Author Topic: Lady Heather graph help, please  (Read 4648 times)

0 Members and 1 Guest are viewing this topic.

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Lady Heather graph help, please
« on: July 18, 2018, 12:24:11 am »
I have Lady Heather setup with nearly standard initialization file.  I looked everywhere I can think of but I cannot find explanations for the graph at left bottom.  Would anybody be so kind to explain this to me?  I'm pretty sure I did not change anything to alter the default behavior of the graph.  Please ignore the unusual nature of the yellow bar.  I forced it by ducting some cold air into the setup just to see what it does.

Starting with the purple line, it says rems, ref, pps.  So I think this is a deviation of PPS.  "ref" is zero.  Where IS the reference?  Top of the graph or bottom?  Does "rms" mean root mean average of what's displayed?  The actual value does not match what's under "Receiver data". 

Green one.  This must be the output of D/A converter.  Same questions....  and why do I need to know this?  At least, in my instance, it closely tracks the first.

White one...  Is this the temperature of the Xtal oven?? 

Now, what's the ideal situation?  What would the graph look like?  I would think all of the lines fairly flat?


Thank you.  Any pointers, including pointing me to somewhere someone has already explained would be very helpful.
 

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Re: Lady Heather graph help, please
« Reply #1 on: July 18, 2018, 04:00:07 pm »
Wait....  12 hours later, still no comments?  Not even good natured fun replies?  What did I DO?
 

Offline metrologist

  • Super Contributor
  • ***
  • Posts: 2251
  • Country: 00
Re: Lady Heather graph help, please
« Reply #2 on: July 18, 2018, 06:02:55 pm »
The items are what is specified at the top of the graph.

I think PPS is referenced to 0 at the center of the graph. You get the instantaneous and historical data in the trace. If you move your mouse over the lines, the tabular data is updated by the mouse (cursor) position.

I think the others are what you said, but I cannot read your screen display well enough.

My DAC reading is showing 20.0 uV/div. That may not ideally be a flat trace as it is showing the dac tuning adjustment.

I would think PPS and DAC could track or be in symmetry.

Temp will usually vary as well - 200m°C/div.

Try seeing the online KE5FX TBolt. Usually there is a desktop icon that gets installed.
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: Lady Heather graph help, please
« Reply #3 on: July 18, 2018, 07:52:42 pm »
The plot references are always the center of the plot area.  That grid line has > and < markers on ends.

Plot scale factors and the reference values can either be fixed values or auto-scaled / auto centered.

Heather calculates various statistics values (RMS, AVG, SDEV, VAR, MIN, MAX, SPAN, etc) for each plot.  The value shown on the screen is for the portion of the plot queue data shown in the plot area.  You can select which statistic you want to see.

All this is documented in the .pdf file or the comments at the start of the heather.cpp file... please read it.

Also, it looks like you changed the color map for the temperature value... it is normally YELLOW.
 
The following users thanked this post: TiN

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Re: Lady Heather graph help, please
« Reply #4 on: July 18, 2018, 10:05:16 pm »
That makes good sense now, thank you, gentlemen...!

By the way, that line is YELLOW.  It just didn't show on that photograph very well. 

A question about the blue line though....  By placing my pointer, I see that it is a TIME itself.  What do I you make it out that it is at bottom of the graph, meaning at almost negative maxima?  How does it know the variance of the time?  What is it referencing it from? 
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: Lady Heather graph help, please
« Reply #5 on: July 19, 2018, 12:24:08 am »
The blue (actually cyan) line is the number of tracked satellites.
 

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Re: Lady Heather graph help, please
« Reply #6 on: July 19, 2018, 03:04:04 am »
How many things can I get in a thread?

My current count is 3 and tend is rising....
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: Lady Heather graph help, please
« Reply #7 on: July 19, 2018, 03:14:46 am »
How many things can I get in a thread?

My current count is 3 and tend is rising....

I don't understand your question...   Heather does not use threads.
 

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Re: Lady Heather graph help, please
« Reply #8 on: July 19, 2018, 03:39:30 am »
"How many things CAN I get WRONG in an EEVBLOG thread?" is what it should have read.


My current count NOW is 4 and the trend is definitely rising....


Thank you, I got what I needed for now.  Now I need to figure out how to run two copies of this on one PC.  I have two GPSDO.  I want to run LH for each on one PC.  That will be for tomorrow.   I've run out of thinking juice.
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: Lady Heather graph help, please
« Reply #9 on: July 19, 2018, 04:20:49 am »
Just fire up another instance of Heather.   You can specify a particular .cfg file on the command line with the /r= command line option like:
   /r=heather2.cfg
 
The following users thanked this post: BravoV

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7549
  • Country: 00
  • +++ ATH1
Re: Lady Heather graph help, please
« Reply #10 on: July 19, 2018, 04:33:41 am »
Just fire up another instance of Heather.   You can specify a particular .cfg file on the command line with the /r= command line option like:
   /r=heather2.cfg

Just knew this as multiple instances of LH can run concurrently, thanks.  :clap:

All running at the same serial too ?

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: Lady Heather graph help, please
« Reply #11 on: July 19, 2018, 04:33:32 pm »
No,  only one instance can have access to a serial port at the same time.   

Heather can echo the receiver data stream to another serial port which can then be connected to a third port running on a second instance... not sure why one would want to do that except maybe as a remote monitoring device...   Heather can also convert the receiver data to NMEA format and echo it to a port.
 

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Re: Lady Heather graph help, please
« Reply #12 on: July 20, 2018, 11:59:55 pm »
Thank you.  I got them to work!

 

Offline tkamiyaTopic starter

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: us
Re: Lady Heather graph help, please
« Reply #13 on: July 21, 2018, 12:01:55 am »
Here's the GPS setup.  Another antenna is outside.
 

Offline bson

  • Supporter
  • ****
  • Posts: 2464
  • Country: us
Re: Lady Heather graph help, please
« Reply #14 on: July 21, 2018, 08:32:28 pm »
No,  only one instance can have access to a serial port at the same time.   

Heather can echo the receiver data stream to another serial port which can then be connected to a third port running on a second instance... not sure why one would want to do that except maybe as a remote monitoring device...   Heather can also convert the receiver data to NMEA format and echo it to a port.
You might want LH to manage the GPSDO while NMEA sentences and PPS output are fed to a stratum 0 (1 actually) NTP server.
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: Lady Heather graph help, please
« Reply #15 on: July 21, 2018, 11:31:52 pm »
You might want LH to manage the GPSDO while NMEA sentences and PPS output are fed to a stratum 0 (1 actually) NTP server.

That was the original intention.   Heather echoes the data in NMEA format whenever it sees a time code message from the reveiver.  But the echoed data is slightly delayed from when the receiver sent the time code message.   Depending upon the vagaries of when the receiver send the time code message in relation to the 1PPS signal, this might cause problems.
 
The following users thanked this post: Housedad

Offline bson

  • Supporter
  • ****
  • Posts: 2464
  • Country: us
Re: Lady Heather graph help, please
« Reply #16 on: July 30, 2018, 09:33:49 am »
I don't think it matters that much, as long as they ideally arrive within the correct PPS window, but more specifically arrive at a fixed latency.  NTP (actually, xntpd, not sure about others) doesn't care when the sentences actually arrive and only uses them for initial time sync.  The skew between the NMEA strings and PPS is configurable so it can knows which pulse and sentence go together.  If the sentences arrive within the correct PPS window (and I think what matters here is the time xntpd is woken from select()/epoll()/kevent()/WSAWaitForMultipleEvents() when there is at least one serial char in the read buffer) then the default zero skew will work out of the box.
 

Offline W0MPM

  • Contributor
  • Posts: 11
  • Country: us
Re: Lady Heather graph help, please
« Reply #17 on: November 25, 2018, 02:01:12 pm »
This thread answered one of my questions; what is the blue line - number of sats tracking.

I have another or two so closely related that perhaps this is a better place than a new thread.

Just acquired one of the Chineses GPSDO which Heather identifies as Dev: 57964-80 Mfg: Trimble. Been running about 20 hours and all seems good. Except:

Lady Heather OSC is 000000 and the graph line (white) also is flat. Since I am primarily interested in a 10Mhz standard, what does this imply regarding the 10Mhz signal?

The Operation Mode is Recovery. For a while it was Settling which then alternated with Recovery. Now seems to be Recovery. edit: nope - Recovery with occasional Settling. Shouldn't it become Locked or Disciplined or some such?

Taken together I'm suspicious the oscillator is free running and not disciplined?

FWIW monitoring the sine output with two Fluke 1925A counters (9 digit 0.1 Hz resolution oven time base) and the count(s) only occasionally flicker to +- 0.1 Hz. So if it's wandering it's well less than my ability to measure. For now. Still, want it to be working correctly.

Thanks. John
« Last Edit: November 26, 2018, 12:20:52 pm by W0MPM »
 

Offline metrologist

  • Super Contributor
  • ***
  • Posts: 2251
  • Country: 00
Re: Lady Heather graph help, please
« Reply #18 on: November 26, 2018, 04:24:46 pm »
Taken together I'm suspicious the oscillator is free running and not disciplined?

FWIW monitoring the sine output with two Fluke 1925A counters (9 digit 0.1 Hz resolution oven time base) and the count(s) only occasionally flicker to +- 0.1 Hz. So if it's wandering it's well less than my ability to measure. For now. Still, want it to be working correctly.

Thanks. John

If you have an oscilloscope, you can trigger a pps signal or other GPS referenced signal and then monitor the 10 MHz of your GPSDO on another channel. You'll be able to see drift a lot easier than trying to measure it.
 
The following users thanked this post: W0MPM

Offline W0MPM

  • Contributor
  • Posts: 11
  • Country: us
Re: Lady Heather graph help, please
« Reply #19 on: November 26, 2018, 09:34:50 pm »
Thank you metrologist. I'll play with that.

As to my other questions; well as the old saying goes - "pfm" - some kinda magic. Took the unit to the local weekly lunch for show-and-tell. Plugged in after lunch and after awhile two notable changes. It's now "locked" and the "loop" parameter (which has replaced "osc" with the installation of Heather 6 beta) show values.

Course I have no idea what that all means  :-// but for now all is good. Looks like the 'Metrology' section might be the logical place to go for information on GPSDO in general and talking with the Lady Heather.
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: Lady Heather graph help, please
« Reply #20 on: November 28, 2018, 01:18:25 am »
Which way is up?
"What the large print giveth, the small print taketh away."
 
The following users thanked this post: W0MPM


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf