[time-nuts] Things your mother (or FEI) didn't tell you about FE series oscillators

Paul Berger phb.hfx at gmail.com
Thu Nov 13 11:10:02 EST 2014

Hi Skip,

Seeing as how you have spent some time poking around inside FE-5680s 
maybe you might be able to help with a problem I have with one.  A while 
back I purchased two 5380As that where reputed to be new they are marked 
with "10MHZ OPTION 57" and would seem to match you description of the 
"Old FE-5680A".  Unlike one I had purchased  earlier they are mounted on 
a piece of aluminium plate about 6mm thick.  One of them starts up and 
works like a champ but on the other one the oscillator that I believe 
feeds the synthesiser, the one on the second card with the thermistor 
soldered onto it, does not seem to start.  I do see a 50MHz (measured on 
a scope so could be the proper 50.255MHz)  signal coming from the 
physics package.  Any idea what I should look for?

I guess one of the alternatives would be to find a parts on that has 
some other problem.

I also noted that on these there is no insulation around the lamp, is 
that normal?  My first one which would seem to match your description of 
the later version does have some insulation around the end of the lamp.


On 2014-11-10 1:24 AM, Skip Withrow wrote:
> Hello time-nuts,
> I have spent a lot of time recently with Frequency Electronics, Inc. (FEI)
> oscillators including the FE-5680A, FE-5680B, FE-5650A, and FE405B and just
> wanted to pass on some of my findings.  Not everything I am including here
> is new, but figured I would include the informattion for the sake of
> completeness.
> As already known, there are a bunch of different variations on some of the
> oscillators.  Here is kind of a quick rundown:
> 1. The 'old' FE-5650A - This unit used a 50.255 MHz VCXO and placed the
> synthesizer outside the physics PLL.  This unit could be programmed over a
> wide range of frequencies and is great if you need an oddball output.  The
> digital board in the unit has the two small pushbutton switches and the
> nearby 5-pin connector for serial communications.
> 2. The 'new' FE-5650A - This unit has the same physical construction, but
> the circuit functionalyity is quite different.  It uses a 60 MHz VCXO and
> the DDS is inside the physics control loop.  The output is divided from the
> 60 MHz.  Many of these units that came from the telecom industry have 15MHz
> output. Unit has serial communications on pins 8 & 9 of the DB-9 connector.
> 3. The 'old' FE-5680A - This unit has a different physical footprint than
> the 5650A, but virtually the same circuit.  Again uses the 50.255 MHz VCXO
> and has the programmable synthesizer that can be tuned over a wide range.
> Circuit board has the same two push botton switches and 5-pin serial
> connector.  Many of these units only have a 1pps output (they came off
> Motorola cell boards).
> 4. The 'new' FE-5680A - This was the popular, inexpensive unit that was
> available a year or two ago.  Unit uses the newer 60 MHz VCXO architecture
> and has serial communications via pins 8 & 9 on the DB-9.
> 5. The FE-5680B - This appears to be a VERY close cousin to the new 5650A
> and 5680A.  It is in the 5680 physical package, but has a compact DB-15
> (think VGA) connector.  Unit only needs single +15V supply.
> 6. The FE-405B - This is a crystal oscillator that is the same physical
> package as the 5650A.  It has a 15 MHz output.  The interface is the same
> as the FE-5650A (except no LOCK signal), including the capability to
> communicate over the serial interface to change its frequency.
> Now for the new information.  Tuning the units with the 60 MHz VCXO has
> been published (see the FEI document that is archived in several places)
> and uses several commands to read and write a 32-bit command value.
> However, only the tuning sensitivity for the FE-5680A has only been
> published that I am aware of.
> FE-5680A - The tuning sensitivity of the FE-5680A is 6.8126x10E-13 for the
> LSB.  The command value runs the full range from 00 00 00 00 to FF FF FF
> FF.  The value is 2's complement with positive values driving the frequency
> higher and negative values lowering the frequency.  Larger values may
> unlock the unit and the positive and negative values may be different
> (depends on how well the VCXO is centered).
> FE-5650A - The tuning sensitivity of the FE-5650A is 1.4900x10E-17 for the
> LSB.  Sounds great, but wait, turns out it does not have any more
> resolution than the 5680A.  It takes a 45,722 step of the command register
> to step the DDS.  So, it retains the same 6.8126x10E-13 actual minimum step
> of the oscillator.  Also, the minimum value of the control register is F0
> 00 00 01, maximum value is 0F FF FF FF.
> FE-5680B - The tuning sensitivity of the FE-5680B is 3.7248x10E-16 for the
> LSB.  Same situation as the 5650A above, but the are 1829 counts per DDS
> step.  Minimum value is F0 00 00 01, maximum value is 0F FF FF FF, same as
> the 5650A.
> FE-405B - The tuning sensitivity of the FE-405B is in the neighborhood of
> 1.143x10E-14.  Minimum value is F7 33 33 34, and maximum value is 08 CC CC
> CC.
> So, if you know how far off frequency your oscillator is divide that by the
> LSB sensitivity, convert the number to HEX, and load it to the oscillator
> (with the right + or - sense).
> Enjoy!
> Skip Withrow
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

More information about the time-nuts mailing list