GPS-Disciplined OCXO
After (a.o.) G3RUH and G4HUP, with PA3GYF additions for monitoring GPS (FIX) status, extra isolation of the (sine) output signals from the PLL, monitoring OCXO heater, monitoring PLL LOCK status, manual frequency tuning (stand-alone OCXO) and stand-alone GPS operation. Includes front and rear panel designs (Schaeffer Front-Panel Designer) and Arduino firmware, both for Arduino Uno and Arduino Micro.
At around 2016, I decided to pick up my RF electronics and HAM hobby again after a decade (at least) absence, and I decided to build a GPS-disciplined OCXO in order to regain lost skills. I wanted a precision frequency reference and had struggled for years with a DCF77 circuit that never seemed to lock. As I was able to purchase a large amount of (HP) test equipment (around 2013), a stable frequency reference turned out to be the ideal restart project for me.
I designed and built most of the circuit in 2016 and early 2017, but never took the time to properly test and document it. As an autumn 2019 winter project, I finished latest testing and modifications, wrote the documentation and submitted everything to github (before that, the repo contained photos only).
The core design of this project, the PLL circuitry including the loop filter, was shamelessly stolen from G3RUH; the distribution amplifier is an unmodified unit (bought) from G4HUP.
My main (modest) contributions have been the extension of the core design with monitoring and control functions, such as monitoring whether or not the GPS receiver has a fix, monitoring the lock status of the PLL, RS-232C interfacing to the GPS receiver, front-panel and rear-panel designs, and Arduino firmware code, including an algorithm for deciding upon PLL-lock from monitoring the OCXO tune voltage.
In addition, I changed (improved?) the interfacing between the OCXO and the PLL by introducing an isolator for separation of the PLL and output signals from the OCXO, and designed and implemented a Low-Noise Amplifier for the OCXO signal in case (well, almost always) the OCXO is insufficient to properly activate the Schmitt-Triggers of the PLL circuit.
My explicit thank-yous to Frank van Vliet for his interest in this project.
The repository is organized as follows:
- doc: Contains documentation; mandatory reading (in particular the 'manual') if you want to rebuilt this project, or use it as a basis for your own design. Most if not all documentation is written with OpenOffice, but I try to keep a set of PDFs in sync, for those that cannot use OpenOffice.
- fpd: Contains the front and rear panel designs (for Schaeffer, www.schaeffer-ag.de, front-panel designer).
- fw: Contains the Arduino firmware; a single file for both Uno and Micro (compile-time switch!).
- pcb: Contains the KiCAD PCB designs for this project. The PCBs were originally designed in KiCAD v3 (well, some v3), and I migrated the crucial ones (part of V1) to KiCAD v4.
- pcb/v1-kicad3: Contains the (original) KiCAD-V3xx PCB designs for all PCBs, including earlier versions that had faults, and including the ones for the Arduino Uno. The KiCAD projects also contain a lot of 3D shape definitions.
- pcb/v1-kicad4: Contains the KiCAD-V4xx PCB designs for the essential PCBs in the V1 unit. It is a subset of the pcb/v1-kicad3 KiCAD projects migrated to KiCAD V4. It does not include previous PCB versions nor the Arduino Uno PCBs. Also, it does not include most of the 3D shapes for the 3D view.
- transformer: Contains details on the transformer used for the Power Supply Unit.
- enclosure: Contains details on the Velleman enclosure (I) used.
Designing and realizing this project has been a true joy!
Currently, my interest has somewhat shifted towards the accuracy of power reading of my equipment, so I do not expect updates soon.
However, it is still on my wishlist to use an ADC with higher resolution and accuracy than the Arduino 10-bit circuits in order to assess PLL lock status. Also, I want to study the PLL status by simply counting the pulses on both inputs over varying intervals, in order to more accurately quantify the PLL locked status (e.g., 10-4 over the last second, 10-6 over the past 10 minutes, 10-7 over the past hour, etc.). I would also love to improve the RS232-C circuit in such a way that it can be used directly as an input to Linux gpsd and ntp/chrony as a PPS-capable time source. At the present time, this does not work. Finally, I would love an Ethernet interface for monitoring and control of the unit.
Unfortunately, at the present time, two hurdles would need to be taken. First, the availability of the required GPS receiver is (AFAICT) close to zero, and I would have to switch to 1PPS use. Second, the memory availability on the current Arduino models is really WAY TOO LOW. I almost forgot what the K symbol means in a computer-memory context.
So, all in all, it might be better to wait for more powerful Arduino models, and perhaps think of a 1 PPS design.
Who knows...
All my parts, both hardware and software, have been released under Apache Commons License V2.
Jan (pa3gyf)