Questions

[Ken.Hendrickson]

EXECUTIVE SUMMARY

 

With the equipment I have, would it be worth while to configure it as a stand-alone time server?

(Even if no, it will be useful as a 10 MHz frequency standard.)

 

DETAILS

 

About 15 years ago, I bought a kit for a GPS disciplined 10 MHz oscillator.  It consists of a Rockwell Jupiter GPS receiver, an Isotemp OCXO (actually a VFO), and a phase locked loop.  The GPS receiver puts out 1PPS, and a 10 kHz reference signal.  The PLL divides down the 10 MHz oscillator by 1000, compares it to the 10 kHz reference signal, and provides an error signal to drive the 10 MHz OCXO on frequency.  The GPS receiver has two serial ports; one can be used for differential GPS input, and the other can be used either for NMEA messages, or a proprietary Zodiac binary message format.

 

I have complete documentation for all of the above.  I will provide this documentation to the gpsd project.

 

This web page shows pictures of some of the components (OCXO, Jupiter GPS, PLL board, PLL board schematic):

http://www.DiyPhysics.com/2012/02/19/d-i-y-gps-disciplined-10-mhz-frequency-standard-gps-based-universal-time-clock/

 

I’m finally getting around to finishing this project, because now I have a need for a 10 MHz frequency standard.  I have options in the way I put the components together.  That is what this email is about.

 

Can gpsd read the proprietary Zodiac binary message format?  If not, I could write that software.  Apparently, the time mark message which precedes the 1PPS pulse is only output in the 9600 bps binary message format, but not 4800 bps NMEA.  (The UTC Time Mark pulse output message provides the number of seconds since the beginning of the week, about ½ second before the 1PPS signal.)

 

What is currently the best way to utilize the 1PPS signal?  GPIO on something like a Raspberry Pi or BeagleBone?  I would be very willing to package a small computer along with the other hardware to run gpsd and ntpd.  What are any accuracy limitations to using GPIO or similar?  At what rate can a GPIO digital input be read?  Is there a better way to read the 1PPS, such as a non-maskable interrupt, to operate at essentially the CPU clock rate?

 

Using the 1PPS signal seems to require being able to read the proprietary Zodiac binary message format.  That is configurable by pulling some pins on the Jupiter GPS receiver high or low.  You can’t have the NMEA standard messages and the Zodiac binary messages simultaneously; it’s one or the other, but not both.  However, if better than 0.5 second clock accuracy could be obtained by other means --- so the binary Time Mark message wasn’t required to identify what second the upcoming pulse was --- then the 1PPS signal could improve its accuracy considerably.

 

One objective I have, if possible, is to have a stand-alone time server with no internet connection.  Is this possible?  What would the gpsd requirements be?

 

Does anybody use differential GPS?  What would be required?  Would differential GPS only improve the geolocation solution, or would it also improve a time solution?

 

I look forward to receiving any replies.  I have already done some preliminary searching of the gpsd-users email list, but not found the answers I’m looking for.

 

Thanks,

Ken

 

  

CONFIDENTIALITY NOTICE: This email and any attachments are for the sole use of the intended recipient and may contain material that is proprietary, confidential, privileged or otherwise legally protected or restricted under applicable government laws. Any review, disclosure, distributing or other use without expressed permission of the sender is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies without reading, printing, or saving.