[Elecraft] K2: PLL reference problem: what to do?

[Rafael García]

Hello friends,

I will describe my recent problem with the PLL reference oscillator just
after 2 days I did the compensated modification. Before to ask what could I
do, I went to the manual and started to make the usual voltage and in this
case, frequency measurements.

The problem:

Two days ago I made the PLL compensated mod. (I had the small kit months ago
but I was too lazy to start the "operation", hehe). All went fine and no
problems at all.
Today, I turn on the K2 this moning. No problem. Lot of noise and almost
none CW stations on any band. At noon, I turn off the rig.
This afternoon, 4 hours later, I turned on the rig again: no signal, no
audio -in fact, with the audio gain almost at full position, something could
be heard. All bands apparently dead. I moved up/down the main dial and found
one signal all the time independently of dial position. Suspect: some went
bad in the oscillator area.

Troubleshooting:

I opened the rig, attached the frequency probe to TP3 and found an erratic
634kHz, :( Could not believe it! The voltage at Q19 was nominal except on
source: 1.5V instead of 0.8. And pin 7 in U4 (PLL synth, SCK signal) was 5V
instead of 0V.
I triple checked the solder joints, bad joints, bridge joints, etc., all
were fine. At this time, I recall someone with a similar problem very
recently. Here is the message again:

----------------------------
>Date: Wed, 28 May 2003 21:15:40 -0700
>From: Wayne Burdick <n6kr at elecraft.com>
>Organization: Elecraft
>To: Daniel Reynolds <aa0ni at yahoo.com>
>Cc: elecraft at mailman.qth.net
>Subject: [Elecraft] Re: PLL Reference Oscillator
>
>Hi Daniel,
>
>The PLL reference crystal we use is a special temperature-stable type with
a
>specific motional inductance that we custom order from ECS. Results with a
>standard type purchased from Digikey are likely to be poor. But we will
send you
>one at no charge (just contact parts at elecraft.com).
>
>There are many other possibilities besides the crystal: a poorly-tinned or
>stripped lead on the toroid in the source circuit; a hair-line solder
bridge
>that conducts once the board reaches a certain temperature, a bad PLL chip,
etc.
>You might contact Gary (support at elecraft.com) for more ideas.
>
>73,
>Wayne
>N6KR
>
>>Daniel Reynolds wrote:
> >
>> I've had an intermittent PLL reference oscillator (based on frequency
readings
>> at TP3) for about the last month. Things would perk up if I grounded or
used a
>> DMM to probe between D16/L31 and ground. All voltages seem to be correct.
I've
>> replaced Q19 (J310) and D18 (1N4148) and the situation has not improved.
Today,
>> I was unable to resuscitate the oscillator. I'm guessing that X1 may be
the
>> culprit.....
--------------------------------------

Well, after reading the message, I thought the problem could be in FET Q19,
or the RFC14, and maybe X1. Because I have not had an FET replacement, I
tried to put in parallel the quartz that came into the kit with the existing
on the RF board, without soldering (only a test touching the pads at the pc
board): the PLL oscillator came to live again! I remove the new crystal
again and the oscillator still works....
I turned off for 20 minutes and on again: the PLL still works.

What should be the solution?

This is my question. Should I change the quartz xtal for the new one? As per
building instructions, I checked that the crystal installed on my K2 was
*exactly* the same that the new one shipped (K2 serial 03330).
I have never experienced this problem before since came to live in March
this year. Only after the mod. Maybe the new L31 in series with the xtal
needs some extra capacitance?

Well, sorry for the looong message,  :) I would like any kind of suggestion.

Thanks in advance.

--
Rafael, EA4RJ
Madrid