[time-nuts] Lars GPSDO

[Bob kb8tq]

Hi

The real question is “how worried are you about phase noise?”. Next past that is the 
bandwidth of the EFC input on the OCXO. Once you have those *and* the LF noise 
spectrum of the DAC …. you can work it all out :)

Since the EFC “does FM” and the noise you worry about is PM ( = phase noise) there
is already a 1/F ( = low pass filter) simply because of the physics involved. There is 
no reason to go insane about this. At the same time, it *could* be an issue. 

One classic example of all this is the TBolt. You have an OCXO with a “many ppm” sort 
of EFC. That’s a lot of sensitivity. Cutting it back to a “< 0.1 ppm” sort of range with a couple 
of resistors  would have a big impact on how things work. 

Bob

> On Apr 5, 2019, at 6:04 PM, Jim Harman <j99harman at gmail.com> wrote:
> 
> I see the AD5680 uses a dithering approach to get the last 2 bits, so the
> noise may be a problem. I suppose I could suppress it with  a LPF on the
> DAC output. Also I have no way to measure the effect of any DAC noise on
> the oscillator frequency, so for the time being I am blissfully ignorant.
> 
> I suppose I could AC couple a high gain amplifier to the DAC output and
> measure the noise, or tune an FM receiver to the 9th harmonic of the 10 MHz
> and listen for noise.
> 
> I have successfully hand soldered a 6 pin SOT-23 but the AD5680 is an 8 pin
> chip and thus even more challenging, and breakout boards for them are hard
> to find and/or expensive. I see Digi-Key has a service where you buy the
> chip and they will deliver it soldered to a breakout board for a few $
> extra. I may give that a try
> 
> On Fri, Apr 5, 2019 at 5:21 PM Bob kb8tq <kb8tq at n1k.org> wrote:
> 
>> Hi
>> 
>> If you do decide to play with the AD5680, be careful of the “tone” on the
>> output
>> at about 180 Hz. You might also want to check the noise inside 100 Hz
>> since the
>> spec sheet is of no help there. If the 180 Hz is the third harmonic of
>> something much
>> larger at 60 Hz …. yikes …..  Since there don’t seem to be noise specs on
>> the MCP4725,
>> who knows if things are getting better or worse. I’d bet better …..
>> 
>> That *is* a weird little package ….
>> 
>> Bob
>> 
>> --
> 
> --Jim Harman
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.