bg7tbl gpsdo master reference

[jpb]

HI jpb. I also did the review with 10 sec avg time. The Allan Dev is as poor as at 1 sec avg.
Right now I'm taking a test with 100 sec. This looks better but about 2-3 X worse
as a BG7TBL 2017-12-16 GPSDO at 1 sec measuring time.
The manual of the OSA Star4 shows, that the PLL time constant can be varied.
Maybe that can make a better behavior of the 10 MHz output.
In Lady Heather's, I couldn't see the command.
Greeting. Jörg

I am very interested in the manual of the OSA Star4.
Any suggestions where?
My GPSDO is exactly the same and quite satisfactory.
However, lack of a manual restricts my options and understanding.
Sofar, looking for a manual was fruitless.

Thanks in advance,
Frans

Yes, I too would like to get hold of the manual. I did a quick search yesterday but couldn't find it on line.

[texaspyro]

Try this...

[jpb]

Try this...

Thank you.

[Jörg]

Hello, thanks to texaspyro for pointing out with "&t". That was very helpful.
The PLL time constant was 200 seconds. I adjusted the TC to 5000 seconds. Now the short-term stability is more determined by the OCXO and at the level of the BG7TBL 2017-12-16 GPSDO. The BG7TBL GPSDO is about 700-1000 μHz lower in frequency compared to my HP Z 3805A.
The OSA Star4 shows very nice close to the HP GPSDO.

Here is an example with 10 sec avg time over 12 hours.
HP Z 3805A as an external reference and OSA Star4+ on input B of the TEK FCA3000.

[Jörg]

A good message for all people who can't mount a Timig GPS antenna at an optimal place.
I took comparison measurements of an antenna on the roof and one on the outside of the window.
I could not notice a significant difference in 10 MHz accuracy, short-term stability or drift.
Both measurements are with the HP Z3805A as ext. reference, and the Oscilloquartz Star4+ GPSDO on Input B to the counter.
Gate Time is 1 second in a 12-hour period. In both measurements, the Star4 unit had completely finished the setup on the antenna.

[metrologist]

I think there is a need to have phase correlation as well...

[Jörg]

Of course. I only checked the 10 MHz output. Many people just need an accurate, stable reference for the frequency, without much technical effort.
For the most accurate 1PPS output in a synchronized environment, a more optimal setup can be the more accurate solution.

[kj7e]

About 2 weeks ago my 2017-12 LCD BG7TBL GPSDO would no longer lock, the supplied Trimble 65256 OCXO aged out and is now beyond the VFC range, slowly drifting down.  I have another 65256 OCXO on order.

[cdev]

Thats too soon! That really sucks.

About 2 weeks ago my 2017-12 LCD BG7TBL GPSDO would no longer lock, the supplied Trimble 65256 OCXO aged out and is now beyond the VFC range, slowly drifting down.  I have another 65256 OCXO on order.

This reminds me that I should likely buy at least one, maybe two of the OCXO that my GPSDO came with.

(But I suppose that if they have been removed from service they might be on their last legs too).

[kj7e]

^ I think I just got a bit unlucky, I suspect most samples will run for years and stay in tolerance.  I'm going to play with a few OCXO's, also have an OSCILLOQUARTZ 8663-XS on the way.

[zveck]

Has anyone tried and been successful adding a LCD to the non LCD version? Mine has the socket installed for the ribbon cable.

[cdev]

Jorg,

You would likely do better if you put a ground plane (even something as small as a CD/DVD) underneath your puck antenna.

They are designed to use one. Any flat piece of steel works well because of the magnet.

I'm surprised people do as well as they have because most ceramic patch antennas are just the top part of the antenna, the ground plane thats supposed to go underneath (say the roof of a car) is the bottom part. The puck antennas that are sold have a small ground plane but its intended to capacitively couple to a larger piece of metal.

[kj7e]

Just a quick update, replaced the original Trimble 65256 OCXO with a China sourced used 65256, the original was at the VFC DAC upper limit of 4.9v.  The replacement is sitting at ~2.3v so seems there is plenty of life in the replacement part.

One interesting thing I noticed, even though the 65256 is spec'd for 12v Vcc its being fed a regulated 5v.  There is a 0 ohm resistor that can be moved to feed it 12v from the input power, but I noticed no difference in performance of the OCXO running from either 5 or 12v other than initial oven warm up power.  Perhaps the thinking here is a regulated 5v to the oven will provide slightly more stable oven regulation, so I left it as such.

[Jörg]

Hello cdev.
It is not always what it seems. The material under the puck antenna is powder coated aluminum.
So i had to fix the antenna for the experiment. Because it was just an experiment.
I could allow myself a temporary solution.
Of course, with the roof antenna, the quality and quantity of the GPS reception is much better. Since I am not dependent on the patch antenna.

[edpalmer42]

Just a quick update, replaced the original Trimble 65256 OCXO with a China sourced used 65256, the original was at the VFC DAC upper limit of 4.9v.  The replacement is sitting at ~2.3v so seems there is plenty of life in the replacement part.

One interesting thing I noticed, even though the 65256 is spec'd for 12v Vcc its being fed a regulated 5v.  There is a 0 ohm resistor that can be moved to feed it 12v from the input power, but I noticed no difference in performance of the OCXO running from either 5 or 12v other than initial oven warm up power.  Perhaps the thinking here is a regulated 5v to the oven will provide slightly more stable oven regulation, so I left it as such.

There seems to be some confusion on the voltage rating for the 65256.  I tested mine and found that the output reaches maximum amplitude with a supply voltage of ~4V.  If it was rated for 12V, that wouldn't happen.  You wouldn't reach maximum amplitude until maybe 8V-10V.  I've also heard of people frying the 65256 by applying 12V.  I suspect that the 0 ohm jumper is to allow for either a 5V or 12V oscillator on the same board.

Ed

[kj7e]

There seems to be some confusion on the voltage rating for the 65256. ...

Indeed, most everywhere you search on the internet Vcc is spec'd at 12v, but I suspect 5v is more likely.  After about a full day of operation, the new 65256 VFC has been extremely steady (<1mv deviation seen for 48 hours after the first 24 hour period) and right about center of the DAC range.  So it appears I got a bit lucky with the used replacement part.

[texaspyro]

There seems to be some confusion on the voltage rating for the 65256. ...

The bare board runs on 6V (the Samsung version wants 5.5V).    There are some packaged up devices that use a 12V (or possibly other) voltage input.

[jpb]

Hi Texaspyro,

you mentioned a while ago that you'd add the Samsung variant to Lady Heather - have you had a chance to do this yet?

[texaspyro]

Hi Texaspyro,

you mentioned a while ago that you'd add the Samsung variant to Lady Heather - have you had a chance to do this yet?

Yes, it works well.   I hope to get the next release out in a (yeah, right) few days.

[jpb]

I like the Lady Heather signal strength (or rather Signal to Noise) plot.

I've not really thought about it before, but I'm quite surprised by the magnitude (and roundness) of the black hole presumably over the north pole where GPS satellites don't go.
(I assume that this is the cause of the large black circle in the plot - there are houses/trees around blocking things but that is true in all directions.)

Unfortunately I have to mount my antenna on the north side of the house (where my study/lab is) so the black hole removes much of the sky. (My antenna is mounted on the top of my garden fence a little way from the house and also from neighbouring houes.)

[WA2SQQ]

I just found this forum, and Im in the process of deciding whether or not to purchase the BG7TBL GPSDO. My intent is to use it in conjunction with my Flex 6500 SDR transceiver. I am realizing that there are several versions of this basic design. I'd like to have both 10 mhz both Sine ans Square wave output in addition to the 1PPS. I've seen a few units with a display of the NEMA info. Wondering if anyone can give me the following info:
1. What is the latest version and who is selling it?
2. Does anyone have any info on a Raspberry PI or Arduino display panel that would display the NEMA data?
3. Can NEMATime Panel be used stand alone, w/o taking control of your PC's time clock?
4. Seen some comments that one versions GPS recdeiver is much more sensitive than the other - truth?

Thanks

Bob, WA2SQQ

[Fennec]

Hi

since weeks I play arround with the FE-5680A but how you wanna change the frequency or the vfc ?
my question is how I can change the accuracy of the FE-5680A. All I found is change the temperature of the lamp.
How you wanna do that ?
pls make pictures if possible

Iam very interested in it. I have the same config, and a few FE-5680A bit just a HP 53131A with 10 digits.

[4thDoctorWhoFan]

I took one of those GPSDO's with the LCD and combined it with a Extron distribution amplifiers.  I did not like having two separate pieces taking up space.
All I had to do was change all the 75 ohm resistors in the amp to 50 ohm & add a pwr supply.

P1010142 (2)

P1010140 (2)

P1010139 (2)

P1010141 (2)

[kj7e]

Made a little how to video on changing out the normally supplied Trimble 65256 OCXO for an Oscilloquartz 8663-XS on my Dec 2017 LCD version.  Just have to move the Vcc jumper from 5v to the 12 volt position and increase the EFC gain to 2x.  The DAC for the EFC is 0-3.3v feeding one side of an LM358 with a default gain of 1.5 for 0-4.95v, 2x gives 0-6.6v and there is a 3x jumper for 0-9.9v as well.  There is a guy selling NOS UCT 108663-01 OCXO's on eBay right now for cheap, which appear to be rebranded Oscilloquartz 8663-XS's.
https://www.ebay.com/itm/UCT-OSA-MY-108663-01-10MHz-12V-Double-Oven-OCXO-Crystal-Oscillator/264230742882

More discussion on the UCT 108663 ocxo's can be found here: https://www.eevblog.com/forum/metrology/ocxo-calibration/msg1303548/#msg1303548

And here: https://www.eevblog.com/forum/testgear/purchasing-newer-vs-older-oxcos-for-gpsdo-use-pros-and-cons/

Oscilloquartz 8663 datasheet: http://www.sungwhatech.com/product/pdf/aging/8663.pdf

https://youtu.be/gBkRFT9TgQI

Testing some of the UCT 10866 ovens;


UCT 10866 oven installed;


Bottom of the board with the UCT 10866 installed, set for 12v and 2x gain for the EFC;


As a side note, I also installed one of the UCT ovens in my BG7TBL 10MHz distribution amp, it also had provisions on the board to utilize the ovens Vref for the EFC trimmer.  I've been very impressed with how little the UCT part drifts.

[KD0CAC John]

Did you forget links ?