Sunday, 24 December 2017

Merry Christmas, more noise and some updates

First of all, have a great Christmas. If you don't celebrate the religious festival then have a good cultural Christmas. You might as well because there is no way of avoiding it, wherever you are.

Frankly, I find the cultural Christmas a bit of a pain. The main problem for me is that it starts in October. Then there is all sorts of goo-gorey-ness, trivia and general wasteful marketing of things which destroy the planet's resources for little benefit.

For me, this time of year is a time for reflecting. And for me this year it has definitely been about reflecting on those less well off than myself. Less well off in every sense, including health, fitness, wealth and also for those who feel excluded from society. This can be a very lonely and frightening time for many people.

Have a good time and be like me - do not over-indulge.
Thanks to several people for your emails, posts and comments over the air.

Particularly motivating were messages from Thomas OV3T and Geoff G0LUJ.

I think I have done my home-spun noise solutions to death on this blog, be here goes again.

Thomas replied to an earlier post about WSJT-X. I could not get WSJT-X to work well with my computers, and especially if I tried to use multiple instances of the software. I had not really written that piece very well, as it looked like I could not get multiple instances to work, but actually what I meant to say was that my computers are not able to process MSK quickly enough - other WSJT-X modes are fine. This seems to apply particularly to AMD processors.

As Thomas has raised the subject, I set to thinking more about it. I certainly do have a faster computer. Although it also has an AMD processor, it is an eight core and quite fast. The problem with it is that it makes a lot of noise, especially on the 2 metre band. In the past I have tried and rejected it as it simply swamps my 2m receiver with hash.

Geoff has also been tackling noise. As well as everything else, he has started a blog - here. Reading Geoff's thoughts set me to work. I decided to install the faster computer and just deal with the noise.

I do like a challenge. I have looked at this machine several times, and each time it produced enough noise to render 2m impossible to use. Here we go again.

After a lot of tinkering, including disconnecting every peripheral device, and removing most of the plug-in boards, I drew a blank. The faster computer was adding a good 20dB to the noise. Then I tried disconnecting the case fans - and the S-meter fell by 12dB.
The rogue computer case fan
There are two case fans, the one at the front does not seem so bad but I have left it disconnected for now. The rear one was making a nasty chopping noise and I  have removed it. It had been running continuously while the PC was running. I have replaced it with a new Arctic Fan Pro temperature controlled fan with a screened lead. The speed varies with the temperature detected by the fan's temperature sensor.

Moving on I found that most of the remaining noise seemed to be coming from the SVGA display lead - which oddly enough was on the previous computer too. Much more tinkering, with moving over to a DVI lead, eventually proved that the noise was coming principally straight off the Asus display board's plugs. So out came the display board and it was swapped with the one on my "Office" computer (which has an Intel processor and which runs only the IC-7300). The Office computer has an on-board display driver and, while it was not quite as good as the separate board, it can manage without a dedicated display board for now.
The dusty rogue display board (with original fitted fan).
This Asus display board is one of two I have - both the same model but the later one came without a fan. It just has a bigger heat sink and the label "High EMC Protection". The one I swapped over is a cheap generic one with no fan and it seems quiet. I do have one application where the fan-equipped board is useful because it has a low profile form factor and can fit in smaller PC boxes. It is banished from the shack though.

The result was another drop of 6dB on the S-meter. With a more normal amount of pre-amplification than I regularly use the S meter is at zero all the time, unless there is a signal on 2m SSB, which is a rare thing these days.

So I am almost there. I can still detect on 2m if this faster computer is in use. The difference is now small enough to work people. You can never have perfection.

The faster computer now drives WSJX-X or MSHV on my FT-817, TS-590 and IC-7100. It is vastly better than the slower 4 core computer. Two instances of MSK144 at the highest settings still shortens the trace by about half a second. However, I generally do not use the highest settings, preferring normal decode and FTol set to 100.

Typically I would run WSPR on 10m, and listen on MSK144 on both 2m and 6m. That is perfectly possible now if I use MSHV for 2m.

The point of all this is to reinforce the idea I put forward in previous postings. The solution to most of my noise problems lies in the shack. Sure, at Christmas there is a lot of noise from down the street, where one house is covered with flashing LED icicles. Most of this internal noise can be cured, and after Christmas I hope the festive lights will come down.

A basic desk top PC has a box which forms a large Faraday cage. This is pretty effective at keeping noise inside the case. It is the cables and plugs that seem to let the noise out. The case fan is beside several holes but I think that the reason for that problem was simple - it was just awfully noisy. And the display board was connected to a long cable to carry its fan noise out of the box.

Although the computer case is fairly effective, the front USB sockets, the front audio sockets and any fans mounted in the front will be fed by cables passing outside the metal case. The sockets are usually mounted on a plastic front panel. In my example, the simple twisted pair wires cross the void and feed to the USB sockets. This could be a big problem. Whether it is the sockets or the cable is not clear, but a simple solution is present - I don't use the front sockets unless it is for a very short term hook up.

So I have managed to improve things a bit and avoid buying a better PC by tackling the noise. I have another set of rear USB sockets to fit to the fast computer but I am not sure if it is worth it right now. I have the box nicely sealed up and noise-tight. Maybe I should just leave things as they are.

I am pretty sure that the snag remains that AMD processors (or at least those of my vintage) are not very good at handling WSJT-X. I doubt if this is a plot, but probably WSJT-X was designed to work on Intels simply because those designing it used Intels. And they probably did not do much with two instances of MSK144. Anyway, I have saved a few bob here by silencing and using an otherwise good computer.
There is not much to report on the radio front as I have been pretty busy with other things. However there have been a few good Es openings on 10m, associated with increased solar activity. I will not go into all the details, but here is one example
28Mhz Es spots on WSPR on 9 December 2017
There is no doubt that these openings are limited geographically (though 10m WSPR coverage is not well spread out in Europe so it is difficult to be sure). They are also usually short. I enjoy them though.

There was also a 6m Es event which did not extend beyond Central Europe to reach me. This was not associated with any solar activity that I could find and I think this was a "Christmas Es" event. There may be more - it is worth watching 6m from now until the middle of January. Christmas Es events can be very good indeed, if you can catch them.

I continue to use 4m meteor scatter regularly. My contact tally with Henning, OZ1JXY, has now reached 70 on meteor scatter alone, which proves that this is a really effective day-to-day method of propagation. Then again, a nice contact with Fred, G4BWP, adds another station to the list I have worked via Es, aurora, tropo and meteor scatter, on SSB, CW and data. That list, of multiple contacts in multiple modes via multiple propagation methods continues to grow, showing how many aspects of this hobby are available.

Thanks to both Richard GI4DOH and David GM4JJJ for alerting me to an article in QST about Es. I have to say that this was one of the least scientifically sensible articles I have read for some time (back in fact to the very dodgy ideas about jet streams). For any theory to stand up it needs to show a mechanism causes the event, not is merely vaguely associated with it. This latest article was (mostly) very wide of the mark.

For some reason I do not understand, people are perfectly happy to associate the variabilities in the D and F layer with solar events and the diurnal and solar cycles. The same goes for E-layer events in general. But when it comes to Sporadic E there seems to be a need to associate those events with storms and winds. Here's a clue which they might like to follow up before charging into print with their scrappy ideas - storms and winds are down here, the Sporadic E is 100km up in the sky. And there is another ionised layer (the D layer) to pass through on the way.

Also, the absence or presence of an amateur signal has got more to do with whether there is anybody there to send it than anything else. An elaborate explanation of why I hear more Es to the South East than to the South West could be worked out, but then in one direction I have Italy, and in the other direction I have the Atlantic Ocean. I suspect that there are more operators in Italy than in the middle of the Atlantic. Something to think about?

It is a pity because that QST article did touch on a few important issues which get very little attention - for example variation in the Earth's magnetic field. This is often associated (carefully chosen word there) with variations in the Earth's crust, especially at Earthquake and subduction zones. Someone comparing the magnetic field and the incidence of Es might be able to write an interesting article. But any such article would need to take account of the fact that more than half the world's amateurs do not have access to 6m, and that activity in the countries which do have the band varies enormously.

Also, and here is a thing somebody should consider before someone rushes into print - sometimes it is night and those funny foreign people don't come on so much after they go to bed, even when it is daytime here.

I have been reading some interesting stuff. Research has suggested a link between thunderstorms on Earth and cosmic ray bursts. This needs more investigation, but you can see that the chain of ionisation down through the ionosphere could indeed bring many lower energy particles down into the weather zone and have some effect.

Even more surprising was a piece on the BBC "Inside Science" programme which explained that two independent groups of researchers had found evidence of atom splitting in high-intensity thunderstorms. These strong storms are rare, but the effect may be similar but weaker in everyday thunder storms. They found not only that the strong electric fields and currents cause ionisation (which we knew from the lightening), but also they could be strong enough to create X-Rays and Gamma rays, which in turn create isotopes of surrounding atmospheric atoms. Indeed, not only did this split atoms, but they found evidence of positrons. Anti-matter made in thunderstorms (admittedly very powerful storms).

Phew. Long lasting isotopes (with half-lives in the tens of minutes) and anti-matter created by nature on Earth. These storm events are very powerful so perhaps I should not be surprised, but I am.

All this is pretty amazing stuff. I have grave doubts about the simple linkages suggested by people between weather and Sporadic E. OK, they might share the same basic source, such as the first research into cosmic rays suggests (though this is still fairly far-fetched). But the second piece of research prompts me to accept that another linkage is possible, that thunderstorms could create high energy particles which could well travel towards the ionosphere. Maybe in less energetic storms this could play a part too. I doubt it.

I am not about to rush off and write an article which links the variation in Es events to thunderstorms, based on these ideas. That is because they probably don't cause Es. I do wish that we were not regularly subjected to the "I got up one summer, looked out at the roses, and realised that rose flowering and Es are related"-type articles. Yes, roses and Es share the times of their incidence. But that does not mean they are linked by the rose perfume rising up to the E layer and activating the ions.

Hey, maybe I could get some payment for an article based on just that.

The way that science progresses is for people to be proved wrong when better theories appear. I may be proved wrong. That is fine with me. But where are those better theories? All I see are ideas that might come from the brain of Professor Frink in The Simpsons.
Again, please enjoy your Christmas.

Activity is usually up around Christmas. Must be due to the roses.



No comments:

Post a Comment