[Elecraft] VCO relationship with 4 Mhz ocsillator

Thu Nov 25 23:03:33 EST 2004

That's not the case for the 4 MHz oscillator in the K2, Jean-François.

First of all, the 4 MHz clock on the control board and the trimmer capacitor
C22 have nothing directly to do with the PLL. That's why adjusting C22
doesn't change your frequency. But what you read in the handbook is right.
Let's look at that first. 

It is true that a PLL requires a reference oscillator, and the output
frequency is based on the reference oscillator frequency. One problem with
PLL's is phase noise - jitter in the frequency caused by the loop constantly
correcting the output frequency to agree with the reference oscillator. 

Elecraft provides superb phase performance by doing something unusual. They
vary the frequency of the reference oscillator. The K2's reference
oscillator is varied over a 5 kHz range. The Phase-Locked Loop tunes in 5
kHz "steps" and the reference oscillator is tuned to provide continuous
coverage within those 5 kHz steps. If you look in the archives (or have an
older K2) you'll see that a year or so ago some effort was put into adding a
temperature compensation circuit to the PLL reference oscillator in the K2.
That was because of a basic rule in radio: If the frequency of an oscillator
can be varied, it will try to drift! The K2's reference oscillator sometimes
showed annoying levels of drift, and the circuit was improved to stop that. 

Let's get back to that 4 MHz oscillator on the control board. The K2 does
something else a little differently from many radios. The frequency readout
you see on the LCD display is *not* produced by constantly monitoring the
oscillator frequency in the K2. Instead, the K2 frequency readout is a lot
like the old vacuum-tube-day dials, in which you turned a knob, that knob
varied the frequency of an oscillator, and it also moved a pointer. The
pointer moved over a scale with numbers on it that showed the frequency. The
numbers on the scale were produced by hooking up a frequency counter to the
rig and checking the frequency produced as the tuning knob was turned. The
correct numbers were then printed on the scale behind the pointer so it
could be set back to any desired frequency later. Not every frequency was
recorded - just frequencies at regular intervals. Frequencies in between the
numbers could be estimated by "interpolation". That is the frequency exactly
halfway between 7000 kHz and 7010 kHz could be assumed to be 7005 kHz, and
so on. 

That's how the K2 works, but its all done electronically. First of all,
there are no variable capacitors or other old-style tuning mechanisms used
in the K2. The oscillators - the beat frequency oscillator or BFO, and the
phase locked loop reference oscillator (called the variable frequency
oscillator or VFO in the K2 literature) are both tuned with voltages.
Voltage-variable capacitors replace the old big mechanical variable
capacitors. The voltage-variable capacitors change capacitance and therefore
the oscillator's frequency in response to the level of a direct-current
voltage applied to them. 

When you run CAL PLL, a frequency counter built into the K2 measures the PLL
frequency. The K2 applies a tuning voltage to the PLL reference oscillator
and notes the frequency measured by the counter. This information is
recorded in memory so it can be looked up later. The K2 tunes through the
whole range of frequencies the PLL can cover noting the actual frequency
measured at regular intervals, and then recording the voltage applied to the
reference that will produce the frequency. 

After CAL PLL is done, all of that data has been stored. 

Now, in normal operation, you choose a band and spin the frequency dial to a
specific number on your LCD display. When you do that, the K2 looks up the
proper voltage from the data that was stored when your ran CAL PLL and
applies it to the PLL reference oscillator. That will put the reference
oscillator on the correct frequency, according to the data that is in
memory. Of course this happens very quickly so it appears that the tuning is
happening in direct response to your turning the knob. 

Notice that the accuracy of the readout depends upon how accurately the
frequencies and tuning voltages were stored when you ran CAL PLL. That's
where the 4 MHz clock and C22 come in. The Control Board 4 MHz clock is the
time base for the frequency counter! So the closer it is to 4 MHz the more
accurately the K2 will measure the K2 PLL frequency and the more accurately
the dial calibration of the K2 will be in use. 

That's why the procedures for setting C22 all remind you to re-run CAL PLL
(and CAL FIL) after adjusting it, in order to make the K2 store new
frequency and tuning voltage data in memory. After you do that you will
notice the change in the dial calibration.

Wayne published an excellent method for setting C22 on the Elecraft web site
(www.elecraft.com). Look under Builder's resources for "Adjusting C22 to
calibrate the K2's Frequency Display".

Ron AC7AC

-----Original Message-----
From: elecraft-bounces at mailman.qth.net
[mailto:elecraft-bounces at mailman.qth.net] On Behalf Of Jean-François Ménard
Sent: Thursday, November 25, 2004 5:56 PM
Cc: Elecraft - Maling List
Subject: Re: [Elecraft] VCO relationship with 4 Mhz ocsillator

Sorry to bother you....

I read in my ARRL handbook about PLL... and now I understand that the 
PLL must have a fixed, precised, reference to phase lock it's loop.... 
the 4 Mhz oscillator clock...

I will find the place to put my scope probe to see the 4 Mhz oscillator 
also... just for my understanding.

Thanks and sorry again for your time.

Le 04-11-25, à 19:49, Jean-François Ménard a écrit :

> Could someone tell me the short story about the relationship between
> the PLL reference oscillator and the 4 Mhz oscillator from the control 
> board ???
>
> I'm asking this question because when I calibrate the PLL reference
> oscillator at TP3 using the internal frequency counter of the K2 and 
> also using my external Instek frequency counter also connected at TP3, 
> trying to match the K2 PLL frequency on both display.... Well.... I'm 
> turning C22 on control board, associated to the 4 Mhz oscillator.... 
> and that make the PLL oscillator to change.... Why ?!?!?
>
> I can see the on the control board that the MCU have an output /PLLCS
> that go to the PLL synthetizer.... that must be there, but why ?
>
> This setup seems to be particular to the K2 concept ?!?!? Right, wrong
> !?!?
>
> Thanks for your time...
>
>
> Sincerly,
>
> ===========================
> Jean-François Ménard / VA2VYZ
> va2vyz at arrl.net
>
> Elecraft K2 #4130
>
> http://homepage.mac.com/jfmenard
> =========================== 
> _______________________________________________
> Elecraft mailing list
> Post to: Elecraft at mailman.qth.net
> 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
>
===========================
Jean-François Ménard / VA2VYZ
va2vyz at arrl.net

Elecraft K2 #4130

http://homepage.mac.com/jfmenard
===========================

_______________________________________________
Elecraft mailing list
Post to: Elecraft at mailman.qth.net
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