2
\$\begingroup\$

My plan is to use my old Garmin GPS 12 as a frequency standard. As it has no pps (pulse per second) output I am thinking about using the on board oscillator as the clock source. So my question: Is the 16.368 MHz oscillator that is controlled by the big crystal on the circuit board (see image below) "GPS disciplined" and thus has a high accuracy and stability?
Garmin GPS12 inside with 16.368 MHz crystal

\$\endgroup\$
  • 4
    \$\begingroup\$ Almost certainly not, but I have no reference for this hence a comment. You would be better identifying the ICs and seeing if any of them have an unconnected / unused PPS output which can be repurposed. \$\endgroup\$ – David Nov 18 '15 at 12:34
  • \$\begingroup\$ IF there was such locking to the GPS system's timebase I would expect that that would be done in software because that is much easier than tuning the crystal. The crystal in the photo look like a simple low-cost model so just standard accuracy and stability. \$\endgroup\$ – Bimpelrekkie Nov 18 '15 at 13:04
  • \$\begingroup\$ I could not find any information about the Garmin ICs on the board, but found out that this type of oscillator if usually not GPS locked - see my answer below. I scrapped the project anyway after I found a cheap GPS module with locked timebase on the market. \$\endgroup\$ – mark_orion Nov 18 '15 at 17:17
2
\$\begingroup\$

The 16.368 MHz oscillator is not GPS disciplined. It serves as the oscillator for the IF stage of the GPS frontend and clock source for its digitized baseband output.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Whilst this is probably true - do you have any reference or design information on why this is the case? Or even information on how these low cost receivers are designed which hints at design decisions that make this unlikely \$\endgroup\$ – David Nov 18 '15 at 17:25
  • \$\begingroup\$ Wikipedia explains the frequency: Crystal Oscillator Frequencies And This post talks only about temperature compensation And I had a look at the chip design of similar GPS chips and they all use this frequency to create the IF without GPS lock (only PLL, xtal and VFO). If there is a timing signal then its created later when the digitized baseband is channelized and processed in another chip. So no locking but software temperature compensation - the temp sensor is under the crystal. \$\endgroup\$ – mark_orion Nov 18 '15 at 20:12
1
\$\begingroup\$

Crystal oscillators can be pulled to a certain extent, for example by varying load capacitance or --- for oven controlled types --- by applying a voltage to a control input. A gps receiver can therefore --- in theory --- pull its oscillator to the nominal frequency after determining the user clock error (a by-product of the position solution). Some integrated RF frontends also have registers to tune the load capacitance on their clock input for exactly this reason.

I am not aware of any consumer grade receiver product that actually does that. Especially for old designs like the GPS12, I would almost rule out that they do this. The clock error is much easier to compensate for in software.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.