[Elecraft] K2 drif problems?

[Ron D'Eau Claire]

Ed, one thing you will see on the reflector consistently are comments
and observations about how to make the Elecraft rigs better. The
Elecraft engineers jump in quite often and more than a few changes and
modifications have been made to the designs following what has gone on
here. That's one of the big advantages of the Elecraft equipment over
typical "factory built" rigs. 

Frequency drift in the K2 is one of those reoccurring issues. First it
is important to understand that the K2 does NOT use a frequency
synthesizer with a temperature-controlled crystal reference oscillator
for "state of the art" frequency stability. While it uses a Phase-Locked
Loop to generate the local oscillator signal, it accomplishes smooth
tuning by using a variable frequency oscillator as the loop reference
oscillator. That means that the frequency stability of the K2 is no
better than the frequency stability of the simple, variable frequency
oscillator. 

Elecraft uses a solid, stable design for that oscillator. Instead of a
free-running variable frequency oscillator (VFO) they used a variable
crystal oscillator (VXO) in which the frequency of a crystal-controlled
oscillator is "pulled" to provide the tuning range necessary to cover
the 5 kHz "steps" in the phase locked loop design. 

The VXO is a lot more stable than a simple VFO, but like any oscillator
that can be shifted in frequency easily, it will tend to shift in
frequency as the temperature changes - what we usually refer to as
"drift". The beat-frequency oscillator (BFO) used in the i-f section of
the receiver is also a VXO and is also subject to drift. 

Over the past two years I have followed several threads from ops who
were trying to improve the overall frequency stability of the K2. Some
of them simply wanted to see how good it could be made. Others had a
serious issue with the drift in their K2's. Most of those were operating
PSK-31. That mode really taxes the stability of any rig. A PSK31 signal
only occupies 31 HZ of bandwidth!

The concerns over the frequency stability of the K2 became a very common
issue when the KPA100 amplifier module was finally released. That module
makes a K2 a K2/100 with 100 watts PEP or CW output. It also creates a
significant amount of heat. The whole top of the K2/100 is a heat sink,
as you can see from the pictures. Those black ribs running front to back
are the top of the heat sink. Almost immediately a number of ops with
the early amplifier modules noted that the drift of the K2 was much
greater, especially on certain bands where the combined drift of the BFO
and VFO added to make the total drift greater than that of either
oscillator alone . The KPA100 amplifier pumps quite a bit of heat inside
the K2 where it warms up the components in the oscillators, making them
move in frequency. 

In response to that, the basic K2 has been improved to reduce the drift.
K2's newer than S/N 3000 have a number of improvements to reduce the
drift of both the VXO and BFO. Elecraft release a kit for the owners of
earlier K2's so they could incorporate these changes in their older
radios. 

Recently I added the KPA100 module to my K2, along with the
modifications to make it more stable. Then I ran some drift
measurements. On 15 meters - one of the bands where the drift is the
worst due to the mixing scheme used in the K2 - I find that it is
disappointingly large. After allowing my K2 to warm up for half an hour,
a CW QSO lasting half an hour of transmitting and receiving can result
in 250 Hz of frequency drift in my K2. That's enough to have moved my
signal outside of the passband of the other station's filters if he were
using a really narrow filter. It's certainly enough to cause the other
operator to "touch up" the tuning in many cases. 

That is on the band where the drift is the worst, however. I ran some
quick checks and it seems to be about half that on 40 meters, for
example. About 100 Hz.

Even that is enough to cause some re-tuning in either CW or SSB. For
PSK-31, it's a really significant amount of drift, and from time to time
you will see a lot of comments on here as ops work on ways to reduce it.
That said, that amount of drift in a rig is not at all unusual on the
Ham bands. 

Compared to the rigs of a few decades ago, the K2 has "laboratory
standard stability". But this is not 1960 and I'm not using my HRO-5
receiver filled with tubes either. Compared to the state-of-the-art rigs
today, the K2's frequency stability is competitive but a lot of
factory-built rigs are better. In true Ham fashion, many ops on this
reflector are of the opinion that the K2 can do better and are working
on ways to make it more stable. The beauty of the Elecraft rig is that
the engineers who designed it are also working on ways to make it more
stable and they participate in this reflector.

In my view, the K2 is a very good rig. It's nothing magic. It's
certainly not without faults. I've never seen a piece of equipment that
is. The biggest feature that Elecraft has to offer, in my opinion, is
their on-going support to the community of Hams who always want to make
it a "little better"...  So we jawbone and fuss over things like making
it more stable. That's what some Hams do as much as they operate. I
guess that I'm one of those.

If it weren't for those of us who want to see it work a little "better"
, this reflector probably would be a pretty dead place. There isn't much
in the way of other "problems" to discuss about the K2. Oh, there is the
other reoccurring issue of a "better" tuning knob - one with smooth
rounded edges and a dimple for "twirling". But  that's been "solved" for
those ops with a replacement knob from a Japanese rig. Personally, I
like the nicely-weighted "stock" K2 knob, so don't worry about that too
much. I always figured that if the worst issue people had to work on was
the design of the tuning knob, the rig must be pretty good. 

For me, the next thing I want to do is to  KILL that #$%@# little fan in
the KPA100. To me, it has all the attributes of a crazed mosquito in the
room. Ah... but that's another "enhancement" to think about another
day...

Ron AC7AC
K2 # 1289