[time-nuts] Plot phase noise spectrum from DMTD measurement?

Stephan Sandenbergh ssandenbergh at gmail.com
Fri Mar 11 20:33:51 UTC 2011


Hi,

Ok some cool advice - this thread is an interesting thought exercise. I'm
going to think about it a some more, but it seems, in comparison at least,
the loose phase-lock technique remains the simplest. Provided you have a
low-frequency spectrum analyser handy.

The sound card idea is clever as well - however, I'd assume one needs to
measure the ADCs clocking oscillator offset, since that will be apparent
when plotting the beat frequency phase (what I mean is that sampling will
then look like another mixing process). What I usually due is  to clock the
sampling system off a clock that's correlated to the clock under test. This
resolves that issue.

However, I'd like to experiment with the cross-correlation idea, since I've
got a setup that will lend itself perfect to that. Maybe I could save myself
some time, with clever post-processing.

Can anyone recommend a fundamental text on the cross-correlation technique?

Regards,

Stephan.

On 10 March 2011 20:53, Bruce Griffiths <bruce.griffiths at xtra.co.nz> wrote:

> Without nulling the carrier the dynamic range of your ADC will limit the
> measurement system phase noise fllor.
>
> With a 300KHz beat frequency the post mixer preamp response need not extend
> to dc.
>
> Bruce
>
>
> Stephan Sandenbergh wrote:
>
>> Thanks,
>>
>> I'm familiar with the designs you posted to measure voltage noise ect. on
>> you home page. These, with some modification, mainly removing the blocking
>> caps, seems that it would do the trick.
>>
>> Cheers,
>>
>> Stephan.
>>
>> On 10 March 2011 10:50, Bruce Griffiths<bruce.griffiths at xtra.co.nz>
>>  wrote:
>>
>>
>>
>>> For conventional phase noise measurements at offsets in the (10Hz, 20kHz)
>>> range one can use a sound card with a low noise preaamp.
>>> Suitable sound card preamps with lower noise floors than Enrico's or
>>> Wenzel's designs can be built using readily available components.
>>> Wider bandwidths ( up to 1MHz or so) are not difficult to achieve.
>>>
>>> Bruce
>>>
>>>
>>> Stephan Sandenbergh wrote:
>>>
>>>
>>>
>>>> Hi,
>>>>
>>>> Cross-correlation a very clever idea! Thanks for the reference - Rubiola
>>>> got
>>>> some good sources of reference on his home page.
>>>>
>>>> One thing though - for a phase-noise kit one will probably need to
>>>> replace
>>>> the ZCD with a low-noise amplification stage of around 80dB to be to
>>>> allow
>>>> sampling at ADC voltage levels?
>>>>
>>>> Cheers,
>>>>
>>>> Stephan.
>>>>
>>>> On 8 March 2011 22:28, Magnus Danielson<magnus at rubidium.dyndns.org>
>>>>  wrote:
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>> On 03/08/2011 07:46 PM, Stephan Sandenbergh wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> I recently noticed something interesting: The DMTD measurement gives a
>>>>>> set
>>>>>> of phase values x(t). From which fractional frequency y(t) is
>>>>>> calculable.
>>>>>> So
>>>>>> now it seems viable to plot the spectrum, Sy(f) and if you scale it
>>>>>> properly
>>>>>> you arrive at Sphi(f). If I'm  not making a gross error somewhere the
>>>>>> math
>>>>>> seems to check out. But, I'm wondering is there a physical reason why
>>>>>> this
>>>>>> isn't valid?
>>>>>>
>>>>>> I have not seen this being done anywhere - so I assume there is.
>>>>>> However,
>>>>>> it
>>>>>> seems possible to plot Sphi(f) for 1Hz<    f<100kHz when having a
>>>>>> vbeat =
>>>>>> 100kHz sampled for 1 second.
>>>>>>
>>>>>> I'm familiar with the loose and tight phase-locked methods of
>>>>>> measuring
>>>>>> phase noise, but am quite curious to know if phase noise from a DMTD
>>>>>> measurement is a valid assumption.
>>>>>>
>>>>>> I would guess that if the frequency domain phase noise measurement
>>>>>> requires
>>>>>> phase-lock then the time-domain measurement requires as well. However,
>>>>>> here
>>>>>> in lies my real interest - two GPSDOs are phase-locked (not to 1Hz,
>>>>>> something far less I know) so can it be possible to measure GPSDO Adev
>>>>>> and
>>>>>> phase-noise using a single DMTD run? Am I making a wrong assumption
>>>>>> somewhere?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>> An architecture not completely different to the DMTD architecture is
>>>>> used
>>>>> in phase-noise kits. Instead of having two sources and one intermediary
>>>>> oscillator is instead there one source and two intermediary
>>>>> oscillators.
>>>>> The
>>>>> oscillators is locked to the carrier frequency rather than an offset.
>>>>> The
>>>>> mixed down signal is then cross-correlated to get the spectrum.
>>>>> Increasing
>>>>> the averaging factor and the spectrum can be suppressed below that of
>>>>> the
>>>>> intermediary oscillators. Since the two intermediary oscillators have
>>>>> uncorrelated noise, the external noise is what correlates over time.
>>>>> This
>>>>> technique is simply called cross-correlation. Such a cross-correlation
>>>>> setup
>>>>> can run very close to the carrier in terms of offsets.
>>>>>
>>>>> In contrast will a DMTD with it's offset frequency be problematic at
>>>>> low
>>>>> offsets since the positive and negative offsets noise will not occur at
>>>>> the
>>>>> same frequency in a DMTD setup. Consider a a DMTD with a 10 Hz offset,
>>>>> pointing a spectrum analyzer on 100 Hz will measure the down-converted
>>>>> average of carrier+(100-10) Hz and carrier-(100+10) Hz, thus carrier+90
>>>>> Hz
>>>>> and carrier-110 Hz.
>>>>>
>>>>> Creating a mixed-mode setup for phase-noise/DMTD will however be
>>>>> possible.
>>>>>
>>>>> So, DMTD as such is relatively limited, but add an RF switch and
>>>>> another
>>>>> oscillator and you get a cross-correlation phase-noise kit.
>>>>>
>>>>> To turbo-charge the phase-noise kit use a quadrature combiner and
>>>>> amplitude
>>>>> adjustment to create a interferometric mixdown, working around part of
>>>>> the
>>>>> mixer limitations. Enrico Rubiola has writen about this approach.
>>>>>
>>>>> Cheers,
>>>>> Magnus
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>> To unsubscribe, go to
>>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> and follow the instructions there.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at febo.com
>>>> To unsubscribe, go to
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>> and follow the instructions there.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to
>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>>>
>>>
>>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>>
>>
>>
>
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>


More information about the time-nuts mailing list