsec)

N/A< 1

31.5

6 of 43

The LMT ScoreboardSpecification Desired Achieved (2014)

Diameter 50m 32m

Surface Accuracy Active Surface RMS (microns) Elevation Range (deg)

Yes75

15 – 85

Yes60

25 – 80

Pointing Accuracy Absolute (arcsec) Relative (arcsec)

N/A< 1

31.5

Two Instruments already in place: AzTEC: A 1.1 mm wavelength continuum camera Redshift Search Receiver (RSR) – 3mm wavelength ultra-wideband

(38 GHz) spectrometer

Third season of “Early Science” Observations starting within a week

7 of 43

AzTEC at the LMT

● 144-pixel bolometer array

● 1.1mm Wavelength● 8” HPBW● Lissajou “On-the-Fly”

Mapping

8 of 43

AzTEC Image of M87

Image Courtesy: G. Wilson

9 of 43 Image Courtesy: D. Sanchez

10 of 43

Redshift Search Receiver

● 73 – 111 GHz– 38 GHz Bandwidth

– 31 MHz Resolution

● Dual Polarization, Dual Beam Receiver

● Excellent baselines● High Sensitivity

– Trx ~ 75K

– Tsys ~ 100K

Arp 220

One Hour IntegrationNo Baselines Removed

Erickson, Narayanan et al (2007)

11 of 43

Arp 220

Image Courtesy: P. Schloerb

12 of 43

Molecule Transition Frequency Amplitude Error SNR(GHz) (mJy) (mJy)

ch3cn 6-5 110.38 17.0 1.4 12.413co 1-0 110.20 29.4 1.4 21.5

5(0,5)-4(0,4) 109.91 9.4 1.5 6.3c18o 1-0 109.78 23.9 1.5 16.1

12-11 109.60 10.7 1.4 7.612-11 109.44 13.2 1.4 9.312-11 109.17 42.2 1.9 22.011-10 100.47 6.3 1.5 4.111-10 100.32 1.6 1.6 1.011-10 100.08 28.1 1.4 20.4

cs 2-1 97.98 25.6 1.4 17.9ch3oh 2(0,2)-1(0,1) 96.74 5.5 1.4 3.9c34s 2-1 96.41 7.8 1.4 5.5n2h+ 1-0 93.17 21.2 1.5 14.1ch3cn 5-4 91.98 12.8 1.5 8.6

10-9 91.33 6.6 1.5 4.410-9 91.20 4.8 1.5 3.210-9 90.98 31.7 1.5 21.61-0 90.66 70.1 2.0 34.31-0 89.19 36.4 1.5 24.31-0 88.63 83.2 1.5 55.1

4(0,4)-3(0,3) 87.93 6.3 1.6 4.1c2h 1-0 1/2-1/2 87.40 11.9 1.8 6.6c2h 1-0 3/2-1/2 87.32 24.5 1.6 15.7

2-1 86.85 13.8 1.8 7.7h13co+ 1-0 86.75 4.7 1.7 2.8h13cn 1-0 86.34 13.9 1.6 9.0

9-8 82.20 2.9 1.6 1.89-8 82.08 4.2 1.6 2.79-8 81.88 27.9 1.6 18.0

hnco

hcccn v7hcccn v7hcccnhcccn v7hcccn v7hcccn

hcccn v7hcccn v7hcccnhnchco+hcnhnco

sio

hcccn v7hcccn v7hcccn

Molecular Line Detections in Arp 220

13 of 43

LMT's Killer AppGalaxy Evolution

Williams et al. 1997

Hughes et al. 1998

Submm View Optical View

14 of 43

What are the “Submm Galaxies”?ULIRGs at high redshift?

Ultraluminous infrared galaxies (ULIRGs):

Dusty, LIR>1012L

Solar

High rates of Star FormationGas rich interacting/merging

Spectral energy distribution: strong peak near 100 micron from warm dust

Peak gets redshifted to 1 mm by z~10 Yun 2000

Surace, Sanders, & Evans 1999

15 of 43

Observations of Distant Galaxies

● Continuum flux from dusty starbursts is almost independent of redshift

● In some models and cosmologies, objects get brighter with z.

TODAY'sLMT 5-sigmaIn 10 sqr arcminMap with 2 hoursIntegration time

0.55.9 2.212.4 10.313.3Age of Universe (GYr)

13.5

ZRedshift

16 of 43

9io9 (z=2.54)

AzTEC (8”)

9io9 identified as lensed object by volunteers in Zooniverse Project!

17 of 43

PJ105353 (z=3.00)

AzTEC (8”)

18 of 43

HJ114637 (z=3.26)

Fu et al. (2012)

Red “arcs” are lensed images of the z=3.26 galaxy

19 of 43

First VLBI Fringes

The Future:

Make an image of the Supermassive Black Holeat the center of the Galaxywith the Event Horizon Telescope

20 of 43

The LMT is an important hub of the EHT

21 of 43

Recap● The LMT works and is doing scientific work!

● Mexico committed to completing the full 50m antenna.

– Assuming timely allocation of funds, expect full 50m will be complete in 2016.

● Next generation of instruments coming online:

– SEQUOIA – 16-pixel 3mm heterodyne focal plane array equipped with ROACH-2 based backend spectrometers

– A 1-mm single-pixel SIS receiver for VLBI observations

22 of 43

● New Instruments

– Phased Focal Plane Array at 3mm wavelength

– Heterodyne Focal Plane array at 1mm wavelength

– Follow-up to AzTEC – a very large format mult-color TolTEC instrument

23 of 43

WARES (Wideband, Array, Roach Enabled Spectrometer)

● SEQUOIA – 16 pixel 85 – 115 GHz focal-plane array using MMIC LNA Amplifiers

● Used at the Quabbin 14m radio telescope for several years

● To be installed on the LMT for the 2014-15 season

● Will need new optics to interface with the LMT, and a new spectrometer

24 of 43

Wideband Spectrometer System for SEQUOIA

25 of 43

Wideband Spectrometer Modes

Mode BW (MHz) NumChannels Resolution (kHz)

Velocity Resolution (km/s)

1 800 2048 390 1.2

2 400 4096 97.7 0.3

3 200 8192 24.4 0.07

26 of 43

Wideband Spectrometer Modes

Mode BW (MHz) NumChannels Resolution (kHz)

Velocity Resolution (km/s)

1 800 2048 390 1.2

2 400 4096 97.7 0.3

3 200 8192 24.4 0.07

Future NB Spectrometer Based on LEDA 16-channel ADC using 2 (or 4) ROACH-2 boards

Mode BW (MHz) NumChannels Resolution (kHz)

Velocity Resolution (km/s)

1 50 4096 12.2 0.037

2 25 4096 6.1 0.018

3 12.5 4096 3 0.009

27 of 43

Phased Focal Plane Array for 70 – 95 GHz

Neal Erickson, Gopal Narayanan, Joe Bardin

Univ Massachusetts, Amherst

Brian Jeffs and Karl Warnick

Brigham Young Univ

Karen O’Neil

NRAO, GBT

28 of 43

Conventional Focal Plane Array

• Large gaps between elements, 2x HPBW spacing.

• No ability to optimally illuminate telescope, particularly over a wide bandwidth.

• Relatively few elements fit into focal plane.

• Difficult to adapt to varying f ratios.

• Physically large.

29 of 43

Phased Focal Plane Arraysomewhat idealized

• Very small feeds, spaced ~/2, are able to collect all of the signal in the focal plane.

• Combine feeds into groups that synthesize the optimum illumination. Requires amplification before combination.

• Correct for large scale telescope surface errors.

• Easily adapted to any telescope f ratio.

• Physically very small.

30 of 43

PFPA Goals

• Build a 64 element receiver.

• Demonstrate at the GBT that this is a viable mm-wave technology.

• RF bandwidth 70-95 GHz. Excellent band for GBT science. SiO, HCN, DCN, HNC, HCO+

• IF bandwidth will be ~50-100 MHz.

• Beam-forming is all digital (no analog summing).

• Budget is very limited, $1.3M, 3 years.

31 of 43

Waveguide Feed Element

Machine the entire feed array including splitter from a single piece of aluminum.

32 of 43

Array ArchitectureOne MMIC gain stage at 20K, 2nd stage at room temp.

33 of 43

First Stage Amplifier Circuit Board

Separate drain bias for each element.

34 of 43

Beam Forming

General process is very much like interferometry.

1. Form complex spectrum from each element.

2. Vector sum elements with predetermined weights.

•. Real-time summing network is very complex, 169 summing engines each with up to 25 inputs

•. If we allow for arbitrary focal plane distribution, then each beam requires all 64 inputs

•. Record and post-process may be needed

35 of 43

Beam Forming

• All signal processing uses ROACH boards with 64 channel ADC board

• Our budget is 8 boards, which may greatly limit the bandwidth (and resolution) we can process.

• ROACH2 boards much more capable.• Interconnections in summing network require fast

ethernet switch.

36 of 43

1mm Receivers

● For upcoming VLBI runs for EHT at the LMT, building a 1mm dual polarization sideband-separation receiver with 4 – 12 GHz IF outputs (successful EHT MSIP proposal)

● Also proposing to build over next 3-4 years, a 16-pixel dual-polarization (2x4x2) one-millimeter array receiver (OMAR) using SIS mixers – will use CASPER-based spectrometers built for SEQUOIA for backend

37 of 43

OMAR – dual polarization 2 x 4 array; Each pixel with sideband separation mixers

38 of 43

Summary

● LMT is not fully complete – but is already a working instrument, collecting valuable science with two state-of-the-art instruments

● Third season of Early Science observations has started

● A wideband array spectrometer based on ROACH-2 and ASIAA ADCs is being put together for the SEQUOIA FPA

● A novel Prototype Phased Focal Plane Array Receiver is being built to be commissioned at the GBT – such instruments are poised to revolutionize heterodyne radio astronomy at mm/submm wavelengths

● 1mm wavelength single-pixel receiver for the EHT project is being prepared for the LMT

● OMAR – a 1mm wavelength focal-plane array receiver for LMT

39 of 43

40 of 43

Supplementary Slides

41 of 43

Combined Feed Element

Waveguide port

Waveguide port must be dielectric filled for best match, =1.3

42 of 43

Cold Preamp Circuit DesignOne continuous soft circuit board for all elements.

Waveguides in and out pass through the circuit board, conductive vias act as waveguide walls.

Circuit is glued to one block, clamped against another.

Pockets cut through board for MMIC amps and bias network.

Single element of circuit board, repeated 64 times.

In

Out

43 of 43