The VCO module I use is the Bliley, 10MHz module. Input is 5V, output is a 10MHz block. Also an input for the voltage control.

I bought them for 20 euro each on Amazon. To be precise:

BLILEY NV47M1008 NVR47M1008 10MHz 5V SC-Cut Square Wave OCXO Crystal Oscillator

Of course there are more options, but the Bliley is not bad at all!

So the Bliley it will be.

I found out that:

- the Bliley needs quite some time to get at rest,.....think of days or even more.

- the module is sensitive to environmental temperature variations.

- It produces quite some heat, and it can easily heat up the inside of the box. 1cm of tempex around it was not enough. The result is that this heat has to be blown away by a fan.

So, my idea was to put the Bliley unit in a wooden box of 18mm thick. I picked wood, because it has a high heat capacity.

1.8 kJ/kgK, so it can absorb a lot of temp-changes without going up in heat to fast. So it works as an additional buffer. Water is the best (4 kJ/kgK), but very impractical. So wood it will be,....around the wood, I glued a second box of tempex, 30mm thick. This is for the isolation of the unit, in order to keep the heat in the unit instead of in the box of the complete GPSDO. This should consume less power, and keeps the box cooler.

Inside the wooden box I also installed a 16bits, I2C temperature sensor, the TMP117.    0.1 C accuracy and a higher resolution than the AD convertor of the NAN0 (10bits), with an analog sensor. And it brings more linearity in the temperature sensing in the system. More of that later.

So this implementation of the OCXO should be much cooler, much less influenced by the environmental temperature, and better controllable by the use of a far better temp sensor. Of course we still need to control the temperature in the GPSDO box, but more of this later.

It looks now like this,....close to finished.