Outline Requirements for a DIY GPSDO

[iMo]

Here is a concept which is even simpler construction than the other MCU based GPSDOs above (except the XOR one).
With modern parts (the TDC7200 55ps TIC), fast MCU and a good DAC it may work pretty well.

https://www.eevblog.com/forum/projects/lars-diy-gpsdo-with-arduino-and-1ns-resolution-tic/msg2785452/#msg2785452

User "thinkfat" wants to elaborate with rubidium (and he wants to start a a new thread with his findings).

PS: with faster MCU (with more ram) you may set the NEO's 1PPS output to say 1kHz and thus gather 1000 phase diffs per second and then do some statistics between the DAC periods.

[FriedLogic]

If a TDC with quite good resolution is being used, it might be as well to just use the 1PPS with the quantization error correction - assuming that a correction message available. It also reduces the chance of ublox 20.8ns jumps causing complications.

[FriedLogic]

So, for a reasonable OCXO to experiment with, is the Isotemp 10 MHz OXCO 134-10 12 VDC Sine Wave Output any good?

The short term stability of  1 x 10^-10/second seems a lot.

An Isotemp 134-10 should work if you have one, but I probably wouldn't bother otherwise. The TC on the ones that I tried was in the region of 1E-10/°C, which is also not brilliant.
Something like an Oscilloquartz 8663 which has a double oven and low ageing and TC would allow longer time constants to be used than a basic OCXO, but either should work.
Also, as often get mentioned here, a lot of surplus oscillators have issues and may not work correctly, so it's best to have more than one.

I don't know if a digital pot would work here or not. A good quality multi turn trimmer should be OK, at least until you know exactly what values are required. Wire wound ones can be better, but old surplus ones can sometimes have a high and variable wiper resistance.
I'm always a bit cautious with pots because of the mechanical and electrical issues around the wiper, but they can work well.

https://www.bourns.com/data/global/pdfs/OnlinePotentiometerHandbook.pdf

https://www.bourns.com/docs/technical-documents/technical-library/trimmers/application-notes/bourns_wiper_setting_apno.pdf?sfvrsn=35e9c48f_0

[Gandalf_Sr]

If a TDC with quite good resolution is being used, it might be as well to just use the 1PPS with the quantization error correction - assuming that a correction message available. It also reduces the chance of ublox 20.8ns jumps causing complications.

I know you've mentioned the 'TDC' and I get the idea that it smooths out the jitter on the 1pps signal but can you please point me to where it's explained in more detail?  I don't even know what TDC stands for (other than top dead center when aligning your engine's timing belt).

[FriedLogic]

Time to Digital Converter. imo mentioned the TI TDC7200.
TIC is Time Interval Counter.
Similar things, but I think that TIC is more the piece equipment, and TDC is more the circuit inside it that does the conversion, although both names often get used either way.

[FriedLogic]

The TDC is just for measuring the time interval between the 1PPS pulse from the GPS and a clock or 1PPS derived from the oscillator. The 4046 circuit and ADC that Lars used in his GPSDO is also a time to digital converter.

Most GPS receivers use their own system clock to generated the 1PPS and other frequencies, and pick whichever clock cycle is closest. There will usually be an error between that and the correct position, and this is the Quantization Error which many receivers will report.

This shows the effect of the quantization error on a Motorola M12:
http://www.leapsecond.com/pages/m12/sawtooth.htm

[thinkfat]

If you're looking into buying an OCXO, Morion MV85 looks quite good. 5V powered, Allen deviation @1s < 5e-12. That's quite decent.

[Gandalf_Sr]

Thanks, I just ordered one.

[Gandalf_Sr]

Well my Morion MV85 5V looks like it's in town and may arrive today.  I have an article from the Cypress website on how to get greater DAC resolution by using 2 DACs, one as coarse and one as fine; I may experiment with this over the next couple of weeks - any suggestions for this or other aspects of the control circuit?

[Edit] It arrived - it's smaller than I thought it would be, 23 x 23 x12 mm

[Gandalf_Sr]

So now my Morion MV85 5V has arived, I start to think about the basic building blocks for my GPSDO. Lar's GPSDO thread here seems to have morphed in multiple directions leaving me confused about what the best way to go is.  I know 'best' is debatable but I'm looking for a summary of what those who have trodden this path before me have figured out - if possible a schematic would be awesome.

Happy holidays to you all.