Elecraft

k2, freq readout calibration

Classic

List

Threaded

6 messages Options

pc5m, Carel

k2, freq readout calibration

Hi,

I have for some time (1 year) a well functioning k2. Never bothered to do a
proper frequency.calibration because I used the k2 as IF for VHF. Recently I
tried to do so (according to the w3fpr method), but no luck. I can't get the
dial correct and have a quite big error (2 KHz or so). It seems that there
is no change at all when performing the calibration (yes I do PLL CAL and
XFIL CAL.).

I do have one question to start with:

If you tune to a carrier (CW) signal and adjust C22 (on the control board,
to adjust the 4MHz xtal osc) what should happen ? Should the pitch of the
carrier change ? should the frequency readout change ? (With my K2 "nothing"
happens, no change in pitch, no change in readout, is this normal ?.).

Thanks in advance,

Carel

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: k2, freq readout calibration

Carel,

The K2 does not use the 4 MHz oscillator in real time for frequency
determination, most of the time it only serves as the oscillator for the
microprocessor, so its frequency is not critical. That is why nothing
obvious happens when you change C22.

The 4 MHz Reference osc is used for frequency determination during CAL FCTR,
CAL PLL and CAL FIL. Before expecting accuracy with any of these menu
procedures, you must accurately set the 4 MHz oscillator beforehand.

If you did not accurately set the 4 MHz oscillator first, that alone may
explain your dial calibration error. Let the K2 warm up for about 1/2 hour
before starting the calibration cycle.

If you have an accurate frequency counter (known accurate to 1 Hertz or
better at about 12 MHz), then you can use the counter to set the 4 MHz
reference oscillator. Set the K2 to the 7 MHz or the 10 MHz band and
connect both the internal counter probe and the external counter to test
point TP1. Enter the menu for CAL FCTR and then adjust C22 until the
frequency displayed on the K2 matches the frequency shown on the external
counter. This method is outlined in the K2 manual.

If you do not have an accurate frequency counter, you can still set C22 with
accuracy as long as you can receive a station of known frequency. A
standard station like WWV is ideal, but other broadcast stations will
suffice if their operating frequency is known (an AM station is best). Tune
the station accurately in LSB or USB mode, then set the K2 menu to CAL FCTR.
Read the frequency at TP1, then read the frequency at TP2 and subtract the
two. Adjust C22 until the result of the subtraction is equal to the
transmitting frequency of the station you used as your 'standard'. This
procedure is the one available on the Elecraft website for download (it is
not in the K2 manual). The resulting accuracy does depend on your ability
to accurately tune the station (be certain to use LSB or USB, not CW because
of the CW offset). I like to use WWV because I can use Spectrogram to
display the 500 or 600 Hz tones and know that I have it tuned accurately -
other standard stations have their own transmitting format, check the one
that you can receive.

After accurately setting the 4 MHz oscillator (and before it drifts
significantly), run CAL PLL and then enter CAL FIL - tap BAND- and cycle
through each of the BFO settings, change each one (and then change it back
to the original setting if it was previously correct) to force the K2 to
write new values into EEPROM based on the new setting of C22. You must do
this for each and every BFO - 8 for SSB, 8 for CW and if you have the RTTY
filters set, yet another 8 BFOs must be set.

When you hgave finished all that, retune the station you originally used for
a standard, and you should find it now tunes within 20 Hz of the correct
frequency. If not, you did something wrong (or out of sequence) along the
way - the sequence is important, set C22 first, then CAL PLL and then CAL
FIL.

73,
Don W3FPR

> -----Original Message-----
>
> I have for some time (1 year) a well functioning k2. Never
> bothered to do a
> proper frequency.calibration because I used the k2 as IF for VHF.
> Recently I
> tried to do so (according to the w3fpr method), but no luck. I
> can't get the
> dial correct and have a quite big error (2 KHz or so). It seems that there
> is no change at all when performing the calibration (yes I do PLL CAL and
> XFIL CAL.).
>
> I do have one question to start with:
>
> If you tune to a carrier (CW) signal and adjust C22 (on the control board,
> to adjust the 4MHz xtal osc) what should happen ? Should the pitch of the
> carrier change ? should the frequency readout change ? (With my
> K2 "nothing"
> happens, no change in pitch, no change in readout, is this normal ?.).
>
>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

pc5m, Carel

RE: k2, freq readout calibration -> now completely dead

Hi Don (and also Tom who responded),

Thanks for the detailed information.). Verified setting of C22 with both
methods (parallel freq counter on TP1 and measuring the offset of TP1 and
TP2 with a 10MHz signal). Did the CAL PLL and CAL FIL routines, But...no
good..still unable to calibrate correctly (still 1-2 KHz off). As mentioned
before I have the feeling nothing is changing at all.

So I bit the bullet and performed a "hardware reset"' (456+power Up). Set
all settings (FIL etc) again --> tried to calibrate: same result: not good.
In addition I got now 2 phenomena's:
1. First, when tuning across a carrier signal I noticed very strong clicks
when tuning away from the carrier, just every 5 KHz or so.
2. During fiddling around in the menu's (resetting my transverter option),
the K2 went all of a sudden completely dead....only backlighting and some
bargraph LED's, no freq display, nothing....

Could it be I just have a faulty CPU or EEPROM ? and part of the code
(calibration and now some other code) just is corrupt ?

Of course I could do a hardware reset again, but maybe someone has a
briljant idea :-)

Thanks in advance,

Carel.

> -----Original Message-----
> From: Don Wilhelm [mailto:[hidden email]]
> Sent: Wednesday, October 11, 2006 10:14 PM
> To: [hidden email]; [hidden email]
> Subject: RE: [Elecraft] k2, freq readout calibration
>
> Carel,
>
> The K2 does not use the 4 MHz oscillator in real time for frequency
> determination, most of the time it only serves as the oscillator for the
> microprocessor, so its frequency is not critical. That is why nothing
> obvious happens when you change C22.
>
> The 4 MHz Reference osc is used for frequency determination during CAL
> FCTR,
> CAL PLL and CAL FIL. Before expecting accuracy with any of these menu
> procedures, you must accurately set the 4 MHz oscillator beforehand.
>
> If you did not accurately set the 4 MHz oscillator first, that alone may
> explain your dial calibration error. Let the K2 warm up for about 1/2
> hour
> before starting the calibration cycle.
>
> If you have an accurate frequency counter (known accurate to 1 Hertz or
> better at about 12 MHz), then you can use the counter to set the 4 MHz
> reference oscillator. Set the K2 to the 7 MHz or the 10 MHz band and
> connect both the internal counter probe and the external counter to test
> point TP1. Enter the menu for CAL FCTR and then adjust C22 until the
> frequency displayed on the K2 matches the frequency shown on the external
> counter. This method is outlined in the K2 manual.
>
> If you do not have an accurate frequency counter, you can still set C22
> with
> accuracy as long as you can receive a station of known frequency. A
> standard station like WWV is ideal, but other broadcast stations will
> suffice if their operating frequency is known (an AM station is best).
> Tune
> the station accurately in LSB or USB mode, then set the K2 menu to CAL
> FCTR.
> Read the frequency at TP1, then read the frequency at TP2 and subtract the
> two. Adjust C22 until the result of the subtraction is equal to the
> transmitting frequency of the station you used as your 'standard'. This
> procedure is the one available on the Elecraft website for download (it is
> not in the K2 manual). The resulting accuracy does depend on your ability
> to accurately tune the station (be certain to use LSB or USB, not CW
> because
> of the CW offset). I like to use WWV because I can use Spectrogram to
> display the 500 or 600 Hz tones and know that I have it tuned accurately -
> other standard stations have their own transmitting format, check the one
> that you can receive.
>
> After accurately setting the 4 MHz oscillator (and before it drifts
> significantly), run CAL PLL and then enter CAL FIL - tap BAND- and cycle
> through each of the BFO settings, change each one (and then change it back
> to the original setting if it was previously correct) to force the K2 to
> write new values into EEPROM based on the new setting of C22. You must do
> this for each and every BFO - 8 for SSB, 8 for CW and if you have the RTTY
> filters set, yet another 8 BFOs must be set.
>
> When you hgave finished all that, retune the station you originally used
> for
> a standard, and you should find it now tunes within 20 Hz of the correct
> frequency. If not, you did something wrong (or out of sequence) along the
> way - the sequence is important, set C22 first, then CAL PLL and then CAL
> FIL.
>
> 73,
> Don W3FPR
>
>
> > -----Original Message-----
> >
> > I have for some time (1 year) a well functioning k2. Never
> > bothered to do a
> > proper frequency.calibration because I used the k2 as IF for VHF.
> > Recently I
> > tried to do so (according to the w3fpr method), but no luck. I
> > can't get the
> > dial correct and have a quite big error (2 KHz or so). It seems that
> there
> > is no change at all when performing the calibration (yes I do PLL CAL
> and
> > XFIL CAL.).
> >
> > I do have one question to start with:
> >
> > If you tune to a carrier (CW) signal and adjust C22 (on the control
> board,
> > to adjust the 4MHz xtal osc) what should happen ? Should the pitch of
> the
> > carrier change ? should the frequency readout change ? (With my
> > K2 "nothing"
> > happens, no change in pitch, no change in readout, is this normal ?.).
> >
> >
>

Don Wilhelm-3

RE: k2, freq readout calibration -> now completely dead

Carel,

Clicks every 5 kHz indeed indicate that CAL PLL is not running properly.
This is not a Microprocessor problem, so a Master Reset will not do any
further good.

If the K2 is still dead, let us know and we will look into that first by
verifying the power supply and the voltage rails inside the K2. I do not
think the power failure is related to your original problem, you likely
bumped something and got a different failure for some unrelated reason. You
may have an old unsoldered connection somewhere that is just now producing
problems.

If the K2 is working now, proceed to the steps below:

Take a detailed look at the whole PLL/VCO area. First thing, check the
resistance of RFC15 - that tiny one on the bottom of the RF Board - it
should be only a few ohms. Then check the voltages at the pins of U4 and U6
and compare to the DC voltage chart near the end of the manual (The voltages
on pins 5, 6, and 7 of U14 will likely not agree with the chart - this is
NOT the problem, just tell us what they are). While you are under the
bottom cover, check the soldering of everything near U4 and U6.

Then start with the PLL Reference Oscillator Range test - probe in TP3, menu
in CAL FCTR, tap BAND+ and record the frequency, tap BAND- and record that
frequency. Tell us the frequencies.

Then check the VCO range: Measure the voltage at the left end of RF Board
R30 when the K2 is set to 4000 kHz (should be 6.0 to 6.5 volts - or 6.5 to
7.0 volts with the K60XV installed), then measure the R30 voltage for the
other band edges. If you have the K60XV installed, I am mostly interested
in the voltage measured when the K2 is tuned to 5300 kHz if the K60XV is
installed, or the voltage when tuned to 7000 kHz without the K60XV.

The R30 voltage for all other band edges should be well within the 1.5 to
6.5 volt range without the K60XV installed or 1.0 to 7.0 volts with the
K60XV. If you find any out of range, let us know.

If all is well in the steps above, then try to run CAL PLL again (the 4MHz
oscillator should still be close enough without re-adjustment). Tell us
what the starting frequency is and watch to see if a "d" is displayed on the
right edge periodically. What was the frequency displayed just before the
End message? The CAL PLL procedure takes about 8 minutes to run.

73,
Don W3FPR

> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]]On Behalf Of pc5m, Carel
> Sent: Friday, October 13, 2006 8:08 AM
> To: [hidden email]
> Subject: RE: [Elecraft] k2, freq readout calibration -> now completely
> dead
>
>
> Hi Don (and also Tom who responded),
>
> Thanks for the detailed information.). Verified setting of C22 with both
> methods (parallel freq counter on TP1 and measuring the offset of TP1 and
> TP2 with a 10MHz signal). Did the CAL PLL and CAL FIL routines, But...no
> good..still unable to calibrate correctly (still 1-2 KHz off). As
> mentioned
> before I have the feeling nothing is changing at all.
>
> So I bit the bullet and performed a "hardware reset"' (456+power Up). Set
> all settings (FIL etc) again --> tried to calibrate: same result:
> not good.
> In addition I got now 2 phenomena's:
> 1. First, when tuning across a carrier signal I noticed very strong clicks
> when tuning away from the carrier, just every 5 KHz or so.
> 2. During fiddling around in the menu's (resetting my transverter option),
> the K2 went all of a sudden completely dead....only backlighting and some
> bargraph LED's, no freq display, nothing....
>
> Could it be I just have a faulty CPU or EEPROM ? and part of the code
> (calibration and now some other code) just is corrupt ?
>
> Of course I could do a hardware reset again, but maybe someone has a
> briljant idea :-)
>
> Thanks in advance,
>
> Carel.
>
>
>
>
> > -----Original Message-----
> > From: Don Wilhelm [mailto:[hidden email]]
> > Sent: Wednesday, October 11, 2006 10:14 PM
> > To: [hidden email]; [hidden email]
> > Subject: RE: [Elecraft] k2, freq readout calibration
> >
> > Carel,
> >
> > The K2 does not use the 4 MHz oscillator in real time for frequency
> > determination, most of the time it only serves as the oscillator for the
> > microprocessor, so its frequency is not critical. That is why nothing
> > obvious happens when you change C22.
> >
> > The 4 MHz Reference osc is used for frequency determination during CAL
> > FCTR,
> > CAL PLL and CAL FIL. Before expecting accuracy with any of these menu
> > procedures, you must accurately set the 4 MHz oscillator beforehand.
> >
> > If you did not accurately set the 4 MHz oscillator first, that alone may
> > explain your dial calibration error. Let the K2 warm up for about 1/2
> > hour
> > before starting the calibration cycle.
> >
> > If you have an accurate frequency counter (known accurate to 1 Hertz or
> > better at about 12 MHz), then you can use the counter to set the 4 MHz
> > reference oscillator. Set the K2 to the 7 MHz or the 10 MHz band and
> > connect both the internal counter probe and the external counter to test
> > point TP1. Enter the menu for CAL FCTR and then adjust C22 until the
> > frequency displayed on the K2 matches the frequency shown on
> the external
> > counter. This method is outlined in the K2 manual.
> >
> > If you do not have an accurate frequency counter, you can still set C22
> > with
> > accuracy as long as you can receive a station of known frequency. A
> > standard station like WWV is ideal, but other broadcast stations will
> > suffice if their operating frequency is known (an AM station is best).
> > Tune
> > the station accurately in LSB or USB mode, then set the K2 menu to CAL
> > FCTR.
> > Read the frequency at TP1, then read the frequency at TP2 and
> subtract the
> > two. Adjust C22 until the result of the subtraction is equal to the
> > transmitting frequency of the station you used as your 'standard'. This
> > procedure is the one available on the Elecraft website for
> download (it is
> > not in the K2 manual). The resulting accuracy does depend on
> your ability
> > to accurately tune the station (be certain to use LSB or USB, not CW
> > because
> > of the CW offset). I like to use WWV because I can use Spectrogram to
> > display the 500 or 600 Hz tones and know that I have it tuned
> accurately -
> > other standard stations have their own transmitting format,
> check the one
> > that you can receive.
> >
> > After accurately setting the 4 MHz oscillator (and before it drifts
> > significantly), run CAL PLL and then enter CAL FIL - tap BAND-
> and cycle
> > through each of the BFO settings, change each one (and then
> change it back
> > to the original setting if it was previously correct) to force the K2 to
> > write new values into EEPROM based on the new setting of C22.
> You must do
> > this for each and every BFO - 8 for SSB, 8 for CW and if you
> have the RTTY
> > filters set, yet another 8 BFOs must be set.
> >
> > When you hgave finished all that, retune the station you originally used
> > for
> > a standard, and you should find it now tunes within 20 Hz of the correct
> > frequency. If not, you did something wrong (or out of
> sequence) along the
> > way - the sequence is important, set C22 first, then CAL PLL
> and then CAL
> > FIL.
> >
> > 73,
> > Don W3FPR
> >
> >
> > > -----Original Message-----
> > >
> > > I have for some time (1 year) a well functioning k2. Never
> > > bothered to do a
> > > proper frequency.calibration because I used the k2 as IF for VHF.
> > > Recently I
> > > tried to do so (according to the w3fpr method), but no luck. I
> > > can't get the
> > > dial correct and have a quite big error (2 KHz or so). It seems that
> > there
> > > is no change at all when performing the calibration (yes I do PLL CAL
> > and
> > > XFIL CAL.).
> > >
> > > I do have one question to start with:
> > >
> > > If you tune to a carrier (CW) signal and adjust C22 (on the control
> > board,
> > > to adjust the 4MHz xtal osc) what should happen ? Should the pitch of
> > the
> > > carrier change ? should the frequency readout change ? (With my
> > > K2 "nothing"
> > > happens, no change in pitch, no change in readout, is this normal ?.).
> > >
> > >
> >
>
>
>
> _______________________________________________
> Elecraft mailing list
> Post to: [hidden email]
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
> http://mailman.qth.net/mailman/listinfo/elecraft
>
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.408 / Virus Database: 268.13.2/472 - Release Date: 10/11/2006
>
>

G3VVT

RE: k2, freq readout calibration

In reply to this post by pc5m, Carel

In a message dated 14/10/06 09:01:37 GMT Daylight Time, pc5m, Carel"
<[hidden email]>
writes:

Verified setting of C22 with both
methods (parallel freq counter on TP1 and measuring the offset of TP1 and
TP2 with a 10MHz signal).

-------------------------------------------------------------------

Hi Carel,

Have a care connecting an extra frequency counter in parallel with the K2
counter as the loading capacitance of some counters may pull the readings off
what they really are. The K2 frequency counter via the supplied probe gives
relatively light loading on the circuits you are measuring. The extra frequency
counter may not cause a problem in the case of TP1, but better to play safe
and isolate the parallel frequency counter with a high impedance probe.

I fell into this trap when setting my K2 4MHz control board clock oscillator
with a Racal Dane counter. This has a ovened xtal clock, hence retains a
fair degree of accuracy from calibration. However the extra load capacitance of
the frequency counter input undid all that and gave an erroneous reading. A
high impedance scope probe gave enough level to operate the frequency counter,
yet still provided sufficient isolation not to affect the K2.

May not be causing problems in your case, but something to be aware of.

Regards,
Bob, G3VVT
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: k2, freq readout calibration

Bob,

While directly attaching a counter without a 10X probe can pull the
frequency of the measured circuit (and that can also happen with a 10X
probe), the beauty of comparing the resulting external counter frequency
count with the K2 internal frequency count is that it does not depend on the
exact frequency of the frequency source (in this case TP1). Even if the
external counter (or the internal counter for that matter) does pull the
frequency at TP1, that frequency is not the parameter being adjusted - one
is adjusting the clocking frequency for the K2 internal counter, and if that
clock is set so both counters read the same, the internal counter will be
almost as accurate as the external counter (within the uncertainty range of
the last display digit).

I have also cautioned about attempting to measure the 4 MHz oscillator
directly - the indirect methods recommended are more accurate and less prone
to errors similar to those you mentioned. It may also turn out that the
'real' frequency for the 4 MHz oscillator is a few Hz different from the
nominal frequency, but even if that is true, it is circumvented by measuring
the displayed frequency rather than the master clocking reference itself,
and any other hidden variables are also compensated out of the 'equation'.

73,
Don W3FPR

> -----Original Message-----
>
> Hi Carel,
>
> Have a care connecting an extra frequency counter in parallel
> with the K2
> counter as the loading capacitance of some counters may pull the
> readings off
> what they really are. The K2 frequency counter via the supplied
> probe gives
> relatively light loading on the circuits you are measuring. The
> extra frequency
> counter may not cause a problem in the case of TP1, but better
> to play safe
> and isolate the parallel frequency counter with a high impedance probe.
>
> I fell into this trap when setting my K2 4MHz control board clock
> oscillator
> with a Racal Dane counter. This has a ovened xtal clock, hence retains a
> fair degree of accuracy from calibration. However the extra load
> capacitance of
> the frequency counter input undid all that and gave an erroneous
> reading. A
> high impedance scope probe gave enough level to operate the
> frequency counter,
> yet still provided sufficient isolation not to affect the K2.
>
> May not be causing problems in your case, but something to be aware of.
>
> Regards,
> Bob, G3VVT
> _______________________________________________
> Elecraft mailing list
> Post to: [hidden email]
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
> http://mailman.qth.net/mailman/listinfo/elecraft
>
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.408 / Virus Database: 268.13.3/474 - Release Date: 10/13/2006
>
>