FE-5680A FAQ v0.1

--------------------------------------------------------------------------------
Frequency Electronics FE-5680A Rubidium Frequency Standard FAQ
culled from "time-nuts" mailing list, and web sites.

CAUTION! OLD, OUT OF DATE, and UN-MAINTAINED!
Please instead visit the current FAQ at
http://www.ko4bb.com/dokuwiki/doku.php?id=precision_timing:fe5680a_faq
--------------------------------------------------------------------------------

NOT MAINTAINED.     Please instead visit the current FAQ at
http://www.ko4bb.com/dokuwiki/doku.php?id=precision_timing:fe5860a_faq


CONTENTS

What is a FE-5680A and why would I want one?
Useful Links (web pages, official manual)
Specifications of the FE-5680A
Connector pinout
Why doesn't the C-field adjustment trimpot through the hole in the side do anything?
What is the available range of frequency adjustment via RS-232 commands?
What if my 5680A output does not lock up after several minutes?
Can I get a square wave 10 MHz output instead of a sine wave?
How can I get a 1 PPS output?
What is the typical drift with temperature (tempco)?
Long Term Frequency stability
Input Voltage Requirements
Is a Rb frequency standard sensitive to external magnetic fields?
Mechanical
Undocumented RS-232 commands (read temperature, VCXO voltage)

--------------------------------------------------------------------------------
What is a FE-5680A and why would I want one?

The FE-5680A is a Rubidium Frequency Reference made by FEI (Frequency Electronics, Inc)
www.freqelec.com. As of January 2012 surplus units were available cheaply (<$50) on
auction sites. The units were probably pulled from cellular base stations, which use Rb
standards to keep transmitter frequencies within spec.  Rb standards can provide a
stable frequency to parts in 10^-11 or better (over some time interval), and now
a surplus 5680A may be the least expensive way to achieve that level of accuracy.

Rb frequency references are used wherever a stable frequency, time or time interval
is needed. One example is narrowband radio communication links at UHF or microwave
frequencies, another is simply an accurate clock independent of GPS or external signals.
A Rb standard can also be corrected over the long term (weeks or months?) by GPS, in that
case it becomes part of a GPSDO.

--------------------------------------------------------------------------------
USEFUL LINKS
--------------------------------------------------------------------------------

Time Nuts mailing list archives:
http://www.mail-archive.com/time-nuts@febo.com/info.html
http://www.febo.com/pipermail/time-nuts/

Frequency Electronics official FE-5680A product page
http://www.freqelec.com/rb_osc_fe5680a.html

FE-5680A Technical Manual
http://www.wa6vhs.com/Test%20equipment/FREQUENCY%20STANDARDS/FE-5680A/5680%20TECH%20MANUAL.pdf

FE-5680A Notes, links, and control software:
http://www.vk3um.com/Rubidium%20Standard.html

"Further Information on Rubidium Std. FE-5680A"
http://www.vk3um.com/Fe-5680A4.pdf

Extensive information page, packaging for ham radio use
http://www.ka7oei.com/10_MHz_Rubidium_FE-5680A.html

Fhotos of unit with some signals marked
https://plus.google.com/photos/109928236040342205185/albums/5680473650837554113

Misc FE-5680A photos
http://www.yeyudo.cn/article.asp?id=230

Atomic Nixie Clock project, using 5680A as timebase
http://www.amug.org/~jthomas/atomicnixie.html

Manual for Datum/LPRO-101 Rb reference, a bit different but has general theory of operation, and usage details
http://www.ko4bb.com/manuals/index.php?dir=05%29_GPS_Timing/Datum/LPRO

How to repair a Rb discharge lamp (heat gun restores precipitated Rb to usable form)
http://www.n4iqt.com/efratom8084008/eeratom_model_frs_lamp_assembly_repair.pdf

=================================================================
NOTE: there are MANY different varieties of "FE-5680A" with many options so any specific
feature, output, or repair procedure may not apply to your unit. See for example,
the options listed in the FE-5680 Technical Manual.

=================================================================
SPECIFICATIONS from FE-5680 Tech Manual Table 3, Sheet 9 (PDF page 10/19):

PARAMETER SPECIFICATION
Frequency 10 MHz
Type Sinusoidal
Amplitude (minimum) 0.5 Vrms into 50O(+7dBm)
Adjustment Resolution <1 x 10-12 over range of 3.8 x 10-5
C-field potentiometer Resolution 1 x 10-11 over range of 3 x 10-9
Drift:
	2 x 10-9/year
	2 x 10-11/day
Short Term Stability: 1 sec = 100 sec 1.4 x 10-11 t
Retrace 5 x 10-11
Phase Noise (fo=10 MHz)
	@ 10 Hz: -100 dBc
	@ 100Hz: -125 dBc
	@ 1000 Hz: -145 dBc
Input Voltage Sensitivity 2 x 10-11/(15V to 16V)
Frequency vs.Temperature (-5C to +50C) ±3 x 10-10
Spurious Outputs -60 dBc
Harmonics -30 dBc
Loop Lock Indication
	> 3Vdc=Unlocked
	< 1Vdc=Locked
Input Power (@25 C) 11 watts steady state, 27 watts peak
DC Input Voltage/Current
15V to 18V @ 1.8A peak and 0.7A steady-state
except Opt 25: +22V to +32V @ 1.25 peak, 0.5A steady-state
Ripple +15V: < 0.1Vrms
Warm-up Time < 5 minutes to lock @ 25C
Size: 25 x 88 x 125 mm, .98 x 3.47 x 4.92 inches
Weight 434 grams, 15.3 oz.

Full manual available here:
http://www.wa6vhs.com/Test%20equipment/FREQUENCY%20STANDARDS/FE-5680A/5680%20TECH%20MANUAL.pdf

=================================================================
Connector Pinout

        pin 1: +15V input (1.7A max when cold starting, 0.6A typ steady state)
        pin 2: GND (15V return)
        pin 3: LOCKn (low=locked, high=unlocked) ACT240 output
        pin 4: +5V input (80mA typ.)
        pin 5: GND (signal)
        pin 6: 1pps (about 1us positive pulse each second)  (some units only?)
        pin 7: 10MHz sinewave (~1Vpp on 50 ohm)
        pin 8: RS-232 RX (receive commands into unit)
        pin 9: RS-232 TX (unit sends responses to pc)

from Jose Camera Apr 20 2011
http://www.febo.com/pipermail/time-nuts/2011-April/056256.html

=================================================================
Why doesn't the C-field adjustment trimpot through the hole in the side do anything?

Some versions of the 5680A (most/all available on auction sites as of Jan. 2012) have this circuit disconnected. So, it doesn't do anything. However, you can program the frequency offset via RS-232 serial commands though, see "control software" link above.

=================================================================
What is the available range of frequency adjustment via RS-232 commands?

The offset setting is a 32 bit signed integer, with a full range of adjustment of +/- 383 Hz. Each individual step is 1.78E-7 Hz, or 1.78E-14 relative to 10 MHz. This is intended as a fine trim control, and the unit may not even lock at the extreme settings. The frequency output has also been reported to become less stable as it is set farther away from 10.000000 MHz.  The DDS chip used is Analog Devices AD9832, see also http://www.analog.com/en/rfif-components/direct-digital-synthesis-dds/ad9832/products/product.html

FREQUENCY ADJUSTMENT 2-3
Introdution 2-3.1
The FE-5680A output frequency can be adjusted digitally over the RS-232 interface (pins 8 and 9). This feature is available as option 2, and is not available on units purchased without this option. The frequency can be adjusted with a resolution of 1.7854E-7 Hz. For an FE-5680A device with an output frequency of 10 MHz, this corresponds to a relative frequency setting resolution of 1.7854E-14.

In order to perform adjustments to the FE-5680A over the serial interface, commands conforming to the protocol described in this section must be sent. The signal levels must conform to the RS-232C requirements. Commands are sent to the FE-5680A using the TX line (pin 9), and response from the FE-5680A are received on the RX line (pin 8). The TX and RX signals are referenced to system ground, pin 5. [...]
[ from 5680A Tech Manual, Sheet 16 (PDF p. 17/19) ]

=================================================================
What if my 5680A output does not lock up after several minutes?
John Beale, Dec. 2 2011

Within 5 minutes after powering up (apply +15V on DB9 pin 1 and +5V on pin 4) the unit should indicate lock (pin 3 voltage drops low).  If it does not, the internal VCXO frequency may have shifted enough so the loop does not pass through 10 MHz and achieve lock. If you have a frequency counter, look at the 10 MHz output signal before lock to see if it sweeps through 10 MHz or not. One of my units would only reach 9.99998 MHz as I received it, but I fixed it with C217 adjustment, as below.

Open up the top cover by removing screws (some hidden underneath labels on top)
Gently move aside the insulating foam blocks (they are fragile).
Rotate trimmer cap C217 slightly with a screwdriver.
Power up again and wait 5 minutes to see if it locks. Repeat as needed.
See also these illustrations:
https://plus.google.com/u/0/photos/109928236040342205185/albums/5680473650837554113/5680683008490223330
https://plus.google.com/u/0/photos/109928236040342205185/albums/5680473650837554113/5681715799377076466

=================================================================
Can I get a square wave 10 MHz output instead of a sine wave?

Yes, by removing the bottom plate, lifting one end of a 15 ohm resistor on the Xilinx XC9572XL CPLD pin 49 (10 MHz square wave output), and installing a cable to bring out the signal.  See J.Beale's modification shown here:

https://plus.google.com/u/0/photos/109928236040342205185/albums/5680473650837554113/5693296809976448530
https://plus.google.com/u/0/photos/109928236040342205185/albums/5680473650837554113/5693295034006787730

The CPLD is a frequency divider, which takes as input a 60 MHz 2.8 Vpp sine wave input from the VCXO, which goes into CPLD pin 64. Pin 22 has a 30 MHz square wave. Some FE5680A units reportedly have a 1 us wide 1 PPS (1 pulse per second) signal on the DB-9 pin 6, but only when locked. It is unknown if this comes from the CPLD or another chip.

=================================================================
How can I get a 1 PPS output?

Several time-nut list members report a 1 microsecond wide, logic level 1 PPS signal appears on pin 6 whenever the unit is locked (hard to see on an analog scope, should be easy on a digital scope)
I have 3 units which as far as I can tell, do not generate this output.

However it is easy to get a 1 pps signal from the 10 MHz output with a single picDIV chip (possibly with a 0.1uF cap and two 10k resistors to bias the sine wave to 1/2 the PIC power supply voltage).

The "PD11" flavor of Tom Van Baak's picDIV (based on PIC12F675 cpu) takes a 10 MHz clock input and generates three different 1 Hz outputs: a short 100 µs pulse is output from pin 7, a 10 ms pulse from pin 3, and a 1 Hz square wave from pin 6.   A low pulse on Pin 4 (ARM/RUN) stops the counter, and a rising edge on Pin 5 (SYNC) restarts it.
See also: http://www.leapsecond.com/pic/picdiv.htm

=================================================================
What is the typical drift with temperature (tempco)?

The official spec says +/- 3E-10 from -5C to +50C (5680 Tech Manual, Table 3
One of my units shows -7E-12 frequency change per degree C increase in case temperature (measured at case temp. of 45C and 55 C). For this reason it is beneficial to stabilize the temperature, for example with a thermostatically controlled fan blowing on the unit's heasink. See also
https://plus.google.com/u/0/photos/109928236040342205185/albums/5680473650837554113/5685636673930493138

=================================================================
Long Term Frequency stability
Bert Kehren, Jan 17 2012

After eight weeks of monitoring to 1 E-12, I still see no aging. Waiting for a change so I can do other tests. One thing I clearly see is a 4 Hz filter response changing the output by +- 3 E-11. It may be more, will have to find a  way to check it more accurately, because of limited response time.

=================================================================
Input Voltage Requirements (official spec is 15 - 18 volts DC)

from Pete Bell, Jan. 18 2012:
One of  the regulators (the one that runs the lamp and lamp heater) is
running at  13.8V and has a 500mV dropout voltage, so there is very
little (~ 700mV)  headroom. The one that runs the cell heater is about
a volt lower and the  one that runs the analog circuits is set to 8.8V,
so has plenty of  headroom.

I suspect the lamp regulator is mostly there to reduce the  start-up
current, though - the FE-5650 has a very similar lamp circuit with  no
regulator and the only significant difference seems to be that  the
current pulled when the unit is cold is  higher.

from Ian, Jan. 19 2012:
With 15.5V DC input power it locks in about 3min.
I backed off the supply to 10.8V and the lock time (from cold) is 8min.

======================================================
Is a Rb frequency standard sensitive to external magnetic fields?

Yes, in fact an internal magnetic field is used to fine-tune Rb standards. The LPRO-101 manual (another Rb unit) mentions sensitivity to DC and AC fields, but "under 2 gauss should not result in measureable permanent frequency offsets for LPRO."  The FE-5680A case is made of a magnetic shielding metal, so it should not be removed (for better heatsinking, etc.) unless the heatsink itself is also magnetically shielded, or the unit is installed away from all stray fields (which is very difficult in most locations, since any electrical current generates a magnetic field).  In particular, power supplies and cooling fans may generate strong nearby magnetic fields.

=================================================================
Mechanical

Size: 25 x 88 x 125 mm, .98 x 3.47 x 4.92 inches
Weight: 434 grams, 15.3 oz.
Some surplus units, which come screwed onto a PCB, are attached with 1/16" hex screws

======================================================
Undocumented RS-232 commands
from Scott Newell N5TNL, Jan 18 2012

I get a response from the unit to several undocumented serial port commands. In particular, the replies from 0x22, 0x57, 0x59, and 0x5A seem to vary each time I read the unit. Not all the commands respond each time; there's certainly a bug or two lurking. (I'm probably getting the poor unit's serial input all confused and out of sync.) Anyway, commands 0x57 and 0x59 appear to send back a lot of data: 0x56 bytes!
I'm not yet validating the data length or checksum of the FE-5680 reply.

[later note: Command 0x5A may be reading an internal 4-channel, 12 bit ADC.
 Ch.1: trim pot, Ch.2: supply voltage, Ch.3: temperature, Ch.4: VCXO control voltage
 see also http://n5tnl.com/time/fe-5680a/ ]

Here's an example of the replies to the commands (in hex):
Reply to command 0x22, 0x0D bytes: [22] [0D] [00] [2F] [20] [5B] [29] [00] [A1] [C8] [4A] [1D] [6C]
Reply to command 0x29, 0x09 bytes: [29] [09] [00] [20] [FF] [00] [00] [00] [FF]
Reply to command 0x2B, 0x09 bytes: [2B] [09] [00] [22] [20] [5B] [29] [00] [52]
Reply to command 0x2D, 0x09 bytes: [2D] [09] [00] [24] [20] [5B] [29] [00] [52]
Reply to command 0x47, 0x08 bytes: [47] [08] [00] [4F] [20] [5B] [29] [52]
Reply to command 0x53, 0x07 bytes: [53] [07] [00] [54] [79] [00] [79]
Reply to command 0x57, 0x56 bytes: [57] [56] [00] [01] [20] [20] [5B] [29] [00] [A1] [48] [4E] [1D] [C0] [00] [D6] [72] [85] [6A] [DB] [25] [80] [14] [8A] [BD] [2E] [5E] [86] [CB] [BA] [E5] [CA] [F5] [2D] [0F] [4F] [7F] [D3] [D1] [37] [2B] [2C] [90] [54] [45] [68] [9A] [AC] [AA] [4B] [50] [33] [CF] [9B] [2B] [FD] [BD] [E6] [AD] [6D] [6B] [BA] [FF] [FA] [17] [03] [62] [F4] [E8] [79] [00] [83] [23] [C1] [17] [69] [37] [18] [8E] [E9] [13] [A3] [FB] [1F] [1B] [F8]
Reply to command 0x59, 0x56 bytes: [59] [56] [00] [0F] [20] [20] [5B] [29] [00] [A1] [C8] [4A] [1D] [40] [00] [D6] [72] [95] [6A] [DB] [25] [80] [54] [8A] [BD] [3C] [5E] [86] [C9] [BA] [E5] [CA] [F5] [2D] [0D] [4F] [7F] [D3] [D5] [37] [2B] [2C] [90] [54] [45] [68] [BA] [BC] [EA] [4B] [50] [33] [CF] [9B] [29] [FD] [BD] [E6] [AD] [6C] [6B] [BE] [FF] [FA] [15] [02] [62] [F4] [E8] [D8] [57] [83] [23] [C1] [17] [68] [37] [18] [86] [E9] [13] [AB] [FB] [5F] [1B] [79]
Reply to command 0x5A, 0x0D bytes: [5A] [0D] [00] [57] [00] [00] [40] [08] [9E] [07] [00] [00] [D1]
Reply to command 0x61, 0x09 bytes: [61] [09] [00] [68] [20] [5B] [29] [00] [52]
Reply to command 0x65, 0x07 bytes: [65] [07] [00] [62] [20] [20] [20]
Reply to command 0x67, 0x08 bytes: [67] [08] [00] [6F] [01] [F4] [B8] [4D]
Reply to command 0xF0, 0x0E bytes: [F0] [0E] [00] [FE] [33] [2E] [34] [00] [00] [00] [00] [00] [00] [29]

======================================================