[time-nuts] FE5680 GPS Disciplined Controller

EWKehren at aol.com EWKehren at aol.com
Wed Jun 18 14:49:02 EDT 2014


 
FE5680 GPS Disciplined  Controller
With all the FE  5680 rubidium oscillators being used as door stops out 
there some of us decided  to develop a GPSDO for it. The main question we have: 
Is there sufficient  interest among time nuts for a discipline controller 
for the FE5680 to make it  available? Looking at the postings over the last 
two years I am not so  sure.
The  construction and preliminary testing of a Brooks Shera style GPS 
discipline  controller for the later version (6.81e-13 resolution) of the FE5680 
has been  completed. We are trying to determine the number of people that 
would be  interested in obtaining an FE5680 discipline controller (if there is 
sufficient  interest about $45 a kit shipping included, $75 for an 
assembled and tested  board, international orders for an additional $5) when it is  
released. 
We are also looking for three Beta  testers that would be willing to 
purchase, assemble, and test our Beta release  controller kit with their own 
FE5680A and GPS receiver or Tbolt and provide  feedback. Please send an email to 
_EWKehren at aol.com_ (mailto:EWKehren at aol.com)   Subject Time-Nuts FE 5680A, 
if you would be interested in being one of the three  Beta testers. A key 
requirement is the willingness to get to it right away, the  board assembly 
takes about 30 minutes. Instrumentation to measure results is  also a 
requirement. We obtained impressive results using a cheap ublox 6M  receiver.

The  FE5680 GPS discipline controller is a small (2” x 2”) board using 8 
DIP’s and 1  SOT23-5 package powered by +5v with 0.1” headers for all inputs 
and outputs. Our  plan is to have the kit supplier solder in the only SMD 
device on the board. A  GPS receiver 1PPS and 10 MHz sine from the FE5680 
feed the board with two 9600  baud serial ports sending TTL level tuning 
commands to the FE5680 and receiving  commands from and sending status data to a 
PC for data logging and system  control via a simple terminal program.  
In the chip count are  two opto couplers that allow the use of isolated TTL 
to USB conversion. These  USB adapters are readily available and furnish 
the 5 V necessary for the  secondary of the opto circuit. An option is to not 
use the opto couplers and  send the PIC TTL level RX and TX into a TTL to 
RS232 adapter. Another option is  to use a TTL to RS232 converter after the 
opto couplers but then an external 5 V  source would have to be supplied for 
the opto couplers.  
As I  mentioned before to get best performance from the FE5680 temperature 
control is  a must and after much fan and metal work I realized that a Lap 
Top heat pipe is  the easiest lowest cost solution. Comments appreciated. As 
an alternative the  temperature correction needs to be disabled.  Otherwise 
two control loops fight each  other.  If you look close on page 7  of the 
brochure temperature stability from –10 to +60 C looks good but a closer  look 
and you see 4 E-11 changes over small temperature changes in the -10 to 60  
C range. Extensive analysis has been done on the FE 5680 A and maybe some 
one  can tackle that problem. Please look at what N5TNL did. It is attached 
and click  on his link. The FE 5680A does have a 4 channel MAX 1246 ADC and 
most likely it  is used to monitor temperature. 
Also  mentioned before the FE 5680 output is not the cleanest, I did 
observe it and  some one posted the attached. I apologize but my records do not 
show who did, so  if you posted the data please come forward. For serious 
applications where you  are using it as your main reference a clean up like the 
Morion MV89 or HP 10811  should be considered. 
This  addition is not required for beta tests but temperature control will  
help. 
I am also  enclosing the express PCB layout, be free to use it but it would 
be more  economical to do a group buy if there is enough interest and some 
one steps up  to kit. 
Bert  Kehren 
To  not exceed the attachment limit the plot will be a separate  posting 





-------------- next part --------------
A non-text attachment was scrubbed...
Name: What is the typical drift with temperature.ZIP
Type: application/zip
Size: 31931 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20140618/be0c5f1a/attachment-0001.zip>


More information about the time-nuts mailing list