[time-nuts] Re: Tuning a GPSDO loop for optimal disturbance handling

[Bob kb8tq]

Hi

> On Mar 22, 2022, at 12:14 PM, Erik Kaashoek <erik at kaashoek.com> wrote:
> 
> Hi Bob,
> By your advice I went for a 6.5 digit DVM and after logging and plotting the DAC output its clear there are some stability issues in the DAC output. The voltage is wandering around at about the level of frequency wandering observed.

That’s a very typical result as you dig into the DAC ( = the DAC isn’t perfect ). 

> A different supply topology for the DAC and VCXO will have to be created.. The DAC resolution is 50 uV (2.048 V / 400000 steps) but the random variations are about 4 times p-p larger.
> Once this is done I hope longer term logging of frequency and voltage will make sense.
> Or is it better to have a low pass filter between the DAC and the VCXO Vtune input so the loop is fast enough to remove the remaining drift?

You can go to a stable ADC and let it “servo” the DAC. That’s going to add cost and
(likely) some noise from the ADC. The most common outcome is to find another DAC
and possibly a better reference. 

> If so, what should the time constant of the filter be compared to the intersect of the ADEV of the GPS and the VCXO (100 s)?
> Having a big time constant (10s?) will be a pain in the initial tuning and will require active components.
> Does owning a 6.5 digit DVM qualify one as a volt-nut? Or should first some voltage references be added?
> Erik.

There’s *always* a way to spend more money on bench gear :) :) :). In terms of this
project, what you have sounds like it’s doing the job just fine.

Bob

> 
> On 5-3-2022 20:50, Bob kb8tq wrote:
>> Hi
>> 
>> One suggestion on the frequency plot:
>> 
>> TimeLab as you have it set does a “per pixel” process on its charts. You have
>> a limited screen resolution and a ton of data. You have to do something. There
>> is no way to display it all.
>> 
>> It’s a good idea to zoom in to some of the “problem regions” and see
>> what is really going on. This reduces the number of data points that
>> map to one pixel and potentially changes what you see a bit. The closer you
>> get to one data point per pixel, the closer you are to “reality” without any
>> processing potentially getting in the way.
>> 
>> On the DVM, eBay (with some time spent shopping) will sell you a working
>> 6 1/2 digit device for < $100 if you are patient. If you need it now, it’s still
>> likely to be below $200. Yes it will be a big clunky box that takes up bench
>> space. It will be HPIB and not serial i/o. The display might be a bit fun in
>> bright lighting.
>> 
>> While you are shopping, I would get set up for phase noise as well. Often
>> it can spot issues that are tough to find with a second to second data stream.
>> The setup does not have to be very fancy or terribly expensive.
>> 
>> Bob
>> 
>