|
Figure 1:
The portable 10 MHz
Rubidium frequency reference.
Click on the image for a
larger version.
For a similar page about
the FEI FE-5680A rubidium reference, click here.
Note: Unlike the FE-5680A Rubidium,
the Efratom LPRO-101 is directly suitable
for providing a 10 MHz reference for microwave
transverters - see below
for more info.
Why a frequency reference?
When operating on the microwave amateur radio bands, narrowband
modes (such as SSB or CW) are often used to maximize the link
margin - that is, be able to talk when signals are weak as the
SSB or CW modes can offer 10's of dB improvement over the wide
FM modes used with Gunn transceivers.
There is a catch, however: The use of microwave
frequencies
and
narrowband modes such as SSB or CW means that the one must
maintain pretty good frequency stability
and accuracy:
- Stability is important as a drift of even a few hundred Hz
at the operating frequency (in the GHz range!) can affect
intelligibility of voice - or, if CW is being used for
weak-signal work, such drifting can move the received signal
outside the receiver's passband filter! Having to
"chase" the frequency around is not only distracting, but it
complicates being able to communicate in the first place.
- Accuracy is also critical because it is important that
both parties be confident that their stated operating
frequencies are reasonably close to where they think they
are, frequency-wise. If a contact is arranged
beforehand, it is vital that both parties be able to find
each other simply by knowing the intended frequency of
communication and as long as the two parties are within
several hundred Hz of each other, it is more likely that
they will be able to complete the contact. If the
error was on the order of several kHz or 10's of kHz,
"hunting" would be required to find the signal and if those
signals are weak, they might be missed entirely -
particularly if, in addition to tuning around, it was also
necessary to move the antennas about as well!
Because achieving such stability and accuracy requires some
effort, it is more convenient if microwave gear is constructed
such that it can use a common, external frequency reference and
lock to it for several reasons:
- Only one reference is required. It's better
to expend the effort in putting together one "master",
stable reference rather than each piece of gear having its
own reference. In this way, the extra money, time and
effort saved can be put toward having this one reference be
as good as you can make it.
- Power savings. Having one common reference
can also be convenient if one is operating portable using
battery power. Using an external reference means that
one doesn't need to keep all of those individual pieces of
gear "warmed up" all of the time to maintain stability,
turning it on (and draining battery power) only when it is
needed. For this reason, many amateur radio operators
(and commercial equipment manufacturers, for that matter)
design their gear to accept a 10 MHz input from a
known-accurate and stable source.
In addition to the Rubidium frequency reference described here,
I also have a
10 MHz
"ovenized" crystal oscillator that I generally use
instead of the "ruby." While not as accurate, the crystal
oscillator's stability and accuracy is more than adequate for
operation at least through 24 GHz (it is within a few hundred Hz
at that frequency) and consuming significantly less power to
operate than the Rubidium reference - an important consideration
when operating from battery. Nevertheless, it's nice to
have something that is portable and "dead on" - and can also be
used as a backup if necessary.
About this frequency reference:
The Rubidium frequency reference - an Efratom LPRO-101 - was
obtained on the surplus market (e.g. EvilBay), having (probably)
been pulled from obsolete cellular telephone gear. Where
crystal-based oscillators use the mechanical properties of the
quartz to determine the operating frequency, a Rubidium
reference utilizes an atomic resonance based on the passage of
light passing through a Rubidium-vapor lamp in an RF
field. This mechanism of operation is fundamentally more
stable than that of a quartz oscillator, allowing orders of
magnitude greater long-term stability.
Unlike a quartz crystal oscillator which has no clearly-defined
"wear out" period and, if well-designed, can actually
improve
as time goes on, a Rubidium reference has a definite lifetime
associated with its lamp: As the unit operates, the
Rubidium within the lamp is gradually consumed and eventually,
too little vapor is available for the atomic resonance to be
detected and the unit fails. For this reason, many
amateurs that use Rubidium references choose
not to leave them on all of
the time. For "base station" use, a GPS-based disciplined
quartz oscillator is often used as the "primary" reference
against which the Rubidium unit is compared. Since a
GPS-based disciplined reference is, by its nature, not
"portable" (that is, you can't just move it around unless you
stay in one location for many hours, giving it time to re-lock)
a Rubidium reference fills the niche providing very high
accuracy and stability in a portable package..
At room temperature (around 68F/25C) the LPRO-101 takes 3-5
minutes to warm up and "lock" (much faster than a crystal-oven
reference!) immediately providing accuracy equal or better than
a good-quality "ovenized" quartz oscillator, gradually achieving
something approaching its ultimate accuracy after a few 10's of
minutes. The LPRO-101 has two "frequency trim" adjustments
- one "mechanical" via a potentiometer and the other
"electrical" (from its connector) - that can be used to provide
fine-tuning of its output frequency. Generally speaking,
one would set the Rubidium's frequency just once using the
potentometer using as close to a "primary" reference as you can
find (such as a GPS disciplined oscillator for most of us!) as a
basis of comparison while the electrical adjustment (on one of
the pins of the connector) would be used if one were to
incorporate the reference into, say, a GPS disciplined system or
were to add a means of temperature compensation to
further-increase accuracy.
The LPRO-101 has a number of other outputs as well. In
addition to the "BITE" (which stands for "Built-In Test
Equipment") signal
(see below) there are several other
signals that indicate the unit's internal health - such as the
"Lamp Voltage" (which provides a relative indicator of remaining
life of the Rubidium lamp) and crystal tuning voltage.
While these signals are useful for diagnostic purposes, but I
chose not to bring them out of the box with the idea that if the
unit were to fail in the field where it might be used, I
wouldn't likely be able to do any in-depth diagnosis,
anyway! That being said, I may add some "pin jacks" at
some point to allow easy checking with a voltmeter.
Comments:
- Once warmed up (after 30 minutes) the lamp voltage on my
LPRO-101 is around 7 volts and the crystal tuning voltage is
around 6.5 volts - both indicative of still-healthy
electronics and physics packages.
- If you don't have something like a GPS Disciplined
oscillator or a cesium reference to which you can calibrate
your newly-acquired rubidium, don't despair: If it
achieves "Physics Lock" it is probably much more
stable and accurate than any quartz-based reference that you
are likely to own even if you can't fine-tune it.
Assembling
the unit:
The LPRO-101 unit was originally mounted to a heat sink plate in
its originally-configured chassis, but as-shipped from the
surplus vendor there was no heat sinking provided for its
baseplate. In order to avoid thermally stressing the unit
- especially at higher ambient temperatures - it is necessary
that it be operated
only with a heat sink.
Shown in
Figure 1 is
the packaged unit, contained in a die-cast aluminum enclosure -
a Hammond 1580D series. Mounted to the lid, the aluminum
cover alone seems to adequately spread and radiate the heat from
the unit's baseplate and when assembled, much of the lid's heat
is also transferred to the body of the enclosure: In
original testing of the LPRO-101 without using a heat sink it
was noted that one part of its baseplate got significantly
hotter than the rest, but with the aluminum lid, this "hot spot"
was considerably reduced as the heat was more evenly distributed
over a larger area. In testing, the lid stayed well with
the unit's specifications when operated at normal "room
temperature" but if long-term operation at elevated temperatures
was anticipated, better heat-sinking should be considered.
Don't forget, too, that running the unit at its rated minimum of
19 volts goes a long way toward reducing heat dissipation.
To maximize heat transfer, it is necessary to make sure that the
inside surface of the box's lid is smooth
(one must grind
off stamping or mold marks and, preferably, remove any paint)
and mates well with the baseplate and that heat-sink compound is
also used. A total of six 4-40 machine screws were used in
the holes provided to bolt the lid and baseplate together.
Using an infrared "non-contact" thermometer it was determined
that the temperature anywhere on the baseplate was well within
the manufacturer's specified ratings.
The wiring - both DC and RF - is connected via a series of pins
on the side. While connectors of this type are readily
available, I simply fabricated something using a cut-down
IDC-type connector from a discarded computer cable, soldering
wires to its pins and stabilizing them with thermoset adhesive
to prevent them from breaking off during handling.
Power supply:
The LPRO-101's specifications note that the allowable voltage
range for the unit is from 19 to 36 volts, but because linear
regulators are used within the LPRO-101 higher voltages result
in more heat being dissipated, higher power consumption, and
higher battery drain. For this reason, I chose to operate
it from the lowest-possible recommended voltage - 19 volts -
which also meant that in order to run it from a 12 volt power
supply (such as a battery) a power converter was required.
A cheap and easy means of up-converting the voltage is to use
the National Semiconductor LM2577 "Simple Switcher" (tm).
This chip can handle several amps and contains most of the
circuitry necessary to perform the voltage conversion, requiring
only a few external components. While "pre-built" units
using this same (or similar) chip are easily found on EvilBay, I
chose to build my own.
This chip works by momentarily shorting the "diode end" of the
100 uH inductor, L103, to ground, causing current to build up
and a magnetic field to be established and then "un-shorting"
the inductor - a process that is repeated at a frequency of
about 52 kHz. When L103 is "un-shorted" by the chip, the
magnetic field collapses and the resulting current "pushes"
against the input voltage and C102 on one end of the coil while
the current from the other end dumps through D103 and charges
the output capacitor, C104. Theoretically, the voltage
produced by L103's collapsing field would be extremely high, but
this is quashed by the charge transfer to C104 and it is with
this energy transfer to the output capacitor that one can effect
an efficient conversion of a lower voltage to a higher
one. When the voltage on the "FB" pin exceeds the
threshold, the switching of the regulator's oscillator is turned
off, allowing the voltage to drop: The chip's
enabling/disabling the oscillator appropriately by monitoring
the voltage allows it to effectively regulate the output
voltage.
|
|
|
|
Figure 2:
Top Left:
Inside the enclosure containing the 10 MHz Rubidium
frequency reference. The Efratom LPRO-101 unit is
partially visible in the background, mounted to the
lid. The LM2577-12 19 volt up-converter can be seen
soldered to the circuit board mounted to the wall of the
box. A closer view of the voltage converter may be
found here.
Top Right:
The 10 MHz rubidium reference, showing the LPRO-101 unit
mounted to the lid of the box and its interconnect
cabling.
Bottom Left:
A close-up view of the distribution amplifier and status
indicator/driver circuit.
Bottom Right:
Schematic of the power supply and interface portions for
the Rubidium reference.
Click on an image for a
larger version.
When ordering parts I found that the desired component - the
"LM2577-ADJ" - was unavailable from the vendor, so I got the
"LM2577-12" instead. These two parts are identical except
that the "-12" version has connected to the "FB" pin some
internal scaling resistors, setting a 12 volt output:
What this means is that instead of a minimum voltage of 3.5
volts or so for this "ADJ" version,
this part's
minimum
output voltage is 12 volts - but since we need 19 volts,
potentiometer (R103) is adjusted to provide 12 volts at the "FB"
pin when 19 volts was present at the output. (
I could
have gotten the "LM2577-15" instead, but the lower voltage of
the "-12" version made it more versatile as it allowed a wider
variety of voltages in other projects.)
The switching regulator's entire circuitry was mounted "dead
bug" on a piece of circuit board material, providing both
low-profile construction (as space inside the die-cast box is
rather limited) and a very solid ground plane - an
absolute
necessity when using a switching regulator.
Doing double-duty, the LM2577 was soldered directly to the
circuit board's ground plane, providing plenty of heat
dissipation: In normal operation, it is just "warm" and
not hot.
To minimize radiation of magnetic fields into other circuits
(which might include the Rubidium box) a toroidal inductor was
used and placed as far away from the LPRO-101 and driver
amplifier circuitry as possible within the confines of the
enclosure.
For power supply bypassing, "Low-ESR" electrolytic capacitors
were used: These types are
absolutely
necessary to provide reasonable filtering and good efficiency as
the use of "ordinary" capacitors will compromise both! The
various components were mechanically secured in place using
silicone adhesive to prevent them from moving around during
transport and breaking away - and to insulate some of the
connections.
The output of the power supply goes through another inductor - a
47uH toroidal unit obtained from a scrapped computer power
supply - followed by another low-ESR capacitor: This
second set of filtering removed residual switching energy from
the LPRO-101's power supply input and if it hadn't been
installed, spectral components of the switching frequency (in
the 52 kHz area) might have found their way into the 10 MHz
output. The inductor and capacitor values for this
additional filtering aren't critical, but I'd not go any lower
than the 220uF capacitor and 47 uH inductor values noted.
Remember: Any "grunge" that appears on the output of your
frequency reference - which may come from the switching
converter - may be multiplied
many-fold at the
local oscillator frequency.
Comment:
- When first powering up the voltage converter, test it without
connecting it to the LPRO101 and preset the voltage adjust
control (R103) so that its wiper is farthest
from the ground end (e.g. the lowest
voltage.) Note that if the wiper of R103 is set
to ground, the voltage converter will attempt to output the
highest voltage which could be as high as 60 volts.
- Unless a load is connected (the "Go/No-Go"
indicator circuit described below should suffice) the
voltage output could be slightly high and it may take a few
seconds for it to go down when R103 is adjusted to reduce
the voltage, requiring the output capacitors to discharge.
- Only after it is verified that the voltage
converter is working properly should you connect it to the
LPRO101.
- Even though this unit uses a switching up-converter to
derive its operating voltage, it has been found to be
"clean" - a word that is not always associated with
switching supplies. This can be attributed to several
factors considered in the design:
- The entire switching
supply was constructed on a solid ground plane.
When designing ANY switching supply it
is absolutely necessary to assure that the ground be solid
and that "ground loops" be avoided. Considering that
there are several amps of peak current present in the
circuit at the switching frequency it's easy to see how
just a few thousandths of an ohm of resistance in a
circuit board trace can cause several 10's of millivolts
of extra "grunge" to appear on the input and output leads.
- Low-ESR capacitors are
used. When constructing a switching supply
only LOW ESR
capacitors especially designed for switching supplies
should be used! These capacitors are somewhat more
expensive than "normal" capacitors, but considering that
you need only a small handful of them in this circuit the
extra cost is minimal, overall. If you try to use
"normal" capacitors (e.g. those only rated for 85C which
are almost never
low-ESR types - and even many 105C capacitors aren't
low-ESR types!) the circuit will not only produce a lot of
"grunge", but the capacitors themselves will probably get
warm and fail very soon.
- Toroidal inductors are
used. Toroidal inductors have the advantage
of being largely "self-shielded" - that is, compared to a
"normal" solenoid-type coil, they radiate a rather low
magnetic field - and considering that it is the magnetic
field that is doing the work to up-convert the voltage,
this can be important! If a strong magnetic field
from, say, a solenoid-type coil were to cut across some of
the other wires or circuits crammed inside this box, they
would likely impart some of the 52 kHz switching energy on
it and spread it around where it wasn't wanted!
- Placement of the
switching converter circuit. You'll notice
that the switching converter was placed as far away from
the Rubidium unit as it could be within the confines of
this small box to minimize the likelihood that some of the
52 kHz magnetic field might somehow enter its
circuitry. The LPRO-101 is magnetically-shielded to
reduce effects from the Earth's magnetic field but that
just lets you know how sensitive it can be to such fields.
- The use of "additional"
filtering. You'll notice that there's
another 47uH choke and low-ESR capacitor on the 19 volt
output from the switcher - and for good reason: Even
with "good" capacitors, there may be a few 10's of
millivolts of 52 kHz "grunge" on the 19 volt output
lead. The use of both inductance and capacitance to
further-filter the output reduced it to a level that I
could not longer see it on my oscilloscope even when set
to the lowest range. I also check the voltage input
lead to see if there was "grunge" there as well that might
be somehow conducted into the 10 MHz distribution
amplifier or into other equipment on the V+ input lead,
but was surprised to see none: If there had been,
it, too, would have gotten its own choke and low-ESR
capacitor.
Overcurrent and reverse-polarity protection is provided on the
Batt V+ terminal with R101, a 3-amp
self-resetting
thermal
fuse and D101, a generic 3 amp (or greater)
diode. While it may seem unlikely that one might connect
the power supply backwards, it
does happen and the use
of a self-resetting fuse such as this will not only protect the
circuit, but will allow the circuit to work once the fault has
been corrected - unlike a traditional fuse that must be replaced
once it has been "blown." I have been using these
inexpensive self-resetting fuses in my projects for several
years now they have been proven to be both reliable and a real
"equipment saver".
10 MHz Distribution amplifier:
In order to be able to drive more than one external device an
LM7171 high-speed, high-output op amp is used to buffer and
amplify the 10 MHz signal. This op amp is mounted on its
own board, located as far away from the switching converter as
possible, and three outputs are provided. While 82 ohm
resistors are shown for R109-R111 in the output legs of the
amplifier, they were only used because - at the time that I
built it - I was very low on resistors in the 47-75 ohm range,
but had
plenty of 82 ohm units! Unless the
cable run is
very long, it is unlikely that a somewhat
mismatched source impedance would have any significant effects,
anyway.
If you cannot find an LM7171 op amp, there are other types that
may be substituted - or one could also use some emitter-follower
buffer amplifiers as well
- see the (this
link) describing the use of the FE-5680
Rubidium for more information.
A "Go, No-Go" status indicator:
Of the several signals output by the LPRO-101, one of them - the
"BITE" (Built-In Test Equipment) signal, noted above - is the
most useful for typical operation. This signal, when
"high" indicates, that an error condition is being detected by
the LPRO-101's internal circuitry. While this indication
can also mean that the unit is still warming up, it could also
indicate that its supply voltage is too low, or that the unit
itself has failed.
Any time this "BITE" signal is high one should not trust frequency
output of the unit to be accurate. While the
unit is warming up its frequency output should slowly sweep
back-and-forth around 10 MHz as it searches for lock from the
"physics package" - a fancy word for the magical Rubidium
lamp. Once the lamp comes to temperature and it can detect
an atomic resonance, it will suddenly "snap" to frequency.
If this "BITE" signal is high, Q101 is turned on which turns
Q102 off, which allows current through R115 to illuminate the
"red" portion of the dual LED, D102, indicating an "error"
condition. If the "BITE" signal goes low, Q101 is turned
off, allowing current through R113 to flow into the "green"
portion of the dual LED and turn on Q102 which, in turn, shuts
off the "red" LED:
This "green" indication
signifies that the unit is operating properly and can now be
trusted to provide a reasonably accurate and stable
reference.
The indicator circuit is powered from the same 19 volt line that
powers the LPRO-101 and with the 12 volt Zener diode in series,
the LED will be dark if the voltage converter has failed.
If desired, the 12 volt Zener may be omitted, but if this is
done the LED will still be illuminated even if the 19 volt
supply quits it would be just one "diode drop" below the supply
voltage rail - which would probably be around 12 volts or
so. The only obvious condition that is
not
readily detected by the LED's status is a failure of the output
amplifier - but in that case you wouldn't have any 10 MHz
output, either.
At the moment the LED turns green, the frequency will be "pretty
darn close" (and probably better than an already-warm
quartz-based reference!) and after 20-30 minutes it should
achieve something close to its rated accuracy - assuming that it
has been adjusted properly and that it is being operated under
environmental conditions similar to those under which it was
calibrated.
Power supply input
filtering/protection:
Finally, the input supply is RF-bypassed using a feedthrough
capacitor and C101 to prevent the ingress or egress of
extraneous RF along the power lead as well as conduction of
switching supply noise along that line. For
power-supply short-circuit and reverse-polarity protection,
R101, a 1.1 amp, self-resetting PTC fuse is used in conjunction
with D101 making the unit nearly fool-proof in the field.
Using the LPRO-101 as a 10 MHz reference for
microwave transverters.
One of the popular uses of surplus rubidium frequency references
is as a precision 10 MHz reference for use with transverters as
used on the microwave amateur bands - such as 10 GHz.
Unfortunately, not all rubidium units are created equal.
For microwave operation, I have three frequency references that
I can use: A crystal-based unit using an IsoTemp OCXO (
described here),
this one, using the Efratom LPRO-101 and another rubidium
reference using the FE-5680A (
described here).
While the OCXO (IsoTemp) unit and LPRO-101 units both work
nicely as 10 MHz references at 10 and 24 GHz, this same could
not be said for the "barefoot" FE-5680A. As described on
its page, this unit -
while being very accurate - output some very low-level
phase-modulated spurious components that, when multiplied
1000-fold to 10 GHz, caused highly objectionable degradation to
transmitted and received signals: I had to add a
"clean-up" oscillator to the '5680A to make it usable for as a
microwave reference.
The upshot of all of this? Not all oscillators are created
equal! While the '5680A appears to be as accurate and
stable as the LPRO-101, it's not as "clean", out of the box.
For an article comparing various frequency
references/sources, see the article "Stability and Noise
Performance of Various Rubidium Standars" by John Miles,
KE5FX, below.
Other links pertaining to Rubidium references:
- The KO4BB
Wiki - This site as a lot of information on
various Rubidium references and "precision timing"
equipment, among many other things!
- Rejuvenating
Rubidium Lamps - The rubidium lamp in these
devices has only a finite lifetime, but this page explains
how you may be able to get more life out of
it if it quits working! Note that this page doesn't
address the LPRO-101 specifically, but the same general
technique may be applicable.
- The
"Time Nuts" Mailing list and archive - Covering all
sorts of nerdy topics related to frequency and time
measurement, the archives of this list contain a wealth of
information about this and other frequency references.
While anyone may peruse the archives, you must join the list
in order to participate.
- Performance
of Low-Cost Rubidium Standards by John Ackerman, N8UR
- This article compares a number of low-cost (e.g. surplus)
rubidium units to determine best short and long-term
stability. The winner? The Efratom LPRO-101.
- Stability and Noise
Performance of Various Rubidium Standards by John
Miles, KE5FX - Another excellent article comparing the
important parameters of various Rubidium devices available
on the surplus market. From this page you can readily
see why the LPRO-101 works "barefoot" as a microwave
reference and an FE-5680 does not.
The usual warnings:
- As with any electronic project, anything described
or referenced above should only be done by those
familiar with the techniques involved.
- Please observe all safety precautions when dealing
with voltage, heat, or potentially dangerous materials
such as Rubidium.
You have been warned!!!
Return to the KA7OEI Microwave
page.
Any comments or questions? Send an email!
This page and its contents copyright
2010-2016 by Clint, KA7OEI. Last update:
20160720