Yo Tom!

On Mon, 18 Jul 2016 18:42:39 -0700
"Tom Van Baak" tvb@LeapSecond.com wrote:

Got some C or Python?

I don't do Windows...  And I want this to auto-generate graphs for web
pages on RasPi's.  Lot's of RasPi's.  :-)

Yeah, that looks sorta like what I have.

Just the facts there, i can do that.

That's what I think I am looking for.

A bit of background.  Ntpd generates a number they call the Allan
Deviation, but the description looks like the Allan Variation, they
describe is as the minima in the Allan Deviation plot, when it has a U
shape.

They then use that number to set the PLL optimally.

Something looks fishy to me, so I want to visualize it, and help the
rest of the NTPsec team see it as well.

RGDS
GARY

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
gem@rellim.com  Tel:+1 541 382 8588

See Anders Wallin's post: https://www.febo.com/pipermail/time-nuts/2014-March/083298.html

For C, I use adev4.c and adev5.c under www.leapsecond.com/tools/ when I don't need the interactive power of TimeLab or Stable32 or when I'm working with millions of data points. Then I use Excel or gnuplot to make the log-log plot. After much trial and error here's a gnuplot example:

http://leapsecond.com/pic/mains-adev-mdev-gnuplot-g4.png

Perhaps PHK can comment if Dave Mill's terminology wrt Allan deviation is strictly correct or not. There's Allan variance, and Allan deviation (which is just its square root). Then less-rigorous usages like Allan intercept or Allan floor, etc. Be careful with your interpretation since AFAIK NTP is working only with past+present data rather than a complete past+present+future data set.

/tvb

----- Original Message -----
From: "Gary E. Miller" gem@rellim.com
To: "Tom Van Baak" tvb@LeapSecond.com
Cc: "Discussion of precise time and frequency measurement" time-nuts@febo.com
Sent: Monday, July 18, 2016 7:01 PM
Subject: Re: [time-nuts] Allan Deviation recipe?

Yo Tom!

On Mon, 18 Jul 2016 18:42:39 -0700
"Tom Van Baak" tvb@LeapSecond.com wrote:

Got some C or Python?

I don't do Windows...  And I want this to auto-generate graphs for web
pages on RasPi's.  Lot's of RasPi's.  :-)

Yeah, that looks sorta like what I have.

Just the facts there, i can do that.

That's what I think I am looking for.

A bit of background.  Ntpd generates a number they call the Allan
Deviation, but the description looks like the Allan Variation, they
describe is as the minima in the Allan Deviation plot, when it has a U
shape.

They then use that number to set the PLL optimally.

Something looks fishy to me, so I want to visualize it, and help the
rest of the NTPsec team see it as well.

RGDS
GARY

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
gem@rellim.com  Tel:+1 541 382 8588

Hi Gary,

Looking good. Not sure how artistically picky you are, but:

• consider changing 'PPS' to '1PPS'
• consider changing 'GPS' to 'NMEA' (at least I think that's what you mean)
• the y-axis title is partially cropped off the chart; center the plot to get the same amount of border as top, bottom, and right sides
• for mobile or printed usage, the axis titles (powers of ten) font size is too small by maybe 2 or 3 points
• consider using a greater depth of adev5 points (e.g., 10 or 20 or 50 per decade) and skip the solid line; or just try both ways to see if one looks better

Also try plotting TDEV instead. In a case like this an ADEV plot is somewhat boring or even misleading -- it's just a -1 slope going down forever. It's not quite what ADEV was meant for. The slope suggests you have a bounded amount of white phase noise and that's that. So, for example, the entire PPS red trace can thought of as 6e-4/tau. In other words, you can represent the performance of the receiver with a single rms number, instead of a featureless ADEV plot. Similarly the blue trace is essentially 0.6/tau.

The weird jump in the NMEA blue trace between tau 1000 and 2000 should be investigated. Increasing the point density (using adev5) or eliminating the false line interpolation or using TDEV may help you look into this.

The other thing, that Bob just alluded to, is that while your NMEA measurements are probably legit, your PPS measurements may be totally skewed by the fact that you're using a plain PC and its operating system (and NTP, and drivers, interrupts, BIOS, caches, etc.) as a measuring instrument. The actual 1PPS out of a typical GPS receiver is orders of magnitude more precise than the "tool" you're using to measure it.

Therefore it's possible that the red trace is more a measurement of how bad your PC/NTP measurement setup is rather than a measurement of how good the 1PPS is. In other words, you have accidentally used 1PPS to measure the limits of your measurement system, rather than use your measurement system to measure the 1PPS. Does that make sense?

Thanks,
/tvb

----- Original Message -----
From: "Gary E. Miller" gem@rellim.com
To: "Tom Van Baak" tvb@LeapSecond.com
Cc: "Discussion of precise time and frequency measurement" time-nuts@febo.com
Sent: Tuesday, July 19, 2016 11:42 AM
Subject: Re: [time-nuts] Allan Deviation recipe?

Yo All!

With help from Tom Van Baak and Bob Stewart, I have 6 weeks worth of
GPS Serial and PPS timing data in a nice ADEV chart.  I think.

Check it out:
https://rellim.com/graphs/adev.png

That is for a SiRF III (MR-350P) with a bad sky view.

How does that look?  Totally wrong?  Somewhat wrong?

Once I get this packaged up I'll have lots more similar graphs.

RGDS
GARY

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
gem@rellim.com  Tel:+1 541 382 8588

Yo Tom!

On Tue, 19 Jul 2016 12:47:12 -0700
"Tom Van Baak" tvb@LeapSecond.com wrote:

I'm picky enough to annoy others.  :-)

For the moment, I'm happy if the data is technically correct.

In NTP land they are synonymous.  I appreciate the vendors now talk
of 5PPS, but my audience is NTP.

Well, in this case, it is actually SiRF Binary.  So maybe 'Serial',
or 'GPS Serial'.

I'm gonna make them the same sie as the other chrony-graphs.

Ditto, they'll be chrony-graph style.  Once they look that way I'll
see about getting the project style changed.

The lines are so straight, is there any data hiding in there?  Well,
surprise, there is!

That was next on my list.  Just done.  Those curves are more interesting,
just not sure what that means...

I gotta figure it out to explain it to the NTP folks.

Which pretty much what I expected.

That is what I love about plots, now it seems to obvious.  Execpt the new
bumps...

More points added.  Yeah, weird...

Yeah, sorta backwards, but that is the data I got.  How to plot it
to show interesting things is what I'm working on.

Yup, exactly, and sorta what I am trying to get to.  Given a good PPS,
how well is the ntpd working?  The point of PPS is to discipline the
NTP, and that is sorta what this shows.

It gets more interesting when using the PPS to be sure the system clock
is good, then using the system clock to measure other refclocks.  Like a
USB GPS compared to the PPS.  That will come in time.

The USB GPS results have surprised a lot of people, and solid proof
required to overcome disbelief.

NTP has some long standing questions on how to really model NTP time
served over real networks.  So using the system clock, disciplined by PPS,
will be a good place to measure that.

If I could get a real good PCI clock (that I could afford) then that
could be the reference for the PPS, ntpd, and/or system clock.

RGDS
GARY

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
gem@rellim.com  Tel:+1 541 382 8588