My external GPSDO is ****

Mark_WS7M

I all,

Being cheap I did not by the Flex GPSDO but purchased a unit recommended in these forums off of eBay.

My 6600 has locked on the GPSDO signal and consistently told me everything is great.  I never thought twice about it.

As we prepare for the WWV 100 year celebration one thing we would like to do for fun is a frequency measurement test using the WWV 10mhz signal as our lock source.

So tonight our frequency measurement expert measured my 6600 signal and low and behold found my signal to be phasing around 10hz.per minute which while it's not bad is worse than expected with a GPSDO lock.

For grins I switched the flex to lock on it's internal crystal and did a frequency call against WWV.  Being that I am just 8 miles from WWV I have a great signal.

The interesting thing is once I did that and had my frequency measurement guy remeasure my signal the phasing had calmed down to like 1 hertz per minute which is very good.

So my eBay GPSDO is not as good as the internal TCXO and and WWV calibration.

Frankly, I was a little surprised!

I mean 10 hz is not bad comparing where I was in the 70's trying to keep my Kenwood on frequency, but I did expect the GPSDO to be better.

Interesting!

Sergey R5AU

Mark, so interesting and could be different behavior , what type of GPSDO you utilize?

DH2ID

Your GPS signal might have been incorrect, due to obstruction in a building, through trees or electric cables. It even can be incorrect when the satellite signal itself is reflected or not correct. Your GPS antenna should be absolutely free of obstructions.

Martin AA6E

Sounds like your GPSDO is not locking up to satellites at all.  Do you have a "lock" indicator?  A good unit (like the Flex's) will tell you how many satellites are visible and how many are being tracked.  I get numbers like 12 visible and 5 or 6 tracked. (That's from memory; my rig is packed away at the moment.)

Sergey R5AU

Alex you are right, it a point of incorrect data by GPSDO. Antenna must be in free space between antenna and satellites

Mark_WS7M

Well I did some more tests.

On Saturday I took my 6600 along with my GPSDO to the WWV site.  I had my GPSDO antenna in perfectly clear view of the sky.  No trees, nothing.

Still the signal was wavering and off.   I switched to another eBay special GPSDO that had never been used.  We gave it 15 minutes to warm up.  It also was pretty bad.

So the guy that was checking signals brought over his Timble GPSDO.  We plugged that in gave it 5 minutes to lock and warm and drift was down to 1 hz in 80 seconds.

The other eBay specials were more like 10 hz in 10 secs.

So it pays to have a really good GPSDO if you are into the frequency measurment stuff!

Clay N9IO

So it really was **** as you said after all?

I'll be darned.

Mark I'm curious which unit you had ordered.
Would you post a photo of the unit in question for us?
I wan to make sure what I ordered is not that unit.
Thanks!

Martin AA6E

We need to do the numbers here.  The Flex GPSDO is supposed to be accurate to maybe a part in 10**11, when properly synced up.  If you're working at 30 MHz, that's an error of 0.0003 Hz.  If you can see ANY offset at all (with a normal radio on the HF bands), your GPSDO is not delivering.  The crystal osc. is probably free running, without being "disciplined" by GPS.

In my experience, the Flex unit takes about 10 minutes to lock up from power on.  This is not just "warming up"; it's finding the geometric model (including, time, freq., latitude, longitude, etc.) that correctly fits the incoming noisy GPS data streams.

The oscillator is so good that it's tricky to measure it, even if you have an atomic standard in the lab.

John - AI4FR

Below are my stats with a 6700 and a Flex GPSDO. The results are from a local AM broadcast station on 1250 in an attempt to avoid doppler shift from WWV. The results were obtained from the FreqCal mode in version 2.1.0 of WSJT-X. The real number to look at is the DF which is the 6th column. This will tell me how stable my radio and the AM transmitter is to each other. As can be seen, with the Flex GPSDO I am well under a Hz in stability(0.0XX of a Hz). To really do it right one would average the readings below while tossing out any flyers. While the radio has been on for several hours, I took the below measurements moments after tuning to 1250. Some improvement may occur if I let things settle down for a while.

Good luck and I hope this helps some.

UTC - Freq. - CAL - Offset - FMeas - DF - Level - S/N

19:08:56 1250 1 1500 1499.903 -0.097 -6.1 38.9

19:08:58 1250 1 1500 1499.913 -0.087 -6.8 39.9

19:09:04 1250 1 1500 1499.906 -0.094 -5.9 38.4

19:09:06 1250 1 1500 1499.909 -0.091 -4.1 36.6

19:09:09 1250 1 1500 1499.907 -0.093 -5.9 38.5

19:09:11 1250 1 1500 1499.905 -0.095 -6.4 38.9

19:09:13 1250 1 1500 1499.907 -0.093 -5.7 38.0

19:09:19 1250 1 1500 1499.917 -0.083 -5.1 37.9

19:09:21 1250 1 1500 1499.915 -0.085 -5.8 38.6

19:09:24 1250 1 1500 1499.918 -0.082 -7.7 40.6

19:09:26 1250 1 1500 1499.908 -0.092 -6.4 39.2

19:09:28 1250 1 1500 1499.896 -0.104 -7.3 39.7

19:09:34 1250 1 1500 1499.910 -0.090 -5.9 38.5

19:09:36 1250 1 1500 1499.913 -0.087 -7.1 39.6

19:09:39 1250 1 1500 1499.914 -0.086 -4.9 37.6

19:09:41 1250 1 1500 1499.910 -0.090 -5.9 38.5

19:09:43 1250 1 1500 1499.906 -0.094 -5.7 38.5

Bill W2PKY

I bought this one: https://www.ebay.com/itm/123049880474
Was able to put in a winning entry in the last ARRL FMT. 
AI4FR and I compared results, his the Flex internal unit, against the above unit and saw pretty much the same prints. My antenna is in the attic and seems to lock just fine.

John - AI4FR

Good point Bill. I honestly could not tell any difference in the results and would have to say that your ebay unit was as good and stable as the Flex GPSDO.

Mark_WS7M

You guys must have gotten better ones than I got.   

Not that I think mine are really that bad but they are not anywhere near as good as the Trimble we tested this weekend.

Ted VE3TRQ

Looks exactly the same as the one I bought from eBay - put it on the scope, and it has very little jitter - none that my scope could see. Either it’s good, or my scope’s jitter matches the GPSDO :-)

Bill W2PKY

Studying for ARRL FMT ran into this webpage: https://www.febo.com/pages/stability/

Interesting conclusion:
Finally, GPSDOs present two operational issues that can't be ignored: (a) they don't like to be turned off, because it takes hours to days for the system to stabilize after power on; and (b) they don't like to be moved, because optimum performance depends on the GPS knowing accurately where it is, and it takes several hours to do an adequate self-survey. These reasons mean that a GPSDO shouldn't be part of a system with a power switch, and that one may not be practical for a mobile or portable system. A battery operated OCXO could be a much better solution.

Finally, if using an internal GPSDO unit in the Flex the 12V source must be constantly applied for best performance.

Tom W3FRG

Bill,

What sort of paper work comes with the BG7TBL 2014-12-09, schematic, manual, instructions?

Tnx  Tom W3FRG

Bill W2PKY

Hi Tom-
Don't remember getting anything but there are a few articles on the internet that document the device. Something else not mentioned the external GPSDO device must be connected to the Flex Radio and turned on before the radio. The 6700 looks for the external source only at power up. 

Tom W3FRG

I intend to use it with the 6600M, which I do turn On/Off each day for use.
I will have the GPSDO on 24/7.

Tnx for the quick response.
Tom W3FRG

Bill W2PKY

yes, even the Flex Radios keep the internal GPSDO running as long as the 12V supply stays on.

Tom W3FRG

Mark,

Which Trimble unit did you test?
Did you just get the Trimble unit or was it a friends unit?

Tom W3FRG

Neal Pollack, N6YFM

If you guys that purchased the BG7TBL units on eBay opened them up and compared notes, you would find that he takes very high quality commercial GPS cards from old cell sites that are upgraded, and wraps them with a crystal oven, and a display unit, in the aluminum box.  Since his business model is getting the older GPS cards that are surplus after cell site upgrades, he is always getting different models of circuit card.   So you can open a BG7TBL GPSDO and find several different brands/models of circuit board.   I own one of the units too.  The log file inside the circuit board, available via the RS232 Connector, showed that mine used to rest on a cell site on a street corner in Tokyo.

Due to constantly changing internal circuit models and evolving design, the BG7TBL units may be of varying quality.    You may find the eevblog.com site valuable for a little information on these units.

More Information:

https://www.eevblog.com/forum/testgear/bg7tbl-gpsdo-master-reference/

http://elecraft.365791.n2.nabble.com/K3s-BG7TBL-10-MHz-reference-with-K3EXREF-td7615297.html


Neal - N6YFM

Tom W3FRG

Very interesting background history Neal. 
I also see that some might also have the Trimble inside, same outside packaging.

Interesting info.

Tnx Neal,
Tom W3FRG

Tom W3FRG

Searching I found a Manual. It appears general purpose for the BG7TBL GPSDO's.

https://www.docdroid.net/EA3BHpD/manual-bg7tbl-gpsod.pdf#page=4

Tom W3FRG

George Haslam

I just ran a test using the SDR-Kits Mini-Precision GPSDO as the 10 MHZ reference to my Flex 6700. Using the WSJT-X FreqCal, and tuned to the local CHU time signal (7.85 MHz), I am getting a delta F of .003. This makes sense to me.

Brian Denley KB1VBF

I bought the BG7TBL GPSDO and it works extremely well.  It keeps my PC clock dead on time and my 6400 dead on frequency.

NelliBalzer

Hi...what I used to do was set the 10MHz oscillator I needed to test as the reference (interface it to the reference contribution at the front).
I'd at that point measure the time of a GPS 1pps yield on the greatest 100sec entryway. You can associate this up to a PC by means of RS232 and return to-back estimations.
What I additionally did was convert the estimations back into checks of the 50MHz clock as this assisted with eliminating some commotion - the counter yields a drifting point number as a string however it is gotten from a whole number tally of 20 nsec periods, by returning to the numbers you can eliminate the adjusting blunder.

You would then be able to average the estimations over an entire day or what ever period you need.

Roger

Mark,

How old was your unit when you did the timing test? It seems some of the older units had an issue: http://www.eevblog.com/forum/testgear/bg7tbl-gpsdo-master-reference/msg949952/#msg949952