fcal - FST4W - 2.0.955 - automatic calibration not working

[PE3ES / F4VTQ]

I installed 2.0.955 to make sure FST4W is reported as FST4W and not as WSPR-2.

I set fcal active yesterday and have NTP active. 

I listen to a stabilised transmitter with a stabilised transceiver and get 14.097010 with my externally disciplined KiwiSDR which is exact and 14.097009 on my QDX which is not disciplined but close.

On the HL2 with SparkSDR I still see 9 Hz difference on 14.097019

Can I do some manual settings to correct this as the automatic correction is not working ?

Thanks, Erwin

[ahop...@googlemail.com]

Hi Erwin,

the automatic calibration does take a while to work, maybe 15 minutes, I shall check that I have not broken it.  With automatic calibration off you can adjust the calibration with the number next to the auto button, on 2.0.955 it is in ppm.

73 Alan M0NNB

[PE3ES / F4VTQ]

Hi Alan, as said it had been on for over a day / 24 hours.

I will check again and try the ppm as well.

Thanks

Op donderdag 8 juni 2023 om 10:43:22 UTC+2 schreef ahop...@googlemail.com:

[PE3ES / F4VTQ]

For me this is always counter intuitive. I had to input a + correction of 0.57 to go down in frequency by 8 Hz in the 14 MHz FST4W band. But now G3ZIL is at 1411 instead of 1419. Now to get to the correct 1410 I will go to 0,641 ;-)

Op vrijdag 9 juni 2023 om 20:54:45 UTC+2 schreef PE3ES / F4VTQ:

[Mike W]

I have revisited my Fcal problem since Erwin was reporting issues too.

I have HL2 and SparkSDR running on a 32bit Raspberry Pi 4. the image
shows a wspr tx on 70.092405 from a gps conditioned RFZero. When the
frequency is reported as 1418 the Fcal is not active and the frequency
is stable with a cal value of 1.0.

Just before 09:00 I engaged the Fcal feature and things went very
strange, random frequency reports with no correlation to the actual
waterfall display.

ideas ? 73 Mike

[Mike W]

Sorry, I appear to have sent the wrong image. apologies... correct image this time...

On 10/06/2023 11:55, Mike W wrote:

I have revisited my Fcal problem since Erwin was reporting issues too.

I have HL2 and SparkSDR 2.0.33 running on a 64bit Raspberry Pi 4. the image shows a wspr tx on 70.092405 from a gps conditioned RFZero. When the frequency is reported as 1418 the Fcal is not active and the frequency is stable with a cal value of 1.0.

Just before 08:10 I enabled the Fcal feature and things suddenly went rather funny so I reset the Fcal value to 1.0 at 08:30 and then, Just before 09:00, I enableded the Fcal feature again and things went very strange showing random frequency reports with no correlation to the actual waterfall display.

ideas ? 73 Mike

[pedion...@gmail.com]

This thread has been a help. I am installing a GPS-disciplined TCXO (Leo Bodnar Mini) on the Hermes 14 with SparkSDR 2.0.33.0.
The frequency of the new GPS-XO appeared to be stepping randomly.
So, when adding a GPS_XO  ( Or, even when checking the basic uncorrected accuracy of txcvr's master oscillator):-
In SparkSDR >Radio Settings, unset Fcal (little green indicator will go Off) . The last correction will only show when UNSET. Change this to be 1.0, leaving Fcal unset.
In linux this is auto-saved in ~/.config/m0nnb/SparkSDR2/settings/hpsdr_104-39-25-152-129-78.json.

Similarly, if using wsjt-x, >File>Settings>Frequencies : check & set slope and intercept Calibration both to 0 ppm. Do not apply the auto calibration available in >Tools
Then wsjt-x will send the uncorrected nominal frequency (eg wspr 14.095600 MHz) to SparkSDR

Regards

[PE3ES / F4VTQ]

I am revisiting this subject to set FCal on for both my radio's on SparkSDR

The Hermes Lite is using 2.0.943.0 which has a radio setting with a button for FCal and a box to set ppm correction

The RSP1a is using 2.0.955.0 which does NOT have a radio setting with a button for FCal but only the box to set ppm correction

Is there a new version that again has the FCal option or is it a radio difference ?

Regards,

Erwin

Op dinsdag 12 december 2023 om 22:41:36 UTC+1 schreef pedion...@gmail.com: