K3EXREF question

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

K3EXREF question

G4HUE
Sometime ago I installed the K3EXREF which I feed from an external 10MHz double oven Xtal oscillator. I am confident the K3EXREF is installed and configured correctly as when in use and the external oscillator is connected the front panel says REF*CAL with the * blinking. Attempting to change the local oscillator frequency gives the message "Locked" just as the manual says.
However, I am puzzled as the local oscillator frequency display slowly drifts down from 49.379.675Hz to 49.379.622Hz over a period of at least an hour from K3 switch on.  I leave the external oscillator on permanently. I was expecting the local oscillator to rapidly get to its final frequency. Can someone explain whats happening as I am puzzled??
My firmware is the latest version.
Andy G4HUE
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: K3EXREF question

David Pratt
Andy - Without the K3EXREF connected, the REF CAL frequency will be
fixed and the K3 will drift, albeit very slightly, particularly when it
is first switched on.  With K3EXREF locked with your 10MHz standard the
REF CAL frequency will vary to correct any drift in the K3. If you
measure the output frequency of your K3 with an accurate frequency
meter, such as a digital counter locked to a standard, you will find
that your K3 is stable to within +/- a couple of Hertz.

73 de David G4DMP

In a recent message, ANDY NEHAN <[hidden email]> writes

>Sometime ago I installed the K3EXREF which I feed from an external
>10MHz double oven Xtal oscillator. I am confident the K3EXREF is
>installed and configured correctly as when in use and the external
>oscillator is connected the front panel says REF*CAL with the *
>blinking. Attempting to change the local oscillator frequency gives the
>message "Locked" just as the manual says.
>However, I am puzzled as the local oscillator frequency display slowly
>drifts down from 49.379.675Hz to 49.379.622Hz over a period of at least
>an hour from K3 switch on.  I leave the external oscillator on
>permanently. I was expecting the local oscillator to rapidly get to its
>final frequency. Can someone explain whats happening as I am puzzled??
>My firmware is the latest version.
>Andy G4HUE
--
 + - - - - - - - - - - - - - - - - +
 | David M Pratt, Kippax, Leeds.   |
 | Website: http://www.g4dmp.co.uk |
 + - - - - - - - - - - - - - - - - +

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: K3EXREF question

GM4JJJ
I -think- the way it works is that when the K3EXREF is locked with the external frequency standard on 10 MHz, this is used to measure the frequency of the TCXO. The value of that frequency measurement is then put into the K3 REFCAL setting. If the TCXO drifts slightly as they all will, then the frequency difference is detected and the value in REFCAL is changed to match, hence keeping the frequency of the K3 correct, without actually phase locking the TCXO to an external reference which could degrade the low phase noise performance of the K3's synth.

Basically the K3EXREF is doing the same thing automatically as you would manually when doing the calibration, by measuring the TCXO frequency and storing that value in REFCAL. The K3EXREF acting a frequency counter with a high stability external 10 MHz reference.

Hope I have that right.

So, it would be normal for the value in REFCAL to change as things warm up even with the external reference into the K3EXREF, because the TCXO is not locked, just measured.


73 from David GM4JJJ

> On 7 Mar 2016, at 11:01, David Pratt <[hidden email]> wrote:
>
> Andy - Without the K3EXREF connected, the REF CAL frequency will be
> fixed and the K3 will drift, albeit very slightly, particularly when it
> is first switched on.  With K3EXREF locked with your 10MHz standard the
> REF CAL frequency will vary to correct any drift in the K3. If you
> measure the output frequency of your K3 with an accurate frequency
> meter, such as a digital counter locked to a standard, you will find
> that your K3 is stable to within +/- a couple of Hertz.
>
> 73 de David G4DMP
>
> In a recent message, ANDY NEHAN <[hidden email]> writes
>> Sometime ago I installed the K3EXREF which I feed from an external
>> 10MHz double oven Xtal oscillator. I am confident the K3EXREF is
>> installed and configured correctly as when in use and the external
>> oscillator is connected the front panel says REF*CAL with the *
>> blinking. Attempting to change the local oscillator frequency gives the
>> message "Locked" just as the manual says.
>> However, I am puzzled as the local oscillator frequency display slowly
>> drifts down from 49.379.675Hz to 49.379.622Hz over a period of at least
>> an hour from K3 switch on.  I leave the external oscillator on
>> permanently. I was expecting the local oscillator to rapidly get to its
>> final frequency. Can someone explain whats happening as I am puzzled??
>> My firmware is the latest version.
>> Andy G4HUE
> --
> + - - - - - - - - - - - - - - - - +
> | David M Pratt, Kippax, Leeds.   |
> | Website: http://www.g4dmp.co.uk |
> + - - - - - - - - - - - - - - - - +
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: K3EXREF question

Kevin Stover
I believe that is the way it is.
53/1000 of a hertz over an hour?
I'd give that up all day everyday for the spectacular lack of phase
noise in the K3s or K3 with the new synth.
It's not a cesium clock.

On 3/7/2016 8:02 AM, David Anderson wrote:

> I -think- the way it works is that when the K3EXREF is locked with the external frequency standard on 10 MHz, this is used to measure the frequency of the TCXO. The value of that frequency measurement is then put into the K3 REFCAL setting. If the TCXO drifts slightly as they all will, then the frequency difference is detected and the value in REFCAL is changed to match, hence keeping the frequency of the K3 correct, without actually phase locking the TCXO to an external reference which could degrade the low phase noise performance of the K3's synth.
>
> Basically the K3EXREF is doing the same thing automatically as you would manually when doing the calibration, by measuring the TCXO frequency and storing that value in REFCAL. The K3EXREF acting a frequency counter with a high stability external 10 MHz reference.
>
> Hope I have that right.
>
> So, it would be normal for the value in REFCAL to change as things warm up even with the external reference into the K3EXREF, because the TCXO is not locked, just measured.
>
>
> 73 from David GM4JJJ
>
>


--
R. Kevin Stover
AC0H
ARRL
FISTS #11993
SKCC #215
NAQCC #3441


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: K3EXREF question

Edward R Cole
In reply to this post by G4HUE
The K3EXREF circuit is essentially a AFC (sort of) where a highly
accurate 10-MHz reference is used to measure the frequency of the
TCXO and send a frequency correction "word" to control the TCXO so
that the error from 49.380 MHz is removed.

As I understand it this comparison is done about every
4-seconds.  The TCXO drifts with temperature so it will definitely
change when you first power on the K3.  This drift will slow at some
point and hover around a offset level (mine is about 49.380.080.  The
TCXO is actually tuned to correct from that so REF*CAL is only
showing the amount of offset that is being corrected.  Output of the
TCXO is not 80-Hz high.  It would be if the EXREF did not correct it.

I have the K3/10 so only have the 12w PA which does not add as much
heat in transmit as the 100w PA.  So that lessens the amount of
temperature drift of the TCXO.  The TCXO-3 is better than the std
TCXO so the improvement the EXREF makes is better with it.  On
transmit on 28-MHz using the EXREF, I see approx. +/- 2 Hz frequency
error with the TCXO-3.  This is equivalent to 0.1 ppm.  Without the
EXREF my error is five time larger or about +/- 14 Hz.

There is still a slight amount of frequency drift since the TCXO is
operating normally in between the times the 10-MHz reference checks
and corrects frequency error.  This explains why a small amount of
drift still occurs vs what happens with a phase-locked
oscillator.  The result is that the frequency will appear to look
like a ramp waveform over the 4-second interval between corrections.

The reason for choosing this form of frequency control is it
preserves the excellent low phase noise of the K3.  PLL are difficult
to make with as low phase noise.

If any of this is incorrect, perhaps one of the Elecraft engineers
will correct what I wrote.

73, Ed - KL7UW
http://www.kl7uw.com
     "Kits made by KL7UW"
Dubus Mag business:
     [hidden email]

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: K3EXREF question

Lyle Johnson
Almost.

The TCXO has no control input and is not tuned; it is only measured.

The measurement information is passed to the radio's MCU which then uses
the data to determine the correct value to command the synthesizer(s),
drive the REF CAL display, and so forth.

73,

Lyle KK7P

> The K3EXREF circuit is essentially a AFC (sort of) where a highly
> accurate 10-MHz reference is used to measure the frequency of the TCXO
> and send a frequency correction "word" to control the TCXO so that the
> error from 49.380 MHz is removed...

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: K3EXREF question

Chuck - AE4CW
Elecraft recently posted a modification to the K3REF to correct a problem caused by the K3EXREF.  Though specific to frequency changes noticeable on on 2M and 6M, I suspect it may directly relate to this discussion.
http://www.elecraft.com/K3/mods/K3EXREF%20freq%20shift%20fix.pdf
---
Chuck, AE4CW
---
Chuck, AE4CW