[time-nuts] FE 5680A GPSDO update
EWKehren at aol.com
EWKehren at aol.com
Thu Jul 3 11:47:46 EDT 2014
First let me thank again Scott Newell for all the work he has done. The
links included in this post where made possible by Scott. All three exceed
the limit of time nuts but will give you a better understanding as to what is
going on. First removing the 10 K resistor changes the Rb frequency any
where between 2 to 3.5 E-10 upward. It is a function of the temperature of the
device. In the first link you will see what happened right after removing
the 10 K.
http://n5tnl.com/time/fe-5680a/gpsdo/JK5680A_Test35.xls Allow me to explain
The first collum is time in seconds followed by time in hours. The third
is phase input measured by the loop The loop will try to set it to 14080. We
call that the set point. The next column shows the delta to the set point
followed by the control word send to the unit and its 2's compliment. 2's
compliment since in this case it is mostly a very large negative number
easier to read. Finally the status word which shows a 1 saying it is unlocked
followed by a 0 other options are glitch hold and alpha. the third column
shows which mode. The way it is presently programmed it will advance to Mode 5
and alpha. That would show up here as 15.
Before the modification this Rb showed when settled -124. The 1 pps comes
from a $ 16 ublox M7 no sawtooth correction.. At about 18.5 hours the GPS
filter is activated with a 12 setting and you notice a dramatic change in the
control word. still has a few more steps. What you see is frequency
correction do to influences like GPS Rb and its temperature. Rb has only a
heatsink, no temperature control.
Next this mornings readings.
http://n5tnl.com/time/fe-5680a/gpsdo/JK5680A_Test35c.xls. This is more of the same but you see a dramatic change between
day (left) and night (right) allowing me to check when he is not working.
Summarizing temperature control is not a requirement but the curve will be
flattened if the Rb is held at a constant temperature. Each digit is a
change of 6.8-13. but it is not directly correlated to the output frequency,
what it says it is the control word that according to the filter function
tries to set it exactly to 10 MHz. The plot that I attached so many times may
contribute to the frequency in 1 second increments that most likely is
reduced by disabling the heater current sensing but there may be other sources.
That is why I feel strong about a clean up loop. Those happy with what you
see stop right now. Again thanks to Scott here is a link to the Auxillary
http://n5tnl.com/time/fe-5680a/gpsdo/AUXILLARYBOARDrevA27-04-2014.pdf I did post a picture but to file limitation was not able to
post the schematic. It has room for power conditioning a PIC and two flexible
outpu circuits. First the PIC takes the output of the controller (attached)
and modifies it in such a way that it can drive a FTDI USB adapter to
store the output without PC requirement. The code is presently in PIC Basic
would like to find some one to write PIC code so we can use FTDI and a I2C
display. Scott is looking at it but if some one has the expertise would help.
The outputs can be individually set to 5 or 10 MHz.
A third board is presently in the works by an other time nut. All boards
will be in Gerber. He is working on a temperature control board. All boards
are 5 X 5 cm. What is missing is a clean up filter board. I am looking for
someone that can do it in Gerber, please contact me off list. We have a
circuit but it is on expressPCB and should be modified for this application.
I look at it the following way. More than half the cost of the kit is
taken up by pay pal, shipping kitting etc. Boards are the smallest cost part
and going from one to 4 will only increase the kit cost by $ 5. In my book a
Comments are appreciated
Some may not get this post because AOL for instance flags it as spam
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Typical RS 232 output.doc
Size: 19456 bytes
Desc: not available
More information about the time-nuts