First, what is a plus for using a digital frequency readout, or having your eye on it when operating CW?
The main benefit is that you can with a high degree of accuracy determine your operating frequency. This is very helpful if you want to avoid going outside the legal amateur radio frequency band limits, provided it is accurately calibrated, which most modern rigs will be. If not, you can do so against a Frequency Time Signal on 10MHz, for example, but this is outside the scope of this article. You can research how to do that with great accuracy.
It can also greatly help when you have a sked or net that you are looking to join, or more importantly to start calling or act as a net control. However, even in such cases, you cannot own a frequency and you will want to vary it according to occupancy, noise, interference etc. And anyone can still find you by tuning around a little plus or minus above or below the agreed frequency.
I really cannot think of any other benefits, other than if you wanted to keep a listening watch on a certain frequency with a very narrow filter, but let us now address the many draw backs of the common place use of the digital frequency meter and the psychological and practical factors that cause problems and losses in CW communications in particular.
Psychological Factor
First let's look at the psychological factor: it just feels so much better to be on a nice looking frequency such as 10.125.000 rather than some random free frequency such as, for example, 10125.475 or 10117.328, right? The human mind likes order. I speak for myself here, but I think it applies to most people, and is the actual prime reason, a psychological one, why most CW operators are to be found on 10110.0, 10111.0, 10112.0 etc or even at best 10121.5, 10122.5 etc. If you look at the band occupancy on a SDR spectrum waterfall or even just tune around, you will see that this is the case. Very few stations are to be found "in between" these 1kHz, or at best, 0.5kHz "channels".
There can also be a practical reason for this with a few rigs, such as the QCX, where the tuning steps are in 1kHz or 0.5kHz usually although one could also set them to tune in 0.1 kHz steps or even 10Hz (0.01kHz) steps etc. This makes such rigs easier to tune across the band faster and even if you set them to 1kHz or 0.5kHz steps, even with the narrow 200Hz (0.2kHz) filter, you are not likely to miss much, due to the above mentioned facts of generaly frequency occupancy.
So, why does this, mostly psychological factor, cause problems for CW operation in general? After all, it makes us feel better, seeing a nice round 000 or 500 on the display, right? Well, I'll give a few reasons why this is a bad approach that we should aim to fight against, at a personal level, this is not a demand for everyone to follow this advice, this is a request for us to reconsider how we do things, and what the down sides of this wide spread practice are, and, how we as self-reflective CW Operators can profit from this situation, to our benefit, as well as others.
QRM
One problem with this approach, is that of QRM. While some might put forward the argument that having 0.5kHz spacing is a very good thing as it minimized interference keeping CW stations apart, this is absolute nonsense in practice and easily demonstrated. The keeping of 90% or more of CW stations to round 1kHz frequencies, and 95% to round 0.5kHz frequencies, is not LESS QRM, but in fact much more, and far worse QRM!
By the way, an example of commercial professional operations not choosing RTTY or CW frequencies to be exact 1 or 0.5 kHz is DDK9 both its centre frequency 10100.8KHz and its actual two tone frequencies 10100.575 and 10101.025 kHz aren't anywhere near an exact 0.5 kHz, are they? And with good reasoning: they wanted their frequency to be free of interference as much as possible over several decades, with receivers glued to this frequency around European waters, and they have achieved this.
QRM is usually not deliberate, and it is usually caused by the nature of HF communications "dead zones" or "skip zones" between hops off the ionosphere and back to earth, as well as paths being open in one or more directions but not others. Therefore, even the best intentioned and most considerate radio amateur, who spends a minute or two quietly listening to the frequency intently, volume up high, narrow filter switched in and out, notch filters off, and moving the BFO up and down a little, before finally timidly sending a short ? and if still nothing heard, a longer QRL? and even a second and third QRL? cannot prevent the situation where his signal is actually heard by one station in an existing QSO, but, he cannot hear that station, or even both.
Not to mention the fact that very few of us are as considerate and meticulous as the above described exemplary CW operator full of patience and virtue! In fact, there exist many at the other end of the extreme, prone to a quick QRL? with the narrow filter on or off, and the antenna wherever it is, and perhaps even the RF gain turned down to avoid noise, or even a squelch turned up so that only nice strong signals delight his ears.
The result of QRM to CW, no matter how it manifested itself via the careful or careless operator, is only severe when the QRM is sitting on the exact same frequency or let's say within 50 Hz or so of the frequency, depending on its signal strength. 50Hz is plenty to separate two CW QSO in the brain if they are of similar signal strength, though it could be annoying at high volume, but 100Hz is absolutely plenty, and 200Hz is complete overkill for separating CW signals. For those who suffer from an inability to separate in the brain or ear there is always the notch filter and narrow filter that can come to the rescue, the latter in far too much widespread overuse, perhaps the subject for another post, on the downsides of over use of narrow CW filters (200, 500Hz are narrow).
You can see where I am going with this. The fact that everyone is using a round kHz or at best a round half kHz means that QRM is actually on average far worse and far more destructive to QSOs when it appears. I'm old enough to remember the delight of QRH (drifting) and QRI (chirp or buzzing) as well as dozens of stations all calling within a few kHz, what a delight! It also meant that because no one had a T9X (pure tone so common and boring these days, please, someone invent a QCX with a button to turn on chirp, if not drift, or a little modulation of the CW signal) it also meant that you could pick out each signal with its unique sound even right on the same frequency. But these days, you cannot…
This has led to stupid and inconsiderate, nay, reckless and psychopathic behaviours of incapable DXpedition operators, never mind that they're also the president of RSGB, IARU Region 2 or IOTA, the title means nothing, as 5T0JL pointed out, these are incapable LID operators, whereby instead of using some skill they solve the problem by sending "UP" and letting the fools spread out 30kHz wide, again mostly using round 1kHz frequencies, thus not solving the problem, and not listening on their frequency, and causing yet more QRM to everyone. Instead of using the wide/narrow/flexible filter of their brain, they try to use narrow filters of modern rigs that compound the problem with pure CW right on frequency!
So, what can the smart CW operator do? And what can be a good compromise to still satisfy the psychological factor without putting on a blind fold, or a masking tape across the digital display?
Either you can move the dial without looking and use your ears only, and don't correct the random frequency when you do look again, or, you can opt to use a nice ending 0 on a 10Hz readout, or at most 01, or .105 on a 100Hz readout, thus compromising with some still "need for order and control" in the mind, but mitigating the problem. So, here a few examples of such operation and its benefits:
1. Calling CQ. Tune until you find a nice clear space, without looking at the dial. If you look and it bothers you, move it to the nearest 10Hz or 50Hz or 25Hz. Have fun calling CQ. When you listen for replies, be ready to use your BFO (RIT) to listen up and down, but do not move your VFO. Chances are someone will reply off frequency, if they do, they are listening to your frequency, do not move it. There is even a benefit when both stations are not exactly on the same frequency: in the event of QRM it will likely affect one station but not both, thus making it more possible to find a solution such as QSY. In the old days, this was done easily: just put the key down or VVV while also turning the VFO so the other station followed you, when QRM came up, this was common.
2. Looking for club members, say, FISTS on 10118. Tune up to 10118.234 or 10117.987 or whatever, if it makes you feel better, 10118.1. Call there. These aren't fixed frequencies anyway, they are "Centers of Activity" — same for QRP on 10126, 10116 or 10106, unless it is a Xtal.
3. Answering stations. Well, it is good to answer right on the frequency of course. But lets say you answer a typical CQ on a round kHz. Go 50 Hz off frequency at least, give the other OP a little practice to remember where his RIT (BFO) knob is, or, call 150Hz off frequency, to aid operators to remember not to keep the 200 Hz or 40Hz filter on all the time. Of course, you may not be heard, or they may follow you, be ready for both. But calling a little off frequency will set you up better in the likely event of ZERO-BEAT QRM appearing later. Or, feign some drift, call back on frequency and drift your VFO off 87 or 99 Hz as you go, or as you please. OK I'm mostly joking here, but you get my DRIFT !
Another solution might be to simply cover the digits below the kHz or below the first hundred Hz on the display, but this is an extreme measure, it might make the rig look a bit ugly.
Anyway, know that a pure CW tone can benefit the use of narrow filters and rock solid RX these days, but, a chirp, a drift, and a non-adherence to the nonsense of CW operating on full or half kHz "channels", are all important ingredients for a more enjoyable QRM-free CW life, and, also a safeguard against future regulations appearing that mandate CW operations only on fixed channels spaced 0.5kHz apart, and only decodable 100% by automated means, and with certain speed limits and keying methods restricted. And about this, I am not joking.
77 DE AL, G4OJW