bg7tbl gpsdo master reference

[kj7e]

Power glitch maybe?  I have not seen mine ever loose lock while powered on.  I have pulled power momentarily now and then for various reasons.  It will come right back without a "No Fix" message but will loose lock for a few min.

Running a test now to see how stable the hold over is.  Going to run it with no antenna for a few days.  I have a good record on the stability my HP 53132A Opt 010 to compare it to.

[Jörg]

In all the situations where the L is lost, I could no abnormalities which see on the power point. If I pull out the antenna plug, the display will immediately "GPS NO FIX". "
I can test a different PSU the next time time and let the results surprise me.
Due to recent events, I'm now a test recording without GPS antenna in the "GPS NO FIX" mode. "
The first 15 minutes point to a more stable 10 MHz signal. We'll see how it looks for several hours.

[kj7e]

I really don't like switch mode power adapters, so I built a compact dedicated 12v 3a LDO linear supply for the GPSDO and 10MHz distribution amp.

[Jörg]

First an example of the characteristic of the BG7TBL LCD GPSDO in hold over mode for about 12 hours.
Below is a plot and statistics in disciplined mode once again as a comparison.
Gate Time 10 Sek. 
Reference for Frequency Counter: HP Z 3805A

[kj7e]

I found the same thing but ran mine for 48 hours in holdover.  The Trimble 65256 oven used is seriously impressive and stable.

[metrologist]

<2E-10 over 12 hours? How did the 48 hour plots compare?

[kj7e]

<2E-10 over 12 hours? How did the 48 hour plots compare?

Inconclusive, I could not tell if my HP 53132A's 10811-60160 or the Trimble 65256 oven that was drifting.  Started with about 800uHz difference then drifted up and down no more than 2mHz by 48 hours.  I would need a second GPSDO to tie my counter to and run a much longer test, like as in 30 days.

[Jörg]

<2E-10 over 12 hours? How did the 48 hour plots compare?

I removed the GPS antenna again and started a new recording of the HoldOver mode for 48h or so.
I hope that there will be no EMC influences in the time.

[kj7e]

My 48 hour test was inconclusive by itself.  Im running two 24 hour tests, first comparing the GPSDO with GPS lock to my counters internal reference, then I will repeat this with the GPSDO in free run.  I suspect I wont see much difference as the Trimble oven used in the GPSDO seems to be very good indeed.  I really need a Rubidum or a second GPSDO for a more stable ref on the counter for this, but thats how good the Trimble oven seems to be.

GPSDO locked, HP 53132 with HP Option 010 HS oven with the following settings;
Gate set to 11 digits (34.4sec gate time), 2512 measurements comes to almost exactly 24 hours.  The LSD is solid and stable when the Gate is set by Digit vs Time.

STD DEV;


MEAN;


MAX;


MIN;

[kj7e]

Same as above but with the GPSDO in Free Run or Hold Over with no GPS signal;

STD DEV;


MEAN;


MAX;


MIN;

[Jörg]

<2E-10 over 12 hours? How did the 48 hour plots compare?

Now another flood of images of the "Free run mode " of the BG7TBL LCD GPSDO without a GPS signal.
Measuring conditions. Tek FCA3000 with HP Z3805A as an external time-base.
Gate-time 10 sec. Ambient temperature min. 18 °c/max 22 °c.
From top to bottom after 12h, 24h, 36h, 48h and current 60h.
I'm going to keep the record running for about two days.

[Jörg]

And here's the rest.

Current modification: 22.4.18
I have added the recording at a duration of 120 h Hold Ovr Mode .
I think you can now estimate the drift well if you construct a regression line over the plot.
On the Y axis are 1mHz/div. On the x axis about 13, 5h/div in this view.

[lars]

Hello Jörg

Nice to see all your measurements. Especially the long term.

Seems that your BG7TBL GPSDO is better in hold than in run

In your reply #596 April  13 you lost "locked" and my guess would be that the OCXO made a jump of a few tenths of a ppb. I have seen jumps after days, after months and even more frequent on my OCXO's bought from eBay. I have one MV102 OCXO that is so bad that I don't get a useful ADEV graph between jumps. Of course I have some OCXO that I haven't seen a jump on (maybe have run them to little )

What is more interesting with your graph is the step wise frequency changes. For me it seems that the BG7TBL GPSDO you have has an FLL. Like the first BG7TBL GPSDO's. So it counts the frequency for about a minute and when corrects. If the error gets smaller it counts longer and make smaller steps. As has been pointed out earlier in the thread and by KE5FX tests, FLL is not the optimum for a GPSDO. This is also the reason you even after 40000seconds averaging has only about 10 digits accuracy.

Even a very simple GPSDO like mine (https://www.eevblog.com/forum/projects/lars-diy-gpsdo-with-arduino-and-1ns-resolution-tic/?all )
has a integral term that normally hold the output phase within 20-50ns (even with a DOT050V TCXO!). 40ns in 40000second should be 1E-12 so 12 digits.
Making a PI-loop is not difficult. A brief explanation is on page 8 in my instruction. On page 15 I have several good links to GPSDO control loops also. But it seems BG7TBL hasn't tried for several years.

Another thought is why nobody (except KE5FX) do time interval measurements between the BG7TBL and other professional GPSDO's with their excellent counters. I really like timelab by KE5FX. Also if the 1PPS comes from the GPS module and not is a divided down 10MHz it would be interesting to see if the phase (time) goes away. 0.1ppb would be 4us in 40000seconds. If I remember correct KE5FX showed this.

Lars

[metrologist]

I was monitoring two Trimble UMMC units on a scope and they are locked in phase. Every so many hours, the signals relative phase would start to drift to +/- one cycle and then be locked again. The cycle repeats... Maybe in HO they will be stable enough for me to identify which one it is...

[Jörg]

Hello Lars. Your guess is quite possible.

The user DajMasta has also shown interesting records of 2016-05-31 model in his post «Reply #448 on: September 15, 2017, 04:32:27 pm».
In comparison, the BG7TBL LCD model 2017 has resulted in a change.
I could only observe these jumps of the output frequency in situations where the letter  "L " disappeared from the display. Not in "GPS no fixed " mode or in "normal" modes.

I have also added a 144h HoldOverMode plot to my post «Reply #611 on: April 20, 2018, 03:42:36 AM».
I think you can now estimate the drift well if you construct a regression line over the plot.
On the Y axis are 1mHz/div. On the x axis about 13, 5h/div in this view.

[Jörg]

Hello kj7e.

I think your option 10 time base is not bad.
In a period of 12-24h you can use the option to exclude the effects of disciplining on the counter reference.
For example, at the moment I cannot tell exactly whether the HP Z3805A GPSDO or the BG7TBL GPSDO is responsible for medium-term fluctuations when both are in the GPS lock.
Can you visualize the data from your HP 53132A on the PC via the GPIB interface?

[kj7e]

Hello kj7e.

I think your option 10 time base is not bad.
In a period of 12-24h you can use the option to exclude the effects of disciplining on the counter reference.
For example, at the moment I cannot tell exactly whether the HP Z3805A GPSDO or the BG7TBL GPSDO is responsible for medium-term fluctuations when both are in the GPS lock.
Can you visualize the data from your HP 53132A on the PC via the GPIB interface?

Working on that, I have an NI GPIB to Ethernet adapter on the way.

[lars]

Hello Jörg,

I am still not impressed by reply 448 saying the algorithm is changed. Could still be a FLL that depending on OCXO drift gives your 10 digits or in DajMastas case 11 digits over the 12 hours.

KE5FX already long ago got information that the resolution of the DAC were in the 1E-11 range so the strange behaviour reported doesn't convince me it is any better algorithms. If someone can show me that the 10Mhz stays within 100ns for 12 hours compared to a good GPSDO as other professional GPSDO's does it would be good.

Even without regression I should say the drift of your OCXO is about -2E-11 per day that is very good.

Lars

[Jörg]

Hello Lars. I have measured the phase and time interval of the BG7TL GPSDO (input B) against the professional hp Z 3805A GPSDO (input A).
In this view, the requires phase is approximately 16 min 30 sec for a run of 0 degrees to 0 degrees. At TI measurement it is about 20 min 10 sec for a complete run from 0 to 0 ns. Input B was measured relative to input A. Both GPSDO are in GPS lock mode. For the use with the scope I have only one old analogue oscilloscopes available.
This is not so comfortable for automated measuring like a DSO

[lars]

Hello Jörg,

and thanks for doing the time interval measurements. I don't think you need a DSO for this measurement as you have such an excellent instrument for time interval.

As the frequency of both sources are very close to 10MHz the period time is very close to 100ns. So both your measurements wraps at 100ns drift. Normally two locked GPSDO's will drift back and forth about 20-50ns but here one drifts away all the time

16min 30s is 990secs and 100ns during 990secs shows an offset of about 100ns/990s that is about 1E-10. 100ns in 20min and 10sec is 1210 secs so 100ns/1210s is about 8E-11. So the frequency varies a little between the measurements. What concerns me is that one of the locked GPSDO continuously drift against the other. Sorry to say I guess it is the BG7TBL that still has the inferior FLL algoritm to steer the frequency.

Lars

[Jörg]

I could now mount a well GPS antenna in an exposed place.
Compared to the antenna outside the window, the stability of the navigation has improved considerably.
I could not determine a measurable change in the 10 MHz output from the BG7TBL LCD GPSDO.

[jpb]

Of the units listed in the OP, which would be the most desirable?
Are any much better than the others?

I'm looking for a GPSDO and really don't know much about them.
Would these BG7TBL units be considered the best bang for the buck or should I look elsewhere for a ready to go box.

I'm very happy with this unit.  I did the management mode and not had a single problem with my unit.   I also added the 10mhz square wave output.
The existing units have the Star Management mode already completed.

https://www.ebay.com/itm/10Mhz-GPSDO-OSCILLOQUARTZ-OSA-OEM-GPS-GPSDO-10MHz-1PPS-STAR-GPS-Clock/262861459973?hash=item3d33c39205:g:sbYAAOSwh2xYAjrp

Do you need to mod these before you get any output? I have just bought one and the vendor claims it works with Lady Heather whilst mine shows no signs of communicating at all over RS232. Pin 2 is permanently at logic 1 (-5.8V) and pin 3 is at around zero volts. Pin 8 has the 1PPS ok. The unit as a whole seems to work ok (locks on to 10MHz as checked against a radio standard).

EDIT : Problem now fixed thanks to help from texaspyro.

[ian.ameline]

How does one control whether the output is a 10Mhz Square or Sine wave (or is it hard-wired one or the other)
The Manual just says that it is either a sine or a square wave -- it gives no other information...

Thanks!

[Brandon B]

Would anyone know a good way to take the data coming into the GPSDO and export it into a data collection sheet like excel?

Thanks!

[kj7e]

@ian,
Default is a sine wave, I dont see a way to output a 10MHz square wave.   If there is, its likely a PCB jumper on the bottom of the board.

@Brandon,
LadyHeather has logging functions that may work for you;
http://www.ke5fx.com/heather/readme.htm