ARRL FMT -- getting the most accurate frequency

Bob Brown - N8OB

I submitted my readings for the latest ARRL FMT and the results they printed are impossible to be anywhere near correct.  It says I was reading a frequency of 100 million.  If you subtract the reading for each band, my readings are very close to the transmitted frequencies.  Check it out...N8OB

Ria

I got curious to find out what the transmitter was. Found this:

http://www.k5cm.com

K5CM FMT transmitting equipment:

A HP Z3801 GPS frequency source clocks a HP-3336B. A 10 db transistor amp follows and drives a 12BY7 / pair 6146. Then to the final amplifier running 300 to 400 watts out.  There are no mixers or any device in the chain of amplifiers that might impact the accuracy of the GPS frequency source.  As a reality check,  I periodically log the frequency source against WWV and other GPS referenced sources.  I monitor each transmission with a completely separate receive system to look for any instability or short term drift. The separate system is not locked to the GPS transmitting system.

I typically see considerably less than 1 mHz difference at 10 Mhz when comparing  two Z3801's or a Z3801 and a Tbolt, over a 10 second period. 

 

Antennas:

160 - Vertical 

80 -  Cage Dipole (favors East, West, and North) or Dipole slopping down from 140' tower.

40 - Dipole hanging between two 80' towers (favors East, West)

30 -  Dipole

20,17,15,12,10  - 4 element SteppIR 

 

 W6OQI FMT transmitting equipment:

My transmitting setup consists of a HP Z3801 clocking a HP 3336B which drives a Johnson Viking I transmitter, the output of which is attenuated by a 6 dB attenuator, to then drive an Icom PW-1 amplifier to about 500 Watts. The antenna is an inverted Vee which favors north and south from the Los Angeles area. The QTH is actually La Canada, CA at 2000 foot elevation on the mountains on the north side of the Los Angeles basin. The Rose Bowl in Pasadena is about three miles to the east.  

I am amazed that these guys went pretty old school for their transmitters. 

Mike va3mw

Did you submit the readings with the decimal in the correct place? It was to be in Hertz.

Dave KD5FX

Steve - N5AC Can you answer my question above?

Bob Brown - N8OB

It was submitted in HZ.  In the past, it would not accept the wrong info.

Mike va3mw

On the results page you entered frequencies around 92mhz. What did you mean to enter?

Bob Brown - N8OB

K5CM - 80  3598118.50
K5CM - 40  7064197.48
W8RKO - 80  3598215.73
W8RKO - 40  7064083.50

These were the readings I measured and thought I entered.
N8OB - Bob

Mike va3mw

This is what you entered.

It looks like 100Mhz on each field.  That must be a bug, so I would email them and ask them to update your entry.

Your results look great and about what I got with my 6300.  

73

Bob Brown - N8OB

I can't find an Email address to send to for them

Ria

The latest version of WSJT-X (1.8 RC) brings a GUI for the frequency measuring tools. Way cool. This is my 6700 (stock with OCXO) on 15 megs. 

Mike va3mw

That is cool.  I will have to see how my stock 6300 plays out.  I was getting used to the command line stuff too.  

Ria

Mode > FreqCal.

Bob Brown - N8OB

I would like to see the same test with an installed GPSDO

Ria

It's on my list of things to get.

Bill W2PKY

If thinking of participating in the WWV 100 FMT Thursday evening, might want to review this thread where Steve documents how to recover the slice error. Below is my slice errors which equate to roughly -.02 - +.02 Hz. The error can be included in the email message to the researchers.

C1|slice get_error 0
R1|0|+20.877246,+11.444087

c1|slice ger_error 1
R0|0|-13.668720,+11.444087

c1|slice ger_error 2
R0|0|+3.970396,+11.444087
                       
c1|slice ger_error 3
R0|0|-18.307796,+11.444087
                         
c1|slice ger_error 4
R0|0|+3.814694,+11.444087

c1|slice ger_error 5
R0|0|+15.260337,+11.444087

c1|slice ger_error 6
R0|0|+7.631039,+11.444087

c1|slice ger_error 7
R0|0|-22.887599,+11.444087

Ronald Panetta

I recently started to play with FMT and found this thread. I have several questions as a result of some experimentation earlier today:

1) Should I be concerned over the large RX error? If I read the thread correctly that is mHz which is 20.042Hz. 
c1|slice get_error 0
R1|0|+20042.419434,+10.652245

2) Using the 6600 and WSJT FreqCal, I measure a 5MHz reference at  4,999,996.191Hz over a 4 minute duration. A Fluke frequency counter concurrently measures the same reference signal to within 10uHz of 5MHz. Given the info in this thread I would have thought I would be much closer. Any thoughts as to why the difference? Summary of the WSJT reported data summarized in Excel is as follows:

Average of Measured Frequency   4,999,996.191
Min of Measured Frequency          4,999,995.822
Max of Measured Frequency         4,999,996.742
StdDev of Measured Frequency     0.151445633

Some insight regarding the configuration. The 5MHz reference signal is generated by an HP ESG signal generator. That 5JHz reference is feeds both the Fluke frequency counter and ANT1 of the 6600. That signal generator, the Fluke frequency counter and the 6600 (yes, it is using the external 10MHz reference) all share a common time base via a TAPR distribution amp. That distribution amp is fed by a Rb oscillator (Symmetricom X72) disciplined by a GPSDO (one of the "good" BG7TBL GPSDOs).

Thanks,
Ron

Mike-VA3MW

Is your frequency offset setting set to ZERO?

Ronald Panetta

If you are referring to Flex Offset (in ppb) no I did not. No mention of that in the original post unless I missed it. Is that a factor when using an external 10MHz timebase input? See screen snapshot below. I did not start a frequency calibration either.

Ron

Mike-VA3MW

End of 4th paragraph

Ronald Panetta

So applies to 6600 (and likely 6400) in addition to 6500 and 6700. Text implies specific to those 2 models. Will rerun tomorrow and report back. Thanks!

Ron

Mike-VA3MW

Any 6400, 6500, 6600 or 6700 with a GPSDO or a 10Mhz reference.  

You can use a 6300, but you will need to calibrate it to known good signals.  A few years ago, I got to less than 1 Hz on all transmissions with a 6300 and using the FreqCal that is built into WSJT.  Don't wait for the last moment as it does take a bit of setup.

It would be very cool to see a LOT of Flex owners give this a shot and then submit their results.  I might start a new community entry to generate some interest for the November FMT.

Mike

Ronald Panetta

Mike,
That (zeroing out the offset) resolved the issue. I now get:

RX error (in milliHertz) = -3.814697

TX error (in milliHertz) = +3.914697

BTW, is there any reason the offset is ignored when using a internal/external GPSDO. Seems like it's primary reason for existence is to a correction mechanism for the internal TCXO/OCXO?

So moving on to your comment to "see a LOTof Flex owners". I started to put together a how to that, if you're willing to work with me, I'll incorporate any corrections/updates. Below is the current table of contents. If you are willing to review it, please send me an email to ron_at_panman_dot_us

Thanks,

Ron

Mike-VA3MW

Hi Ron

I'm sure several of us would love to help you out by reviewing it.  You can email it to me at michael@flex-radio.com.  

And, I think the reason we zero it out is that we expect the GPSDO or the 10Mhz reference to be more accurate than the onboard TCXO.  However, I have asked Engineering to be sure.  

Ronald Panetta

Draft sent via email.

Thanks,
Ron

John KB4DU

Anything new here? FMT coming up.Can I help with anything?

Ronald Panetta

Thought the Flex folk were going to post the guide. Have not heard any feedback from them. If you email me (email good on QRZ.COM) I'll email you the current version.

Mike-VA3MW

My bad. Let my see about it tomorrow morning.  

Mike

Ronald Panetta

Mike, Thanks!
Ron

Ronald Panetta

Those interested in my FMT document can find it viia https://helpdesk.flexradio.com/hc/en-us/articles/360035279112-ARRL-FMT-Technique-with-FLEX-6000-Series-Radios. Thanks to the Flex folk for their insight and posting it on my behalf. I will continue to maintain the master. Feel free to either post suggestions/corrections here or email to me (email address is good on QRZ.COM). Good luck on the FMT.

73,
Ron, WB2WGH 

W9WLX_Chuck

Thanks, Ron.  Great document for setting up FMT.  I did some testing on my 6600 with GPS DO and I noticed one inconsistency.  On page 2, Steve stated that there were improvements at SmartSDR 1.10.9 and we should see errors +/- 28.61 mHz.  I ran the test at four frequencies and got these RX_Error values:
0.6185 MHz      19.073486 mHz
1.1285 MHz      -3.81300 mHz
9.9985 MHz      38.146973 mHz
14.9985MHz     -22.88184 mHz

I was surprised that the get_error query returned a value of 38.146973 based on Steve's statement.  Any ideas? I have 14 satellites visible and tracking 11.

Chuck, W9WLX