Upgrade of the VLBA CBand Feed Receiver Design

Upgrade of the VLBA C-Band Feed & Receiver Design Review Presented by S durand Designed by Bob Hayward and: Sivasankaran Srikanth (Feed Design) Hollis Dinwiddie (Mechanical Design) Everett Callan (Master Receiver Builder) Gordon Coutts & Craig Hennies (OMT Testing) Marian Pospieszalski & the CDL Amplifier Group Pat Madigan & the VLA Machine Shop Michael Hedrick & the Green Bank Machine Shop 21 April 2011

Why Upgrade the VLBA C-band ? • Astronomers were eager to: • Access the Methanol Maser at 6, 668 MHz GHz • Access to OH at 6016 -6049 MHz • Carry out continuum observations with two IF’s separated by several GHz to allow the effects of the atmosphere to be subtracted out Example : IF-AC=4. 0 -4. 5 and ID-BD=7. 5 -8. 0 GHz • Limitations of the existing C-Band system: • Mediocre sensitivity due to old GAs. FET LNA’s • Narrow bandwidth due to: – – Septum Polarizer (4. 5 -5. 2 GHz) LNA’s (4. 5 -5. 2 GHz) Warm IF (4. 5 -5. 2 GHz) T 105 Conversion (4. 5 -5. 2 GHz)

T(Rx) for Old VLBA vs. New EVLA Rx’s Average of 8 Dual-Channel Receivers Receiver Noise Temperature - VLBA vs. EVLA Receivers 50 45 40 Noise Temperature (K) 35 30 25 VLBA Rx 20 15 10 5 0 4. 5 5. 0 5. 5 6. 0 Frequency (GHz) 6. 5 7. 0 7. 5 8. 0

Required Specifications • To modify the VLBA Receiver to give us the RF performance that is as close to that achieved by an EVLA Receiver as possible.

Sri’s VLBA 4 -8 GHz Feed Design • • Profiled Corrugated Horn Four machined sections Ring-Loaded Mode Converter Plugs into old feed location Total Length = 37. 5” Outer Diameter = 18. 5” Weight = 55 lbs

Installation of the New Feed at VLBA Pie Town • Done with Feed S/N 01 • And Unmodified C-Band Rx • Uses a Transition Plate so an old Receiver can be attached to a new Feed and thus allow installation of lateral support brackets and turnbuckles. • Luckily feed is small enough to fit through the hatch (i. e. , a crane is not required for hoisting it over the dish).

Existing VLBA C-Band Receiver Key: Cryo Pamtech CTB 1107 Narda 4014 C-30 Keep -30 d. B Toss New Atlantic Microwave AMC 0935 LCP Narda 4014 C-30 Pamtech CTB 1107 High-Cal ( Not implemented on VLBA) +20 -10 d. B Omni-Spectra 2020 -6617 -10 LNA -30 d. B Pulse Cal T TCal & PCal T Noise Diode -6 d. B Omni-Spectra PCal 2089 -6203 -00 PA T RCP Septum Pol LNA TCal & PCal N/A Ditom Reactel Miteq Ditom D 3 I 4080 -1 6 B 1 -4850 -700 S 12 AMF-2 B-4552 D 3 I 4080 -1 4 -8 GHz 4. 5 -5. 2 GHz 4 -8 GHz Omni-Spectra 2020 -6616 -06 T GAs. FET 4. 5 -5. 2 GHz Noise Diode PA GAs. FET Ditom Reactel Miteq Ditom 4. 5 -5. 2 GHz D 3 I 4080 -1 6 B 1 -4850 -700 S 12 AMF-2 B-4552 D 3 I 4080 -1 4 -8 GHz 4. 5 -5. 2 GHz 4 -8 GHz The Cryogenic Pamtech Isolators were a surprise. There was no mention of them in the VLBA Technical Report #3. There must have been added after the 1 st prototype Receiver was built.

Upgraded VLBA C-Band Receiver Key: Cryo Narda 4014 C-30 Keep Dorado 3 ICC 60 -1 -30 d. B Toss New T LCP Ciao CA 48 -281 G=26 d. B PA Noise Diode Omni-Spectra 2089 -6203 -00 Mactech CA 7205 U Hybrid RCP TTE K 5221 -4/8 G 4 -8 GHz LNA TCal & PCal N/A OMT In HEMT 4 -8 GHz Ditom D 3 I 4080 -1 4 -8 GHz Noise/Com NC 3205 -G TCal T -10 d. B T Narda 4014 C-30 Dorado 3 ICC 60 -1 In. P HEMT 4 -8 GHz Pulse Cal LCP 4 -8 GHz Out PCal LNA -30 d. B RCP 4 -8 GHz Out Omni-Spectra 2020 -6617 -10 10 d. B TCal & PCal Ditom D 3 I 4080 -1 4 -8 GHz PA Ditom D 3 I 4080 -1 4 -8 GHz TTE K 5221 -4/8 G 4 -8 GHz Ciao CA 48 -281 G=26 d. B Ditom D 3 I 4080 -1 4 -8 GHz Note that we leave out the High-Cal feature, which has never been used on the VLBA

Modifications & Problems • Making the modifications to the RF path of the receiver were relatively straight forward. • However, cooling the beast was a problem. – The old VLBA receiver had its Septum Polarizer tied to the 50°K Stage. – We wanted the new OMT to be tied to the 15°K Stage since it has more resistive losses. • Hollis spent quite a while fighting to minimize the final temperature that the 15°K Stage reached. – It was hard to get it much below 30°K and the temperature was very dependent on how many other receivers were on the same Helium line. – Required reducing the weight of the OMT as well as adding Space Blankets around the 50°K Radiation Shield. • It was also decided that the Thermal Gap assembly used in the EVLA design would need to be improved.

EVLA-Style OMT & Thermal Gap AN 72 Absorber wrapped around Thermal Gap to prevent Cavity Resonances. Includes: “Old” Thermal Gap + Heavy Blocks + Absorber Strip

VLBA-Style OMT & Thermal Gap Includes: New Circular Thermal Gap with Choke Ring and long standoffs + “Swiss Cheese” Blocks to reduce thermal mass + Mount for Hybrid makes for an integrated Circular Polarizer

Recent Cool-down Results • Thanks to Hollis’ new Thermal Gap Assembly, the Model 22 is now cooling the RF Tree in the Prototype Receiver down to acceptable temperature levels. • The 15°K Cold Stage temperature is also less susceptible to variations from other fridges being connected to the Helium lines: – – – ~12 -14°K when alone ~14 -16°K with one Model 350 on the loop ~16 -18°K with two Model 350 on the loop • The 2 x Model 350 scenario is what occurs on a VLBA Antenna. • VLBA C-Bands typically 15 -25°K

VLBA C-Band Prototype (VC#11)

Early RF Sensitivity Tests on VC#11 Old EVLA vs. New VLBA Thermal Gap Receiver Noise Temperature on VLBA C-Band #11 -RCP Unmod (Circ) & Modified (Linear) Rx vs. EVLA Rx C#16 (Circ) (RHH : 13 April 2011) 50 45 Receiver Temperature (K) 40 35 30 25 VLBA C#11, Unmodified, Circular EVLA C#16, Old T-Gap, Circular 20 VLBA C#11, New T-Gap, Linear 15 10 5 0 4. 5 5. 0 5. 5 6. 0 6. 5 Frequency (GHz) 7. 0 7. 5 8. 0 Using the new T-Gap assembly with its proper Choke Ring yields a vast improvement over that achieved on the EVLA with the old T-Gap design.

VLBA C-Band Receiver Old vs. Interim vs. New Configuration RCP 4. 5 -5. 2 GHz C RCP 4. 5 -5. 2 GHz Old T 105 4. 5 -5. 2 GHz LCP Old C-Band Rx with Old T 105 Converter Module IF Output A&C 500 -1000 MHz L 104 #2 4. 5 -5. 2 GHz RCP 4. 0 -8. 0 GHz C RCP 4. 5 -5. 2 GHz 4. 0 -8. 0 GHz LCP Modified C-Band Rx with Old T 105 LCP (From old Rx) RCP 4. 0 -8. 0 GHz C SMA Modified C-Band Rx with New T 405 4. 5 -5. 2 GHz Converter Module IF Output A&C 500 -1000 MHz L 104 #2 RCP 4. 0 -8. 0 GHz LCP Old T 105 LCP New T 405 4 - 8 GHz Converter Module IF Output A, B, C & D 500 -1000 MHz L 104 #2 L 104 #3

Downconverter Design 3900 -5900 and 5600 -7900 MHz

Conclusions • Thanks to the new Thermal Gap design, the upgraded VLBA C-Band is now superior to the EVLA receiver. • Expect T(Rx) < 10°K from 4. 2 -8. 0 GHz • It is now probably the most sensitive wideband system in the world at this frequency.

Questions ?