[time-nuts] TCVCXO Adjustment

Bob kb8tq kb8tq at n1k.org
Sat Apr 14 13:26:26 EDT 2018


Hi



> On Apr 14, 2018, at 11:10 AM, Adrian Godwin <artgodwin at gmail.com> wrote:
> 
> What if you iterated toward a suitable minimum-error setting, then looked
> for cyclic corrections with a period of weeks to months. Once you start to
> see that, choose the centre of the cycle and track it (or perhaps just
> increase the time constant).

If you only get one data point per month, then it will be a pretty basic plot. 
Even with a data point a week, it will be pretty sparse. 

We get very used to “data everywhere” sorts of situations. With most of the
proposed approaches *if* tight set is desired the data will be pretty sparse. 
It’s still a bit unclear what the target accuracy is (or even if the TCXO needs
setting at all …).

Bob


> 
> On Sat, Apr 14, 2018 at 2:37 PM, Bob kb8tq <kb8tq at n1k.org> wrote:
> 
>> Hi
>> 
>> The gotcha is that you do not have a calibrated adjustment. Put another
>> way,
>> there isn’t a perfect correlation between DAC bits and ppm. Each
>> adjustment
>> you make is subject to a bit of error. When you are trying to get within a
>> ppm,
>> your measurements are quicker, so the larger error ( percentage of step)
>> may
>> not be as big a deal. When you get close, it is likely to become a big
>> deal.
>> 
>> You could track all of your changes (month to month). The issue there is
>> that the
>> drift in the TCXO month to month is not likely to be the same. Sorting all
>> of that out
>> could be a bit nasty …..
>> 
>> ====
>> 
>> TCXO drift is not the only contributor to the accuracy of a time code
>> generator.
>> The other obvious one is setting it to the correct time in the first
>> place.  If the
>> objective is to compare data from different locations, getting it set may
>> be as
>> big an issue as the TCXO. If it’s a single location and the time is
>> arbitrary, then
>> maybe not so big a deal. If it’s all arbitrary … why worry about drift? ….
>> 
>> GPS on the board looks like a good thing to have to me ….
>> 
>> Bob
>> 
>> 
>>> On Apr 14, 2018, at 6:35 AM, Adrian Godwin <artgodwin at gmail.com> wrote:
>>> 
>>> If you compare VCXO time with UTC or GPS once a month to an accuracy of
>> 1s
>>> (with NMEA or even a time signal and manual pushbutton) and make a
>>> correction for the 2.5 million seconds that occurred since the last
>>> correction, you'll be better than 0.5 ppm.
>>> 
>>> Is that good enough ?
>>> 
>>> On Sat, Apr 14, 2018 at 5:59 AM, Hal Murray <hmurray at megapathdsl.net>
>> wrote:
>>> 
>>>> 
>>>> kb8tq at n1k.org said:
>>>>> The alternative is to plug a USB GPS into the mac and do a bit of code
>> to
>>>>> compare things. If you want to pass the gizmo around to your friends ….
>>>> that
>>>>> can be done. Pretty good ones are “sub $10” delivered.
>>>> 
>>>> USB GPS gizmos generally don't have a PPS and the timing on the NMEA
>>>> sentences is generally crappy.  (I may be biased by a few bad examples.)
>>>> 
>>>> Does anybody have a list of ones known to work well?
>>>> 
>>>> There is at least one GPS-USB with PPS, the Navisys
>>>> GR-601W/GR-701W/GR-801W,
>>>> They were hard to get retail.  Looks like idealez sells them Taiwan for
>>>> $47
>>>> for the 701W (Ublox 7), $50 for the 801W (Ublox 8) (plus shipping).  I
>>>> haven't tried ordering through them.  Gary and/or Mark may have some for
>>>> sale.
>>>> 
>>>> ----------
>>>> 
>>>> Plan B is a GPS breakout and a USB-serial breakout and a few wires.  4x
>>>> less
>>>> USB jitter if your USB-Serial chip is full-speed.
>>>> 
>>>> I got mine from SparkFun:
>>>> Venus GPS with SMA Connector
>>>> https://www.sparkfun.com/products/11058
>>>> USB to Serial Breakout - FT232RL
>>>> https://www.sparkfun.com/products/12731
>>>> 
>>>> ----------
>>>> 
>>>> Google found this.  I don't know anything more:
>>>> https://www.zti-communications.com/z050-gps-dongle/
>>>> 
>>>> 
>>>> --
>>>> These are my opinions.  I hate spam.
>>>> 
>>>> 
>>>> 
>>>> 
>>>> _______________________________________________
>>>> 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