feed & front end pdr

Bob HaywardReceiver Engineer

Feed & Front End PDR12-13 Feb 2002

1

Feed & Front End PDR

Monitor & Controland

Cryogenics



2

The “Card Cage” (Photograph)



3

The “Card Cage”

• Originated with X-band Receiver in 1980’s• Different for each Receiver Band• Hand-Wired

• Labor Intensive & Mistake Prone

• Many components now obsolete or hard to get• Mixes high voltage (150 VAC) & low-level

signals (ie: LNA bias)



4

Old M&C(for L, X, K & Q)

F14Module

Rx#2

CardCage

Data SetModule

Rx#1

CardCage

Rx#3

CardCage



5

Current Monitor Points (L, X, K & Q-Band)

• Monitor Points:• LNA VGate (RF1/2 & LF1/2) & LED Voltage• Dewar Temperature (300, 50, 15K Stages)• Vacuum Pressure (Dewar & Pump Line)• Cryo Status (Off, Cool, Stress, Heat, Pump)• Pump & Solenoid Valve Status • AC Current• Rx Band ID, S/N & Mod Level (4+6+2 bits)



6

Current Control Points (L, X, K & Q-Band)

• Control Points:• Cryo Status (Off, Cool, Stress, Heat, Pump) • Control LNA Bias Settings (with potentiometers)

• Signals that pass thru the Card Cage:• Lo-Cal Drive (TCAL)• Hi-Cal Drive (SCAL)



7

Current “Local” M&CFunctions

• Monitor Dewar Temperatures & Vacuum• Control Refrigerator and Vacuum Pump

• based on monitored parameter

• Control LNA Bias Settings • Communicate with Rest of World

• For VLA, used F14 Module• For EVLA, via the MIB



8

New M&C Guidelines

• Minimize Digital “Traffic”• Fail-Safe Design• Replace Obsolete Components• Minimize Space Required• Standardize Components & Systems



9

Cryogenic and Vacuum Control

• Operating Sequence Controlled “Locally”– “Fail-Safe” if Computer is “Down”

• Refrigerator will not start without vacuum in Dewar

• Current Design Works Well – Don’t change the basic logic– But do it in new hardware



10

“Stress” Mode and AC Current Monitor

• VLA has the capability to– Add “stress” heat load to refrigerator– Monitor total current to refrigerator and vacuum solenoids

• This was meant for remote diagnosis, but is not used– Cryo performance trends monitored instead

• This capability not needed for EVLA



11

Amplifier Biasing

• “Servoed” Bias Circuit – Maintains ID Constant by Varying VG

– Uses 8 pots to control 4 stages per card– Provides Buffered Monitor Voltages– Provides Over-voltage and ESD Protection

• Current Design Works Well – Don’t change the basic circuit design– But do it in new hardware



12

EVLA Receiver M&C Options

1) A dedicated Card Cage for each receiver• “Old” Philosophy• Need 240 units + spare cards

2) Single M&C Rack Unit for all 8+ Rx’s / antenna• “New” Philosophy• Total of 30 units + spare cards

• Surely easier and cheaper to build 30 rack units than build/modify 240 Card Cages



13

New M&CBlock Diagram

Q-Rx

Ka-Rx

K-Rx

Ku-Rx

X-Rx

C-Rx

S-Rx

L-Rx

Temp SensorCard(s)

Vacuum SensorCard(s)

Pump ControlCard(s)

LNA SerialCard(s)

Inte

rfac

e

LBC

Mic

roco

mpu

ter

AD

C /

PIO

/ SI

O /

Ethe

rnet LBC

LBC

LBC

LBC

LBC

LBC

LBC



14

Multi-Receiver M&C Unit

• Common Temp & Vacuum Monitoring• 4-wire DT-471 Temp Sensors good for cables many

feet in length• Need to investigate maximum cable length Hasting

DV-6 pressure sensors can handle (10-500 mV)

• Remote Fridge & Pump Control • Avoids 150 V / 60 Hz near sensitive LNA bias lines



15

M&C Computer

• Embedded Control Computer• How smart does the Rx C need to be?• MIB or dedicated Front End Subsystem ?• Engineering M&C and Diagnostic Software• Since no “Local” access, need identical units for

Lab testing



16

New M&C CardsFunctions

• New Temperature Sensor Cards:• Cryo Temps : 2 per Rx times 8 = 16 (+spares)• Room Temps : 1 per Rx times 8 = 8 (+spares)

• New Vacuum Sensor Cards:• 2 per Rx times 8 = 16 (+spares)

• New Control Card:• Pump / Fridge sequencing and Heater On/Off control

for each Receiver (ie: 8+)



17

New Amplifier Bias Cards

• Can be much smaller than old design• Possibly mount inside Dewar• Bias lines shorter, shielded by Dewar• Fewer pins in Dewar feedthru connector

(typically 9 lines required for a 4-stage LNA)• Cheaper• Better seal• Better RFI bypassing



18

New LNA Bias Card

• LNA Bias Card (LBC):• Minimum of 4 stages x 2 Polarizations = 8 stages• Sync Serial Interface to Rx M&C Unit• Programmable Drain Voltage and Current settings• Monitor all VDrain and IDrain settings plus VGate

• Use NRAO-Tucson has a Quad LNA (16-stage) card ? (8-layer board, 800 cpts, 32 channel DAC

SCLK, SYNC, DIN + VD/ID/VG-Mon)• Non-volatile Bias Settings ?

(or controlled by a “remote” PIC micro like Quad LNA)



19

NRAO-TucsonALMA Quad LNA Bias

Card Block Diagram

LNA2_VG_MON[1..5]

LNA2_D5

LNA4_G3

IDS_MON

LNA2_G3

LNA2_VD_SET[1..5]

U1

SN74HC594DW

14

11

10

12

13

151234567

9

SER

SRCLK

SRCLR

RCLK

RCLR

QAQBQCQDQEQFQGQH

QH'

MON_A[0..3]

D1MMSZ5240BS

VCC

SYNC2

LNA1_G3

MON_A3

LNA3_VG_MON[1..3]

MON_A0

DIN

VSS

VCC

AVCC

DIN

LNA1_D5

10.0V

LNA1_D2

R43.92K

LNA3_G2

LNA4_D2

R1

752-091-100K

1 2

34

56

789

LNA4

3-Stage_HEMT_LNA_Bias

ID_MON[1..3]DRAIN[1..3]

GATE[1..3]

VD_SET[1..3]

ID_SET[1..3]REF

VD_MON[1..3]

VG_MON[1..3]

ENAB

LNA3_ID_SET[1..3]

LNA1_G1

U2

MAX8216CSD

1

9

10111213

7

56

14

34

VREF

DOUT

OUT4OUT3OUT2OUT1

DIN

+15V-15V

VDD

+5V-5V

LNA1_ID_MON[1..5]

VGS_MON

LNA3_D1

LNA3_VD_MON[1..3]

LNA2_ID_SET[1..5]

LNA1_G5

MON_A1

LNA2_G5

LNA3



GATE[1..3]

VD_SET[1..3]

ID_SET[1..3]REF

VD_MON[1..3]

VG_MON[1..3]

ENAB

LNA4_G1

LNA1_D1

VDS_MON

ENAB_LNA4

LNA4_D3

LNA2_G2

LNA3_D[1..3]

LNA4_ID_MON[1..3]

R2

1K

LNA3_D2LNA3_G1

MON_A2

VDS_MON

ENAB_LNA3

LNA1_G2

LNA2_D[1..5]

SCLK

LNA2



GATE[1..5]

VD_SET[1..5]

ID_SET[1..5]REF

VD_MON[1..5]

VG_MON[1..5]

ENAB

LNA2_G[1..5]

LNA2_G1

LNA1



GATE[1..5]

VD_SET[1..5]

ID_SET[1..5]REF

VD_MON[1..5]

VG_MON[1..5]

ENAB

LNA2_VD_MON[1..5]

LNA1_VG_MON[1..5]

IDS_MON

VSSVDD

LNA1_VD_SET[1..5]

LNA3_G[1..3]

LNA1_D[1..5]

LNA1_G4

ENAB_LNA2

LNA3_ID_MON[1..3]

LNA1_D4

LNA2_D2

SYNC1

VDD

SCLK

LNA3_G3

VGS_MON

SYNC2

LNA2_G4

R311.0K

LNA4_VD_SET[1..3]

LNA3_VD_SET[1..3]

J1100-964-053

A1A2A3A4A5A6A7A8A9

A10A11A12A13A14A15A16A17A18A19A20A21A22A23A24A25A26A27A28A29A30A31A32

C1C2C3C4C5C6C7C8C9C10C11C12C13C14C15C16C17C18C19C20C21C22C23C24C25C26C27C28C29C30C31C32

ALMA04023SX0016

ALMA Quad LNA Bias Card

Atacama Large Millimeter Array (ALMA) Project

National Radio Astronomy Observatory949 N. Cherry Ave.Tucson, AZ 85721

B

1 23Thursday, August 23, 2001

Title

Size Document Number Rev

Date: Sheet of

LNA2_ID_MON[1..5]

LNA1_VD_MON[1..5]

ENAB_LNA1

LNA1_ID_SET[1..5]

AVCC

POWER_ON

LNA4_VG_MON[1..3]

LNA4_G2

LNA4_D[1..3]

LNA1_G[1..5]

LNA2_D4

LNA4_G[1..3]

LNA2_D1

LNA4_D1

VCC

VDD

LNA4_VD_MON[1..3]

LNA1_D3

LNA4_ID_SET[1..3]

LNA3_D3

LNA2_D3

To 5-StageLNA #1

To 5-StageLNA #2

To 3-StageLNA #3

To 3-StageLNA #4



20


Card OverviewLNA1_VD_MON[1..5]

3VGS2

U5

AD5532ABC-1

A2

A3

A4

A5

A6

A7

A8

A9A10

B1

B3

B4

B5

B6

B7

B8

B9

B10

B11

C1

C2

C10

C11

D1

D2

D10

D11

E1

E2

E10

E11

F1

F2

F10

F11

G1

G2

G10

G11

H1

H2

H10

H11

J1

J2

J6

J10

J11

K1

K2

K3

K4K5 K6K7

K8

K9

K10K11

L2L3L4

L5L6 L7

L8

L9L10

A4

A2

A0

CS/SYNC

DVC

C

SCLK

OFFSET_SEL

BUSYTRACK/RESET

VO16

A3

A1

WR

DG

ND

DIN

CAL

SER/PAR

DOUT

REF_IN

VO18

DAC

_GN

D1

AVC

C1

REF_OUT

VO20

DAC

_GN

D2

AVC

C2

OFFS_OUT

VO26

VO14

AGN

D1

OFFS_IN

VO25

VO21

AGN

D2

VO6

VO24

VO8

VO5

VO3

VO23

VIN

VO4

VO7

VO22

VO19

VSS2

VO9

VO11

VO17

VO15

VO27

VSS3

VSS1

VSS4

VDD

2

VO2

VO10

VO13VO12

VO28VO29VO30VD

D3

VDD

1

VDD

4

VO31

VO0VO1

1IDS2

2IDS1

4VDS1

3IDS1

MON_A1

3SET_VD2

VDD

2IDS2

1SET_VD2

+

C8

10uF

2IDS[1..5]

2SET_ID1

4SET_ID2

2SET_ID4

U7SN74AHC1G04

2 4

4IDS[1..3]

3VDS[1..3]

LNA1_ID_MON[1..5]

MON_A3

2SET_VD3

DIN

1VDS[1..5]

1VGS1

1SET_ID4

1VGS[1..5]

1SET_ID1

C5

0.1uF

4VGS2

VSS

1VGS2

4SET_ID[1..3]

MON_A3

C7

0.1uF

+

C1

10uF

2VGS2

1SET_VD[1..5]

LNA1_VG_MON[1..5]

LNA2_ID_SET[1..5]

MON_A2

4SET_VD1U6

DG406DN

151617

18

1920212223242526

28

456789

1011

14

127

A2A1A0

EN

S1S2S3S4S5S6S7S8

D

S16S15S14S13S12S11S10S9

A3

+VCC-VCC

VDS_MON

1VDS5

2SET_VD1

2VGS5

ALMA04023SX0016

ALMA Quad LNA Bias Card


National Radio As tronomy Obs ervatory949 N. Cherry Ave.Tucson, AZ 85721

B

2 23Thursday, August 23, 2001

Title


Date: Sheet of

1IDS5

2VGS1

1SET_ID5

VCC

2IDS4

3VDS3

VDD

3VGS[1..3]

LNA4_VG_MON[1..3]

2VDS4

4IDS3

VD_MON

1VGS5

3SET_ID3

SYNC2

2VDS[1..5]

MON_A[0..3]

IDS_MON

2VDS5

-

+

U4B

TL034CD

5

67

LNA2_VD_MON[1..5]

4IDS2

MON_A3

VCC LNA2_VD_SET[1..5]

VGS_MON

2SET_ID[1..5]

LNA3_ID_SET[1..3]

4SET_VD2

2SET_VD4

LNA1_ID_SET[1..5]

LNA3_VG_MON[1..3]

MON_A[0..3]

-

+

U4A

TL034CD

3

21

1IDS[1..5]

3IDS2

VCC

ID_MON

VDD

2SET_VD2

LNA4_VD_SET[1..3]

VDD

1VDS2

2VDS1

1SET_VD3VCC

AVCC

1IDS43SET_ID1

3VGS1

3SET_VD1

LNA4_VD_MON[1..3]

2IDS3

4IDS1

MON_A2

-

+

U4C

TL034CD

10

98

LNA3_ID_MON[1..3]

1VDS3

MON_A1

2SET_VD5

VSS

3VDS1

4SET_VD3

U3

DG406DN

151617

18

1920212223242526

28

456789

1011

14

127

A2A1A0

EN

S1S2S3S4S5S6S7S8

D

S16S15S14S13S12S11S10S9

A3

+VCC-VCC

1SET_VD4

+

C3

10uF

1SET_VD1

VCC

SCLK

MON_A2

Connect AGND and DGND at asingle point, close to AD5532.

3SET_ID2

2SET_VD[1..5]

1VGS3

3IDS[1..3]

4SET_ID3

3VGS3

4VDS2

2VGS4

4VGS[1..3]

2IDS5

1VGS4

1IDS3

U9AD780BR

2

4

3

6

58

IN

GND

TEMP

OUT

TRIMOP_SEL

AVCC

VSS

3VDS2

LNA4_ID_SET[1..3]

2VGS[1..5]

4VGS1

+

C2

10uF

LNA2_VG_MON[1..5]

1SET_ID[1..5]

3SET_VD[1..3]

2SET_ID2

LNA2_ID_MON[1..5]

1SET_ID3

C6

0.1uF

1SET_ID2

4VDS[1..3]

3IDS3

MON_A0

2VDS3

4SET_ID1

LNA3_VD_MON[1..3]

2SET_ID3

LNA4_ID_MON[1..3]

2VDS2

2VGS3

VG_MON

LNA3_VD_SET[1..3]

MON_A0

4VDS3

3SET_VD3

U8

DG406DN

151617

18

1920212223242526

28

456789

1011

14

127

A2A1A0

EN

S1S2S3S4S5S6S7S8

D

S16S15S14S13S12S11S10S9

A3

+VCC-VCC

C4

0.1uF

1VDS1

1SET_VD5

1VDS4

MON_A0

4SET_VD[1..3]

2SET_ID5

LNA1_VD_SET[1..5]

MON_A1

4VGS3

3SET_ID[1..3]

VSS

1IDS1



21


LNA Bias Circuit

R3627.4

1. All resistors are thick film SMD (0603), 1% unless specified otherwise.

R331K

VG_ON

NOTES:

2. All capacitors are ceramic SMD (0603). ALMA04023SX0016

ALMA Quad LNA Bias Card: LNA1, Stage1


National Radio Astronomy Observatory949 N. Cherry Ave.Tucson, AZ 85721

A

8 23Saturday, July 28, 2001

Title


Date: Sheet of

R35

27.4K

R30

100K R3247.5K

VD_SET

R38

274K

R2910K

U26A

EZADT

102

Q2BFMB3946

24

3

DRAIN

R31274K

U23A

EZADT

102

-

+

U24C

TL034CD

10

98

C109

0.1uF

U27BADG202AKR

151614DIN

S

3.3V

R24100

R26332K

DRAIN

C111

0.33uF

ID_MON

-

+

U24B

TL034CD

5

67

VD_MON

GATE

R23

100K 0.1%

-

+

U24A

TL034CD

3

21

-

+U20A

TL034CD

3

21

C1070.1uF

VD_ON

GATE

-

+U24D

TL034CD

12

1314

-

+U22A

TL034CD

3

21

R27

100K 0.1%

VG_MON

R28

2.21KQ2AFMB3946

5

1

6

R373.01K

D8MMSZ5226BS

R25

100K 0.1%

C110

1uF

C1080.033uF

R22

100K 0.1%

VDD

D111N4148WS

GATE

D91N4148WS

U21A

EZADT

10 2

D101N4148WS

ID_SET

R344.75K

U25AADG202AKR

21

3D

INS

U27AADG202AKR

2 13 D IN

S



22

Monitor by Exception

• If go with minimum amount of processing power:– Instead of monitoring all the voltages, all the time

• Each VG compared to its saved “OK” value– If all are OK, a monitor bit is “low”– If any are not OK, the bit goes “high”

– For diagnostics, the MIB can cycle through and read back all voltages and currents on command

– Bias voltages set by non-volatile “digital pots”• Could be changed, if required, at any time

– But Cryo & Vacuum need continuous monitoring



23

Monitor Continuously

• If use a Rx C with lots of processing power:– Monitor all the voltages, all the time

• Each VD, ID & VG compared to its saved value– Reload if “set” values are not correct– Catches instances where LBC card has lost power

• Provides processed data to an Engineering M&C Lap Top as well as Master Control Computer

• Temp inC, Pressure in microns, etc

• Can the MIB handle this?



24

Design Work

• Will need New Cards– Temperature, Vacuum, Control Boards

• Straight forward - just update our existing design• Do more on less real estate

– LNA Bias Card• New (or modified NRAO) design

– Software• Depends on whether done in MIB or dedicate Rx C

• May need to hire a new junior engineer?



25

CryogenicsRefrigerator Drives

• The Refrigerator Motor on each Rx requires a 150 Volt, 2-Phase Supply

• We now derive the Second Phase using a R-C Network– Requires a separate “box” for each Front End– Requires “tuning” to get phase shift right

• Proposed Alternate: Scott Tee



26

CryogenicsScott Tee

• A Way of Connecting Two Transformers to derive 2-phase from 3-phase

• Smoother Drive Without Tuning• Transformers Already In-House• Mount in Module or Bin

– Include Vacuum Pump Control, He Pressure Sensors, etc.



27

CryogenicsHelium Compressor

• VLA currently has 2 compressors supplying 80 S.C.F.M. (total) to 5 refrigerators per antenna

• EVLA will need 120 S.C.F.M. for 8 refrigerators• New compressors required

• Testing of a higher flow compressor already started• Direct replacement• Cost – on the order of $350K

feed & front end pdr

Documents

sparesnew control card

pump control lna bias

bitscurrent control

xband receiver

dedicated card cage

vacuum pumpbased

engineering mc

vacuum solenoidsthis