Hi

The quick and simple answer to “how to check it?” is to compare to something better. Yes, this is a really deep rabbit hole to wander down. There are various things you can use and various approaches. As with anything in life, there is a tradeoff between how much you spend and how much time you put in.

How crazy do you want to get?

Bob

Bob,
I want to stay married, so only the equipment I have already described.  I guess I could use each of the three GPSDO in turn as the reference for the PM6690.  Take the readings and see where it goes.

I've always had big question looming over my head.  Adev is the summation of a pair of frequency measurements made over a period of time.  So, the question is do I need a reference for the counter that is superior than the frequency references I am trying to measure?

Regards,
Bob

-----Original Message-----
From: Bob Camp via time-nuts time-nuts@lists.febo.com
Sent: Monday, June 23, 2025 4:24 PM
To: Discussion of precise time and frequency measurement time-nuts@lists.febo.com
Cc: Bob Camp kb8tq@n1k.org
Subject: [time-nuts] Re: Synergy GPS replacement module for the Z3801/05
Importance: High

Hi

The quick and simple answer to “how to check it?” is to compare to something better. Yes, this is a really deep rabbit hole to wander down. There are various things you can use and various approaches. As with anything in life, there is a tradeoff between how much you spend and how much time you put in.

How crazy do you want to get?

Bob

time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

Would a 3-cornered hat ADEV using the 3 GPSDO show any relevant
information of each individual GPSDO?
Erik.

On 24-6-2025 1:36, rbenward--- via time-nuts wrote:

Hi

You are always in an A to B comparison (or to Eric’s point something even more complex).

Unless you know one of your devices is better than the rest, any comparison will only get you so far. A three way comparison needs to be done “real time” so you would have to add counter(s) to the mix.

Depending on what you are trying to look at, something fancier than a counter may be needed. Finding a counter that will directly measure <1x10^-12 (so 1 picosecond resolution) at 1 second is not at all easy. Finding some that do strange math to make you believe they might be doing that is much easier :). That’s not what you are after for ADEV.

So, what to do?

First up would be to monitor the data from the GPS modules. Log it for a few days. See how your “sats in view” is doing. If it’s not doing well, move your antenna to a better location.

Second might be to accept the limitations above and simply log the delta between the 1 pps out of this device vs that device. Collect data for a few days and see what it looks like. Plots can tell you a lot more than ADEV can.

Third might be to save up for something like a telecom Rb. Their eBay prices swing wildly. If and when they go below $100 delivered …. maybe you sneak one in the back door :).

Fun !!

Bob

So, for me this has always been a pressing question.  Forget the three cornered hat.  Can you do an ADEV on a GPSDO and derive any relevant information from it, without having a hydrogen maser as your reference?

Just as the rise time of a scope is a function of the scope native risetime and the risetime of the probe combined.  Maybe this is the best we can get outside a lab.

When I get the antenna splitter, I will take some data using each as a reference.  I also have a E1938 OXCO which I might bring into the fold.  It will be interesting to see the difference between the single oven 10811 vs the UCT 8663 double oven.

We shall see.
Bob

-----Original Message-----
From: Erik Kaashoek via time-nuts time-nuts@lists.febo.com
Sent: Tuesday, June 24, 2025 12:57 AM
To: time-nuts@lists.febo.com
Cc: Erik Kaashoek erik@kaashoek.com
Subject: [time-nuts] Re: Synergy GPS replacement module for the Z3801/05
Importance: High

Would a 3-cornered hat ADEV using the 3 GPSDO show any relevant information of each individual GPSDO?
Erik.

On 24-6-2025 1:36, rbenward--- via time-nuts wrote:

time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

Hi

If you want to have some fun, grab one of those UTC 8663’s and tear them open. Good luck finding any trace of a true double oven in there. Yes, the are always advertised that way. One clue might be their 1 to 2 ppb temperature spec. A true double oven would be at least 10X better than that.

To answer the question, yes, there are a whole lot of ways you can get relevant info on a GPSDO without having a hydrogen maser. However they do involve both time and at least some money. This immediately gets back to “budget”. If the budget is zero then the project pretty much stops there.

Things that might be used for evaluating a local standard’s accuracy:

1. A (possibly cheap) Rb standard. Likely combined with this or that.

2. A multi-band GNSS device like a Trimble NetRS or a Mosaic-T. Also combined with a few other things.

Yes there are other possibilities. If you are just after 1 second ADEV and nothing else, you head down a different rabbit hole. The first thing you bump into is the “noise floor” that you are trying to get to. Simply measuring the noise floor of your counter isn’t really very informative GPSDO wise :).

Bob

Hi Bob,
I have a Rubidium oscillator somewhere.  It's been a long time, I am not sure it's working anymore.  I also have one of those old Bulova OXCO.  I have two variables here, the first is the Synergy GPS, and the second, is the bare 10811, vs the 8663 double OXCO.  I don't plan on re-assembling the 10811 into the double oven configuration.  I started making a flex to replace all of those wires internally, but that's on hold right now.

For my purposes, using any of the GPSDOs I have will more than satisfy my needs.  It's that preverbal rabbit hole that beckons....So, they are all on the bench warming up....

Bob

-----Original Message-----
From: Bob Camp via time-nuts time-nuts@lists.febo.com
Sent: Tuesday, June 24, 2025 9:15 AM
To: Discussion of precise time and frequency measurement time-nuts@lists.febo.com
Cc: Bob Camp kb8tq@n1k.org
Subject: [time-nuts] Re: Synergy GPS replacement module for the Z3801/05
Importance: High

Hi

You are always in an A to B comparison (or to Eric’s point something even more complex).

Unless you know one of your devices is better than the rest, any comparison will only get you so far. A three way comparison needs to be done “real time” so you would have to add counter(s) to the mix.

Depending on what you are trying to look at, something fancier than a counter may be needed. Finding a counter that will directly measure <1x10^-12 (so 1 picosecond resolution) at 1 second is not at all easy. Finding some that do strange math to make you believe they might be doing that is much easier :). That’s not what you are after for ADEV.

So, what to do?

First up would be to monitor the data from the GPS modules. Log it for a few days. See how your “sats in view” is doing. If it’s not doing well, move your antenna to a better location.

Second might be to accept the limitations above and simply log the delta between the 1 pps out of this device vs that device. Collect data for a few days and see what it looks like. Plots can tell you a lot more than ADEV can.

Third might be to save up for something like a telecom Rb. Their eBay prices swing wildly. If and when they go below $100 delivered …. maybe you sneak one in the back door :).

Fun !!

Bob

time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

Hi

Just like the 8663, the 10811 never really was a real double oven. There is no value in the outer shell unless you need to warm it up from -40C. In many ways a “normal” 10811 actually is a better choice.

OCXO’s once they have warmed up for a couple months are a good reference for the “under 100 seconds” stuff. They might in some cases be ok for somewhat longer time spans.

Rb’s are typically a good thing for the 100s to 10,000 s sort of range.

A GPSDO is something that should do well in the > 1,000s sort of range.

In terms of noise floor, your counter likely will not show much at under the 100s sort of range.

Taking all this into numbers:

A good 10811 can have an adev around 1x10^-12 at 1 second. A really good one might be 4 or 5 times better. Count on sorting a few hundred to find that one :).

A “fine old OCXO” from somebody might have an adev around 1x10^-11 at 1 second. It could be better. It might be worse.

If an OCXO stays below it’s 1 second number out to 100 seconds that’s doing ok. Holding it to 1,000 seconds is rare.

A typical telecom Rb likely is in the 2x10^-11 range at 1 second. It should get to 2x10^-12 by 100 seconds.

A GPSDO at 1 second might be in the > 1x10^-11 range. It might stay there for a while. It could be better. Somewhere out around the 100’s of seconds range it should start getting better. By a thousand seconds 1x10^-12 is not out of the question. Past that various odd things come into the picture.

Your counter has a “one shot resolution” number. If it’s 100 ps, that comes out to 1x10^-10 at 1 second. You get to 1x10^-12 at 100 seconds. Yes, this assumes the reference to the counter does not get in the way. Also this is a resolution number and not an ADEV. Still, you want the resolution to be better than the ADEV you are after.

If you are looking for a resolution floor of 1x10^-13, your counter begins to be useful at 1,000 seconds. Put your pps from the GPSDO in channel A and grab a reading on channel B once every thousand seconds.

Since you would like to have at lest 100 samples for an ADEV number, you will be taking data for a day to get a reasonable answer.

If you are looking at OCXO’s you are measuring aging at that point.

Are there a whole bunch of assumptions there? Yes there are.

Is there a whole lot more to it all? Most certainly.

Fun !!

Bob