GPSDO vs Rubidium with a 53230A counter

[TheSteve]

I recently picked up an Agilent 53230A frequency counter. I thought I would use my GPSDO as the reference in and compare it to my Rubidium standard.
This is a plot of 10000+ readings, 12 digits, 1 second gate time, the min/max are set by the first 500 readings.

I know HighVoltage has the same counter,a GPSDO and Rubidium standard, hopefully he can post a shot of his to see how it compares.

[0.01C]

I will try

[Dr. Frank]

GPSDO have mediocre short - and mid-term Allan stability, compared to an Rb standard.
Better use the TimeLab package and the 53230A as a T.I. counter for the 1pps signals, to adjust the Rb standard to 1E-11 level, which will at least take several hours to mitigate even these mid-term variations of the GPSDO.

Frank

[HighVoltage]

My old HP GPSDO seems to be very stable but I do not have a working Rubidium standard at the moment.
I will set up this test with my stable external OCXO, it will be a good basis for a future RB test.

[TheSteve]

My GPSDO is a Trimble 63090 which came from a WCDMA site in Japan. My Rubidium standard is a repaired LPFRS-01. So neither are anything super exciting. The GPSDO does have a 1 PPS output but the Rubidium is 10 MHz only. I thought it might be interesting to see how others compared.
I'll certainly have to fire up TimeLab with the 53230A

I left it running overnight, 59000 + samples with the GPSDO as external reference and the rubidium as the input.
Here are the results:

[HighVoltage]

I am warming up my equipment.
Last time I used these three counters was many month ago.
Isn't it just amazing how close they are to each other, when measuring the HP GPSDO?

The HP Z3805A needs about 2 hours to be in GPS Lock mode.

[HighVoltage]

TheSteve
How did you get such a nice histogram of your measurement.
This was just measured on one channel, right?

Here is my attempt to get a good relationship between my GPSDO and external Piezo 10 MHz OCXO
Plus, how did you get a screen dump on your 53230A, it seems my instrument does not want to do that.

So, here is a picture of the screen
After 3587 measurements, only 582 pHz Pk to Pk
I find this kind of remarkable, or I am wrong?

The 53230A runs on its own internal OCXO reference.
 

[VK5RC]

When I get home I can set up a similar trial, 53230, z3805 (it stays on continuously) and a Efratom lpro Rubidium.
Aren't we really comparing the different disciplining protocols, both use a quartz crystal one disciplined by the Ru RF transition, the other indirectly Caesium?

[TheSteve]

HighVoltage - Saving a screen shot is anything but intuitive with these things - The sequence is Utility, System Setup, Manage Files, then for the action select "capture" - you can pick internal or external memory such as a USB drive.

VK5RC - I am really just looking at how my GPSDO and Rubidium references differ. Obviously any error I see can be from either device as my GPSDO is external reference for the counter and I am logging the data from the Rubidium standard.

My first post is a Histogram with a readings set to 500, a 1 second gate time, 12 digits. My second is the same config but I have math enabled which can display Allan Deviation.
I just bought another GPSDO today (Symmetricom 58503B), so once it arrives and has warmed up for a few days I'll compare it to my Trimble to see how much they wander around.

[VK5RC]

When I got home my wife had unplugged my z3805 to charge some garden tool battery.
Yesterday I was getting result of 1000uHz StanDev and AllanDev of 150uHz or so in the first few hours.
Quite a lot better tonight.
After 24hrs warm up , I get the results below over 20mins or so.
Edit; input is 50 ohms (or so) via a 3dB attenuator.