Hi Bob,
said:
"You have roughly 25 ns p-p in the data you show."
OK, here's a misunderstanding on my part right off the bat.  You see the swing as a p-p value, when I've been looking at it as only +/- 12.5ns from the trendline.

said some time ago:
"Now toss in the basics of GPS. Depending on the day, you will get <10 ns to  >100 ns swing over a  24 hour period. Today may or may not be the same as tomorrow."
So maybe I'm thinking too much about the >100ns figure, and not so much about the <10ns figure you mentioned.  The average doesn't seem to do much for me, either.  So, is the probability curve between 10ns and 100ns, where 100ns is least probable, of the type  y=2^-x?  IOW, in a year, I might see one 100ns swing, I would probably see at least one or two 50ns swings, and will probably see anything less than that multiple times, with the probably increasing as the value gets lower.

Not trying to crucify you with your own words, Bob.  Like many of the time-nuts who don't post, I'm just trying to make some sense of this in terms I can deal with.

Bob

 From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: Magnus Danielson magnus@rubidium.se
Sent: Thursday, April 27, 2017 1:18 PM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi

You have roughly 25 ns p-p in the data you show. There are a number of 10 ns “cycles” in the data.
Any of this may be due to ionosphere. They also could be due to other issues.  With ~4.4 days of noisy
data, it may be tough to spot a trend. Since the ionosphere is a bit random, there is no guarantee that
you will always see a pretty sinusoidal trend line through the data. It’s a good bet that things quiet down
around midnight. There is no guarantee that they always go nuts (or go nuts to the same degree) around noon.

Bob

On 4/27/17 12:10 PM, Bob Stewart wrote:

I don't think it necessarily has a nice distribution.  precise phase
measurement with GPS has lots of effects that are in the "about a meter"
range.

Ionosphere is one - if you're near a ionosonde you might be able to get
a data set to correlate against.  It's not necessarily a "worst at noon,
best at midnight" thing - some of the satellites contributing to your
fix will be on slant paths, so the solar ionization is not uniform.

Solid earth tides are another effect in that general magnitude (10s of
cm) - see Wikipedia or:
https://www.unavco.org/education/professional-development/short-courses/course-materials/strainmeter/2005-strainmeter-course-materials/tidenote.pdf

Thermal expansion and contraction of a variety of things:

1. the structure supporting your antenna
2. the earth's surface
3. the coax from antenna to receiver
4. any filters
5. the antenna

For most of these, they are periodic in some sense: solid earth tides
are affected by moon and sun, which have reasonably well known periods.
Thermal effects typically have a strong 24 hour cycle (although I've
seen some weird ones caused by shadows intermittently falling on stuff,
but still, it has a 24 hour periodicity)

Then there's various geometry/multipath things - choke rings help if
you're in a wide open area.

Hi

Dig into space weather if you want to get into the details of the why and how often. It’s all out there
Google is your friend. Things like sun spot cycles are one of many drivers. The more perturbed the
space weather is day to day, the more likely you are to see changes in the GPS.  Monitor the space
weather sites on a regular basis and you will be able to make some guesses about what’s likely to
happen. It’s only going to be a guess, no better than the 10 day weather forecast :) One of many
sites: http://www.swpc.noaa.gov http://www.swpc.noaa.gov/ . A bit about what to watch for: http://www.swpc.noaa.gov/impacts/space-weather-and-gps-systems http://www.swpc.noaa.gov/impacts/space-weather-and-gps-systems
This one is a bit HF oriented: http://www.sws.bom.gov.au/HF_Systems/6/5 http://www.sws.bom.gov.au/HF_Systems/6/5

That’s only the first part.

The second part is that the delta ionosphere is fit by GPS data. What you are concerned with is not just a
"bump' of space weather. You are concerned with one that does not fit well. That’s a normal
thing when the weather is rough, but not an always thing.  A peak solar flux event that ramps up
slow and decays slowly is very different than one with faster changes. You do get both. Patchy
disruptions in the ionosphere are worse than a uniform high. They are hard to fit.

Next up …

The goodness of fit depends a lot on the sat’s you are using for your processing and where they are
located in the sky. If you happen to have a sat that sends a signal across a big long patch of poorly fit
ionosphere,  you have a problem. If every single sat you are using is straight over head and your
house is well fit, there is no problem.  Longer paths by their nature are more likely to be an issue.
Bad fit only matters if you are depending to some extent on that part of the sky.  How much longer
vs shorter contributes to your solution right now is always a “that depends” sort of thing.

Is that all there is?

No, not hardly, that’s just the easy part. The troposphere also gets into the act and it flies around a
bit. Last time I checked, they just use a static model there so it’s not a broadcast vs reality issue.
You also get into things like location and sat angle from your location. If you are in northern Greenland
things will be a bit different than in Ceylon. There are a few other issues I could probably dive into with a bit of research.

So no, it’s not simple. How often do you see > 100 ns? Best data I’ve seen is that you hit that
range a few times a year on average. More so at solar maxima and less so at solar minima. It’s no
different than propagation on 10 meters. If you are looking for 100 ns every day, day in and day out,
that’s not going to happen.

You are looking for a peak to peak sort of swing. If you already have 20 ns wander in the data, you are
going to have a hard time seeing anything much below 20 ns. What you are looking for is most likely to have
a 86,000 second period (= day - night cycle). My guess is that you don’t see it because it’s buried in the noise of your data.

Not at all easy….

Bob

Hi Bob,
OK, since we are in a low sunspot cycle, then it would follow that the 100ns movements would be rare.  Also, since I'm at about 29.8 degrees north, with few obstructions or reflectors to cause a problem, that improves what I'm capable of seeing out of what's available.  As to what problems are buried in my data due to equipment limitations, you're welcome to give me a long term loan of a nice H Maser and Timepod.  I promise I'll treat them well!  =)

Anyway, thanks for the explications.  As mentioned, I see figures bandied about on timenuts, but no explanation of the variables that cause those figures to be true or not true.  When I don't see them happening in my "lab", then I get confused.  I'm pretty certain I'm not the only one this happens to.  Several of you have been adamant about ionospheric effects north of 60ns.  I just haven't seen them.

Bob 

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net
Cc: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Thursday, April 27, 2017 4:17 PM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
Dig into space weather if you want to get into the details of the why and how often. It’s all out thereGoogle is your friend. Things like sun spot cycles are one of many drivers. The more perturbed the space weather is day to day, the more likely you are to see changes in the GPS.  Monitor the spaceweather sites on a regular basis and you will be able to make some guesses about what’s likely to happen. It’s only going to be a guess, no better than the 10 day weather forecast :) One of manysites: http://www.swpc.noaa.gov . A bit about what to watch for: http://www.swpc.noaa.gov/impacts/space-weather-and-gps-systemsThis one is a bit HF oriented: http://www.sws.bom.gov.au/HF_Systems/6/5
That’s only the first part.
The second part is that the delta ionosphere is fit by GPS data. What you are concerned with is not just a "bump' of space weather. You are concerned with one that does not fit well. That’s a normalthing when the weather is rough, but not an always thing.  A peak solar flux event that ramps upslow and decays slowly is very different than one with faster changes. You do get both. Patchydisruptions in the ionosphere are worse than a uniform high. They are hard to fit.
Next up …
The goodness of fit depends a lot on the sat’s you are using for your processing and where they arelocated in the sky. If you happen to have a sat that sends a signal across a big long patch of poorly fitionosphere,  you have a problem. If every single sat you are using is straight over head and your house is well fit, there is no problem.  Longer paths by their nature are more likely to be an issue.Bad fit only matters if you are depending to some extent on that part of the sky.  How much longer vs shorter contributes to your solution right now is always a “that depends” sort of thing. 
Is that all there is?
No, not hardly, that’s just the easy part. The troposphere also gets into the act and it flies around a bit. Last time I checked, they just use a static model there so it’s not a broadcast vs reality issue.You also get into things like location and sat angle from your location. If you are in northern Greenlandthings will be a bit different than in Ceylon. There are a few other issues I could probably dive into with a bit of research. 
So no, it’s not simple. How often do you see > 100 ns? Best data I’ve seen is that you hit that range a few times a year on average. More so at solar maxima and less so at solar minima. It’s no different than propagation on 10 meters. If you are looking for 100 ns every day, day in and day out, that’s not going to happen. 
You are looking for a peak to peak sort of swing. If you already have 20 ns wander in the data, you are going to have a hard time seeing anything much below 20 ns. What you are looking for is most likely to have a 86,000 second period (= day - night cycle). My guess is that you don’t see it because it’s buried in the noise of your data.
Not at all easy….
Bob

On Apr 27, 2017, at 3:10 PM, Bob Stewart bob@evoria.net wrote:
Hi Bob,
said:
"You have roughly 25 ns p-p in the data you show."
OK, here's a misunderstanding on my part right off the bat.  You see the swing as a p-p value, when I've been looking at it as only +/- 12.5ns from the trendline.

said some time ago:
"Now toss in the basics of GPS. Depending on the day, you will get <10 ns to  >100 ns swing over a  24 hour period. Today may or may not be the same as tomorrow."
So maybe I'm thinking too much about the >100ns figure, and not so much about the <10ns figure you mentioned.  The average doesn't seem to do much for me, either.  So, is the probability curve between 10ns and 100ns, where 100ns is least probable, of the type  y=2^-x?  IOW, in a year, I might see one 100ns swing, I would probably see at least one or two 50ns swings, and will probably see anything less than that multiple times, with the probably increasing as the value gets lower.

Not trying to crucify you with your own words, Bob.  Like many of the time-nuts who don't post, I'm just trying to make some sense of this in terms I can deal with.

Bob

 From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: Magnus Danielson magnus@rubidium.se
Sent: Thursday, April 27, 2017 1:18 PM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi

You have roughly 25 ns p-p in the data you show. There are a number of 10 ns “cycles” in the data.
Any of this may be due to ionosphere. They also could be due to other issues.  With ~4.4 days of noisy
data, it may be tough to spot a trend. Since the ionosphere is a bit random, there is no guarantee that
you will always see a pretty sinusoidal trend line through the data. It’s a good bet that things quiet down
around midnight. There is no guarantee that they always go nuts (or go nuts to the same degree) around noon.

Bob

Hi Jim,
said:"Ionosphere is one - if you're near a ionosonde you might be able to get a data set to correlate against.  It's not necessarily a "worst at noon, best at midnight" thing - some of the satellites contributing to your fix will be on slant paths, so the solar ionization is not uniform." Are the Ublox timing receivers a lot better at getting rid of ionospheric phase shifts because they can see more sats than the older Motorola receivers?
And as a point of reference, the dataset I linked begins at ~9:21AM CST on 4/22/2017.  I'm in west Houston, TX.

Bob

  From: jimlux <jimlux@earthlink.net>

To: time-nuts@febo.com
Sent: Thursday, April 27, 2017 5:03 PM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

On 4/27/17 12:10 PM, Bob Stewart wrote:

I don't think it necessarily has a nice distribution.  precise phase
measurement with GPS has lots of effects that are in the "about a meter"
range.

Ionosphere is one - if you're near a ionosonde you might be able to get
a data set to correlate against.  It's not necessarily a "worst at noon,
best at midnight" thing - some of the satellites contributing to your
fix will be on slant paths, so the solar ionization is not uniform.

Solid earth tides are another effect in that general magnitude (10s of
cm) - see Wikipedia or:
https://www.unavco.org/education/professional-development/short-courses/course-materials/strainmeter/2005-strainmeter-course-materials/tidenote.pdf

Thermal expansion and contraction of a variety of things:

1. the structure supporting your antenna
2. the earth's surface
3. the coax from antenna to receiver
4. any filters
5. the antenna

For most of these, they are periodic in some sense: solid earth tides
are affected by moon and sun, which have reasonably well known periods.
Thermal effects typically have a strong 24 hour cycle (although I've
seen some weird ones caused by shadows intermittently falling on stuff,
but still, it has a 24 hour periodicity)

Then there's various geometry/multipath things - choke rings help if
you're in a wide open area.

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Hi

Not really. Most of the time you have less than a full set of sat’s visible on a uBlox no matter how it is set up. If you crank the elevation mask up, you may
not see much difference between the two at all.  Also consider that the “old” Motorola receivers are (at this point) very old. The “new” 12 channel units
have been around for quite a long time. We have even been through a couple of generations of replacements for them.

Bob

On 4/27/17 3:10 PM, Bob Stewart wrote:

As a general thing, more satellites means more values into the average.
But it's also where they are.

If you've got a big clump of ionization to the west of you, then it
affects the satellites that are west of you, but not those that are
east. But if your current solution is more west than east, that will
pull it.

OTOH, most of the receivers also weight the contribution by the SNR. A
higher Total Electron Content (more delay) also results in more
attenuation.

1 TECU is about 16 cm in delay (call it 0.5 ns) (I think..)

typical TECU values are 50-100.. so you can see there's a pretty big effect.

ftp://ftp.ngdc.noaa.gov/STP/GPS_GNSS/Mihail's.pdf