[Elecraft] maybe I don't understand drift?

Thu Jul 24 00:36:01 2003

Thanks for the note, Ron.

I forgot to remind Tim and others of the first part of the "YR Frequency
Calibration Procedure." That is to check the frequency of one of the WWV
tones (400, 500 or 600 Hz) with the rig in AM mode so that dial calibration
doesn't enter into matters. That way any soundcard error, provided that it
is reasonably constant, can be factored into the final analysis.

Now, the K2 cannot be operated on AM so that poses a problem right there
with potential soundcard frequency errors. So, I suggest that one use
another receiver that can tune WWV in AM and run a check on the soundcard
and note any consistent error. Then when the K2 is being checked, the
soundcard error can be factored in.

BTW, most PSK31 programs such as DigiPan allow for soundcard calibration.
Instead of using the normal sampling rate of 11,025  Hz I find that I must
use 10472 in order to get the frequency exactly right on. But, once set, it
stays steady as a rock.

Another aspect of my suggestion that could stand improvement is that a
passive receive-only test really doesn't subject the radio to the
temperature rise and other effects that transmitting does. But, as a first
cut as seeing if the problem is major or minor, the receive-only test will
get you in the ballpark.

Thanks for your comments!

73/72, George
Amateur Radio W5YR -  the Yellow Rose of Texas
Fairview, TX 30 mi NE of Dallas in Collin county EM13QE
"In the 57th year and it just keeps getting better!"
<mailto:[email protected]>

----- Original Message -----
From: "Ron D'Eau Claire" <[email protected]>
To: <[email protected]>
Sent: Wednesday, July 23, 2003 11:10 PM
Subject: RE: [Elecraft] maybe I don't understand drift?

True about using Digipan to check frequency drift, George, IF the PC sound
card ref oscillator is that stable. Actually, they are quite good, bit I've
seem some drift several Hz... and quickly at times.

Another way that doesn't require the PC is to zero beat WWV using SPOT.
Leave it there for half an hour and come back and see what's going on to see
how the long term drift is doing. Simply retuning the K2 to re-zero the
signal if necessary will let you read the drift right off of the LCD.  Just
be sure you zero beat the carrier and not the sidebands on WWV, Tim. Those
tones change regularly. Check the zero beat during the voice announcements
to be sure.

The K2 is spec'd at less than 100 Hz/hour after a warmup, but mine is much,
much better than that. I think most of them are. Especially with the
stability mod installed.

You are right about the "dial indication", George. It only indicates the
frequency indirectly. When CAL PLL was run, a table of values was created
that said, "if this tuning voltage is applied, the K2 will be receiving (or
transmitting) on that frequency". A table of tuning voltages and
corresponding frequencies is stored in the EEPROM. Then, when you set the K2
tuning knob to show a certain frequency, the logic sends off the
corresponding tuning voltage to the local oscillator. If anything has
happened to change the oscillator frequency that goes with that tuning
voltage, the K2 won't be on the right frequency.

It works just like an old mechanical tuning dial. When a certain frequency
is displayed, the system assumes the receiver is actually on that frequency,
but it doesn't have any way of verifying that. All you do when you run CAL
PLL is to recalibrate the "dial". The subsequent accuracy is wholly
dependent upon the oscillator being absolutely stable after that.

Ron AC7AC
K2 # 1289

-----Original Message-----

Tim, it is uncommon for the dial reading of the rig to change if the signal
frequency is "drifting." However, depending upon how the dial frequency is
developed and displayed, whatever is causing the output frequency to change
can also cause the dial reading to change. I don't recall enough of the K2
logic to say much more.

I would think that 80 Hz drift on 40 meters in an hour's time is a bit much.
But the K2 is not a precision instrument by any means when it comes to dial
calibration and frequency stability.

I suggest that you tune in WWV on 10 MHz with your dial set at 10,001 kHz
LSB and look at the nominal 1000 Hz beatnote on a PSK31 program as a
waterfall track. DigiPan, for example, will measure the beatnote frequency
down to a tenth of a Hz very accurately. If you will note the indicated
frequency on the DigiPan display when you first turn on the radio and make a
chart of frequency vs time for an hour or so, you can get a good idea of
what the rig is doing. The dial indication will not change while making this
test, so any change in beatnote frequency is due to drift within the K2
circuitry.

73/72, George
Amateur Radio W5YR -  the Yellow Rose of Texas
Fairview, TX 30 mi NE of Dallas in Collin county EM13QE
"In the 57th year and it just keeps getting better!" <mailto:[email protected]>

_______________________________________________
Elecraft mailing list: [email protected]
You must be a list member to post to the list.
Postings must be plain text (no HTML or attachments).
See: http://mailman.qth.net/mailman/listinfo/elecraft
Elecraft Web Page: http://www.elecraft.com