[NLRS] WSPR on Tuesday

Tue Jun 10 15:53:53 EDT 2025

On Tuesday (06/10/2025 at 02:40PM -0500), Chris Cox, N0UK wrote:
> If you didn’t have the OCXO option installed in your IC-746, then it was most probably the transmit fan coming on that caused the rapid and significant drift, rather than the PA heating up.

Ya. Except I did have the OCXO and it still heated it :-(

The fan issue was known to me but I could see the drift even at low power,
when the fan never even came on.

Getting that reference oscillator outside the box (essentially) finally
cured it for me.

> I wrote up a short doc many years ago right around the time of the first WSJT release that I think is still somewhere on www.chris.org <http://www.chris.org/> and copied elsewhere for a CWXO fix (Cotton-Wool XO!).  The PA fan blows directly across the 60MHz reference crystal and the cotton wool provides enough insulation to virtually eliminate the drift.

My 746PRO had a 32.0 MHz reference so maybe you worked on an earlier model
with 60 MHz reference?

In any case, whatever you can do to isolate the osc from external thermal
influence is a very good thing, at least for these narrowband modes.

Chris NØJCF

> > On Jun 10, 2025, at 14:34, Chris Elmquist <chrise at pobox.com> wrote:
> > 
> > There is an old discussion of "drift" on the WSJT mailing list here,
> > 
> > https://wsjtx.groups.io/g/main/topic/how_does_wsjt_x_calculate/78207034
> > 
> > When WSPR was a new thing, a very long time ago, I played on HF quite
> > a bit with it and my IC-746PRO which I had at the time.
> > 
> > What I discovered was that whenever the rig went into transmit, there
> > would be subtle but noticable heating of the entire radio, even at
> > low power levels like 1, 5 or 10W.  This heating drove the reference
> > oscillator off frequency a little which manifested as a "slant" in my
> > TX frequency.  After it unkeyed, you could then see it slew back to the
> > correct receive frequency as the radio cooled back down, with another
> > slant in the received signals.
> > 
> > This drove me to build a PLL to lock the 32 MHz reference oscillator in
> > that radio to 10 MHz GPSDO which eliminated all of this drifting.
> > 
> > I did the same with this FT-736 for VHF+ use.
> > 
> > My IC-7610 for HF accepts 10 MHz external reference out of the box,
> > so I definitely use that.
> > 
> > Chris NØJCF
> > 
> > On Tuesday (06/10/2025 at 02:07PM -0500), Chris Elmquist wrote:
> >> On Tuesday (06/10/2025 at 01:53PM -0500), Dave Fugleberg wrote:
> >>> for those of you with an IC-9700...how stable to you find it to be?  I'm
> >>> seeing a lot of traces in the waterfall with significant drift (like 30-50
> >>> Hz) over the course of the 2 minute WSPR window on 70cm.  Not nearly as
> >>> pronounced on 2m.   I have yet to copy anyone on 70cm - maybe that's why.
> >>> Has anyone added that ref lock board from Leo Bodnar to their 9700?
> >> 
> >> I'll let Jon respond on his own situation but I think his -9700 is
> >> externally referenced-- as is my FT-736.  And between Jon and I, we
> >> are always exactly on the same frequency.
> >> 
> >> I am transmitting with offset 1480 Hz (per the WSJT setting) and Jon
> >> reports all of my spots on 144.49080 +/- 1 Hz and 432.301480 +/- 1 Hz.
> >> 
> >> Similarly, I spot him at 144.490500 +/- 2 Hz.  I have not seen Jon on
> >> 432 yet today but I conclude his TX offset is 1500 Hz, the default.
> >> 
> >> Yesterday when we were playing on 1296, we were both on 1296.501500
> >> exactly for each of us, on each spot-- no drift.
> >> 
> >> You can look in the WSPRNet database here,
> >> 
> >> https://www.wsprnet.org/drupal/wsprnet/spots
> >> 
> >> and set up the query to show what others are seeing from you and vise-versa
> >> as a good way to check this sort of thing.   This of course assumes those
> >> spotting are uploading their spots.
> >> 
> >> Chris NØJCF
> >> 
> >> -- 
> >> Chris Elmquist
> >> 
> > 
> > -- 
> > Chris Elmquist
> > 
> > ______________________________________________________________
> > NLRS mailing list
> > Home: https://mailman.qth.net/mailman/listinfo/nlrs
> > Help: https://mailman.qth.net/mmfaq.htm
> > Post: mailto:NLRS at mailman.qth.net
> > 
> > This list hosted by: https://www.qsl.net
> > Please help support this email list: https://www.qsl.net/donate.html
> 

-- 
Chris Elmquist