Frequency electronics, inc. Rb and quartz oscillators for space applications

1. FREQUENCY ELECTRONICS, INC.

Rb and Quartz Oscillators for Space Applications
Martin Bloch
1

2. Frequency Electronics, Inc.

• FEI was started 51 years ago with a
mission to bring precision timing
technology from the laboratory to the
real world
• FEI is a Long-Recognized Technology
Leader
High
Precision Timing, Space Applications
Low Phase Noise Microwave Sources and
Synthesizers
Low “G”- Sensitivity Clocks
Timing/Frequency Systems for Severe
Environments
2

3. Development of Rb and Quartz Oscillators for Space Applications

1.
2.
3.
Quartz Clock Performance in Space
Rb Atomic Clocks in Space – The Results So Far
Next Generation Rb Atomic Clock for Space
Applications
A.
B.
Design, Rb Atomic Frequency Standard (RAFS)
Performance Results, RAFS
3

4. Introduction

• For the past 50 years, most satellite systems have used
precision quartz oscillators as the time and frequency
reference
– Reliable
– Low Power
– Light Weight
• Atomic Frequency Standards have demonstrated
reliable operation in space during the last 25 years on a
limited number of programs
–
–
–
–
GPS
Glonass
Milstar/AEHF
Galileo
4

5. Technology Trade-offs

• Quartz oscillators are the most reliable, lowest cost
technology for providing precision frequency and time in
space applications
• Quartz is inherently sensitive to nuclear radiation
• Atomic frequency standards cost more, are less reliable,
but provide improved frequency stability, and are
relatively immune to radiation
Frequency Stability (Allan deviation)
10 sec
100 sec 1000 sec
10,000 sec
100,000 sec Drift per day
Technology Power
Weight 1 sec
Quartz
2 Watts < 1 kg
1.00E-13 1.00E-13 1.00E-13
3.00E-13
1.00E-12
3.00E-12
1.00E-12
Rubidium 15 Watts 2 - 5 kg
1.00E-12 3.00E-13 1.00E-13
3.00E-14
1.00E-14
1.00E-14
7.00E-15
5

6. Development of Rb and Quartz Oscillators for Space Applications

1. Quartz Clock Performance in Space
2.
3.
Rb Atomic Clocks in Space – The Results So Far
Next Generation Rb Atomic Clock for Space
Applications
A.
B.
Design, Rb Atomic Frequency Standard (RAFS)
Performance Results, RAFS
6

7. State of the Art Quartz Performance in Space

• Usage of “Premium Q Swept Quartz” or radiation
hardened quartz material.
• SC-cut crystals (SC-cut crystals stabilize faster than ATcut crystals. The retrace of SC-cut crystals is orders of
magnitude better than AT-cut crystals).
• 5th overtone resonators (aging is significantly affected
by the thickness of the resonator, hence, the thickest
quartz blank should be used at the highest practical
overtone for best aging performance).
• Crystals exhibiting monotonically-positive aging slope
(radiation offsets the positive aging trend of quartz as
further explained below).
7

8. Radiation Effects on Quartz

Quartz sensitivity to most radiation
has been shown to be
approximately:
-1E-12 per Rad
Typical background radiation for
geosynchronous orbit is:
6 Rads/day
Typical frequency aging on earth
is ~1E-11/day for high precision
quartz oscillators
On-orbit compensation resulting in
frequency aging of ~1E-12/day is
possible for positive aging quartz
resonators
Quartz sensitivity to protons is
similar magnitude, but less
predictable; and can result in
positive or negative frequency
changes
Even with compensation for background
radiation:
• best performance observed in space is
~2E-13/day
• effect of solar flare is 1E-10 to 4E-10 over
several days
8

9. State of the Art Quartz Oscillator

-133
Lᵠ(f) = -133 dBC/Hz @ 1 Hz offset
ADEV = 7E-14 @ 10 sec
9

10. Development of Rb and Quartz Oscillators for Space Applications

1.
Quartz Clock Performance in Space
2. Rb Atomic Clocks in Space
– The Results So Far
3.
Next Generation Rb Atomic Clock for Space
Applications
A.
B.
Design, Rb Atomic Frequency Standard (RAFS)
Performance Results, RAFS
10

11. Overview

• Compared to crystal oscillators, atomic clocks are inherently insensitive
to space radiation.1
• Compared to other types of atomic clock (e.g., Cs),
Rb clocks have the advantage of size, weight, and power,
Rb clocks have the (believed) disadvantage of relatively large frequency
aging.
Parameter
3
Rb Clock
size (in )
100
weight (kg)
2.5
power (W)
20
frequency aging (day-1) 10-12 to 10-13
Cs Clock
1000
20
50
< 10-14
• To better understand the utility of Rb clocks for long-term space
missions, we have investigated Milstar Rb clocks (on-orbit) over the past
two decades.
• Contrary to popular mythology,
Rb clocks can have exceptionally long life with stable operation,
Rb clocks can have extremely low frequency-aging rates: 10-14 to 10-15/day
1) J. Camparo, S. Moss, & S. LaLumondiere, Space-system timekeeping in the presence of solar flares, IEEE
Aerospace and Electronic SYSTEMS Magazine, 19(5), 3-8 (2004).
11

12. Milstar Rb Clocks

• The Milstar Rb atomic clocks are manufactured by Frequency
Electronics, Inc. (FEI)
• The clock is a classical Rb vapor-cell design, with optical
excitation using an rf-discharge lamp.2
– The clock has a 10 year design life.
– The clock has a weight of 2.3 kg (i.e., 5 lbs.).
– The clock frequency can be tuned with a resolution better than 1
10-12.
Solenoid Windings
Microwave Cavity
To Generate VCXO
Correction Signal
87Rb
rf-Discharge
Lamp
85Rb
6,834.7 MHz
Filter Cell
Resonance Cell:
87Rb
Photodiode
& Buffer Gas
2) T. McClelland, I. Pascaru, M. Meirs, Development of a Rb Frequency Standard for the MILSTAR Satellite
System, 41st Annual Symposium on Frequency Control, 1987, p. 66
12

13. On-Orbit Data: Rb Satellite Clocks Functional Fit: y(t) = Ae-gwt + Be-get + Dt

• Empirically, we find that as a family the data is well fit by a biexponential plus linear term:
g
t
g
t
w
e
y
t
A
e
B
e
D
t
• Clock “warm-up”
– Ae-gwt
– This term may be due to movement of the liquid Rb pool in the discharge
lamp.
• Clock Equilibration4
– Be-get
– The mechanism of equilibration is not yet understood.
– Nevertheless, helium permeation remains an open possibility.
Linear Frequency Aging
– Dt
– The mechanism of linear frequency aging is not yet understood.
13

14. Milstar Rb Satellite Clocks Summary

• Six clocks have operated on orbit (data for three presented
here)4
• Longest operating time on-orbit is 14.8 years
• Including estimates for ground operation, 3 of these clocks
have logged more than 18 years of operation!
Clock ID
A
B
C
D
E
F
A term
time constant
(1/γw)
(in months)
A
-10
-2.91 10
-10
-2.12 10
-10
-2.00 10
-10
-15.3 10
-10
-0.91 10
1.4
0.2
3.4
4
0.7
B term
time constant
(1/γe)
(in years)
B
-10
-7.62 10
-10
-0.00 10
-10
-7.44 10
+9.96 10
-10
-5.64 10
-10
N/A
-2.43 10
-10
1.62
-2.20 10
-0.00 10
averages -3.87 10
-10
1.13
Operating time (years)
Ground
total
On-Orbit (estimate) (estimate)
D term
(1/day)
-14
-(2.7 ± 0.2) 10
N/A
+(0.07 ± 0.11) 10
1.49
-14
0.36
0.76
-10
0.59
-10
0.72
12.7
5.9
18.6
6.1
2.4
8.5
14.8
3.4
18.2
-14
+(2.3 ± 0.2) 10
-14
8.4
2.2
10.6
-14
13.9
4.3
18.2
-14
5.8
1.9
7.7
10.3
3.3
13.6
+(0.24 ± 0.16) 10
-(3.6 ± 0.1) 10
-(7.3 ± 2.2) 10
-1.83 10
-14
4) J. Camparo, T. McClelland, and J. Hagerman, “Long Term Behavior of Rb Clocks in Space,” European
Frequency and Time Forum, 2012.
14

15. Development of Rb and Quartz Oscillators for Space Applications

1.
2.
Quartz Clock Performance in Space
Rb Atomic Clocks in Space – The Results So Far
3. Next Generation Rb Atomic Clock
for Space Applications
A.
B.
Design, Rb Atomic Frequency Standard (RAFS)
Performance Results, RAFS
15

16. Design Goals

• Precision Rubidium Atomic Frequency Standard
(RAFS) for Space Applications
– Best possible frequency stability
– 20 year operating life
• Emphasis on frequency stability rather than size, weight
and power (SWAP)
• “Classical” Rb vapor frequency standard
– Optical pumping using Rb lamp to create ground state hyperfine
population difference
– Temperature controlled filter cell with Rb85
– Temperature controlled resonance cell, with Rb87, in microwave
cavity
• One prototype, 3 engineering models
16

17. Development of Rb and Quartz Oscillators for Space Applications

1.
2.
3.
Quartz Clock Performance in Space
Rb Atomic Clocks in Space – The Results So Far
Next Generation Rb Atomic Clock for Space
Applications
A. Design, Rb Atomic Frequency Standard
(RAFS)
B.
Performance Results, RAFS
17

18. Rb Atomic Frequency Standard (RAFS)

• 15 x 4.6 x 5.2 inches
(381 x 117 x 132 mm)
• 16.5 Lbs (7.5 kg)
• 28 VDC Input
• 30 Watts
• 10 MHz Output
18

19. Block Diagram

• Temperature controlled chassis (±1°C) operates
from -34°C to +25°C
• Modular Design
• Digital Rb control loop
19

20. Digital Rb Control Loop

• Digital processing of analog signal from physics
package
• Space qualified FPGA
• Direct Digital Synthesizer (within FPGA) to tune
output frequency (1 x 10-14 tuning resolution)
20

21. 6.8 GHz Frequency Synthesizer

• Phase locked CRO at
2.278 GHz
• Output signal at 6.8
GHz has no
sidebands within a
±2.278 GHz window
21

22. Radiation Hardening

• Component selection
• FPGA
– Frequency setting stored on select resistors connected to input
pins
– Fuse programmed (write once)
– Hardware triple redundant logic, with three way voting to
minimize single event effects
– Software triple redundant logic with 3 way voting of critical
values (digital output to DAC (quartz oscillator control voltage))
• Radiation shields
– Chassis, covers (material and thickness)
– Spot shields for critical components
22

23. Development of Rb and Quartz Oscillators for Space Applications

1.
2.
3.
Quartz Clock Performance in Space
Rb Atomic Clocks in Space – The Results So Far
Next Generation Rb Atomic Clock for Space
Applications
A.
Design, Rb Atomic Frequency Standard (RAFS)
B. Performance Results, RAFS
23

24. Relative Frequency vs Time prototype unit in vacuum

24

25. Allan Deviation Prototype Unit in vacuum

25

26. Allan Deviation multiple units, in vacuum

1.00E-11
Allan Deviation
1.00E-12
Performance Goal
SN 00, Feb 2010
SN 02, Sept 2010
SN 02, Aug 2010, corrected
SN 03, Sept 2010
maser 20 vs maser 73
1.00E-13
1.00E-14
1.00E-15
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
1.E+06
1.E+07
Averaging Time (seconds)
26

27. Warm-Up and Stabilization SN 01 in vacuum, -4°C

Warm-Up and Stabilization
1.0E-09
70
0.0E+00
65
-1.0E-09
60
-2.0E-09
55
-3.0E-09
50
-4.0E-09
45
-5.0E-09
40
-6.0E-09
35
0
1
2
3
Input Power (Watts)
Relative Frequency
SN 01 in vacuum, -4°C
Frequency
DC Power
4
time (hours)
27

28. Initial Warmup SN 01 in vacuum, -4°C

1.E-09
70
Rb loop locked
(13 min.)
65
-1.E-09
60
-2.E-09
55
-3.E-09
50
-4.E-09
45
-5.E-09
40
-6.E-09
35
0
0.25
0.5
0.75
Input Power (Watts)
Relative Frequency
0.E+00
Frequency
DC Power
1
time (hours)
28

29. Frequency vs Temperature SN 01, in vacuum

1E-14 per °C
29

30. Power Consumption SN 01 in Vacuum

SN 01 Power Consumption (Vacuum)
Base-plate
Temp (˚C)
71
61
9
5
-4
-5
-20
-24
-34
Power
Consumption
(W)
19.91
20.52
30.85
31.61
36.06
37
42.62
44.41
48.22
30

31. Frequency vs Magnetic Field SN 01 in vacuum

Optical Axis (X - Axis)
magnetic field
Δ
df/dH
(gauss)
(freq - fit line) (1/gauss)
3
-2.2742E-13 -7.5807E-14
-3
2.7864E-13 -9.2880E-14
Optical Axis (X - Axis)
4
3,E-12
3
Relative Frequency
4,E-12
2,E-12
2
1,E-12
1
0,E+00
0
-1,E-12
-1
-2,E-12
-2
-3,E-12
y = -9,3134E-17x + 4,1915E-13
-4,E-12
0
1800
3600
-3
5400
7200
Magnetic Field (Gauss)
mean sensitivity: -8.4344E-14
frequency
frequency (100 sec
avg)
frequency (1800 sec
avg)
magnetic field
-4
9000
time (seconds)
31

32. Allan Deviation SN 03 SN 03, vacuum

Unit Name:
fit function
a Ln b t
start slope
1
c, with a
SN 03
1.69609 10 11 , b
1.14892 10 13 day
end slope
0.00677393, c
1.01406 10 9
8.67013 10 14 day
1.0 10 13
GOAL
5.0 10 14
Allan Deviation
START: Sun 19 Sep 2010 23 : 58 : 20
STOP: Sat 6 Nov 2010 23 : 58 : 20
7.0 10 14
SN 03
3.0 10 14
2.0 10 14
1.5 10 14
1.0 10 14
100
1000
104
105
Averaging Time seconds
106
107
32

33. Summary Next Generation Rb Atomic Clock

• Demonstrated performance over environments
–
–
–
–
Temperature (-35°C to +71°C) in vacuum
Warmup from OFF condition (-4°C) in vacuum
Magnetic Field (-3 gauss to +3 gauss)
Continuous extended operation at 8°C in vacuum
• Performance at constant temperature in vacuum:
– Demonstrated: y( ) < 9 x 10-13 / + 2 x 10-14
for between 1 and 1,000,000 seconds
– Target:
y( ) < 7 x 10-15 at = 100,000 sec
33