Search This Blog

Sunday, 30 June 2013

Can we do it with less?


My 30M QRSS Arduino based beacon is still on the air; and I am receiving reports via e-mail from time to time. The most recent was from Andre, DL8WX; he is in JO30WE 716.75 km (445.390 miles), bearing 112.3 degrees from me and said "I found ur beacon last night with a strong signal on 10140.02 kHz for the first time!". The beacon project build is here:

Last night, I left my FT-DX-3000 WSPRing on 40M using the dipole I made here:

and here's the result:

Before I left it alone I measured the power out exactly using my Arduino based power meter from here:

and the list of stations heard/hearing looks like this:

So, you have to wonder, if this is achievable with 40dbM (10dBW or 10 watts), what could we achieve with far less power?

Given that the most recent 50mW QRSS beacon project is Arduino based, I could write some software to generate WSPR using that, I would need a GPS to give me an accurate time source, there's also exactly that in a (not quite finished) project from here:

or even this Hans Summers project from here will do it exactly as I describe above (GPS and WSPR):

So, time to fiddle!

As a slight aside, I'm going to Malta for a week soon and I have 9H3JM for use there; as my FT-450 is in Doha I have bought myself an FT-817 to take with me. This radio is capable of outputting much lower power levels than the other radios I have here, so perhaps the easiest option for some really low power WSPR is to use that?

Lets see.

Friday, 28 June 2013

So, Why would you want one?


It's a great question - we're talking Spectrum Analysers here, by the way - my first offering would be, why not? Second would come, because I can.... but really, let me try and show you.

None of this has any real importance, but you can say that about anything, can't you?

I've just been aligning (read fiddling with) an SSB transmitter that I made a while ago. Now, part of the circuit mixes the BFO frequency (in my case it's about 12.9MHz) with the audio that's being injected. Under normal operation this would be from a mic, but for this exercise I'm using an audio signal generator and injecting a 1KHz tone.

The signals are mixed together in a simple diode based doubly balanced mixer, and the output contains a whole pile of stuff. There's the BFO frequency, the audio frequency, the sum, the difference and all sorts of other harmonics in this signal. On a scope, when the balanced modulator is un-balanced it looks like this:

As you can see the signal is a complete mess, and very hard to see what's going on. This was a single shot take on the signal as there is so much that the 'scope cant really decide what to trigger on.

Here's the same signal on the Spectrum Analyser:

Now, this makes a whole pile of sense. The left hand peak is the LSB, it's exactly 1KHz (the audio frequency) below the next peak which is the carrier or BFO (remember I said the one of the things in the output was the difference between the carrier and Audio frequency). The right hand peak is the USB - this is the sum of the carrier and the Audio frequency. What we actually have here is an AM transmission - a carrier and two sidebands.

So it's very clear here - this part of the circuit contains some components to suppress the carrier (remove it) but the circuit needs to be balanced before this will happen correctly. Using the adjustment available and looking at the Spectrum Analyser I can adjust to this:

So, I can clearly see that the carrier is now well down into the noise. The next part of the transmitter chain is the crystal filter, that will suppress the unwanted LSB to create SSB as what we have here is DSB (double side band). Once again, using a Spectrum Analyser, I can watch the RF output and alter the BFO to place the USB inside the filter passband and the LSB well down the skirt of the filter. Only then will we have what we are trying to create; Single Sideband (SSB).

Simple, egh?

Saturday, 22 June 2013

It's Keying Time with CMOS


My chum Vince, G0ORC bought himself an antique recently in the form of a TS-830S by Kenwood. The radio is solid state except the PA which is a valve design.

The radio doesn't contain a keyer so I said I would have a bash at making something simple. My tendency was to dive straight into developing something software based using a PIC, but first I decided to use some simple logic gates. A quick trawl on the internet found me this site belonging to N1HFX:

The schematic at the link above looked to be exactly what I was looking for. The first step was to try the circuit as a prototype:

This seemed to do roughly what I expected so I built the circuit on some veroboard and stuck it in a box:

Now, once built and with a paddle connected I found if I leant on the dot or dash paddle, after a continuous stream of dots or dashes the keyer would "miss a beat" in that there would be an unexpected gap or change in the timing. Trying to investigate this problem I attached a 10x scope probe to the clock output - this resulted in the clock slowing down but the keyer never miss behaving! I figured the only thing this could actually be doing was introducing some capacitance to the clock output which seemed to be making the circuit more inherently stable.  I decided to measure the capacitance that the probe was introducing into the circuit:

This was quite interesting as 20pF is certainly enough to de-tune RF circuitry and was clearly effecting the running of this clock...So all I did was modify the schematic as follows:

I've added 18pF to the output and sped up the clock by reducing the R in the timing components.

Here's the end result in it's new home:

Fun though, egh?

Thursday, 20 June 2013

I Never Fail to be Amazed


Just been rummaging through recent email in my inbox and a digest from the QRSS Knights has caught my eye. There are a set of reports from Joze, S52AB who is in locator JN75ou.

Here's the image that one of his reports are based on (Joze has done the annotation, not me):

So the distance from IO93ga (me) to JN75ou (him) is 1439.44 km or 894.470 miles at a bearing of 116.9 degrees, so my 50mW is making it all the way to Slovenia.

Would you Adam and Eve it?

Sunday, 16 June 2013

QRSSing Again, Again


Been fiddling a little more here, and stuck the Arduino QRSS beacon inside a box - it's currently being powered by an old bench PSU that I made many years ago - this PSU has a really nice toroidal  transformer inside it.

Here's the box on the shelf in the shack:

Anyhow, PA2OHH, has a really neat on-line grabber for QRSS monitoring on 30M here:

tnx Onno!

And here is a screen grab that I've annotated to aid reading:

So, my Arduino QRSS beacon, which has a p-to-p output of about 4.8V (that's about 60mW- calculated by p-to-p squared / 400) is being decoded in PA2 land (actually in square JO33df) which is about 520km.

I'm lower in frequency than I thought, although I didn't check the frequency after I put the gubbins in the box - so a little tweak up a smidgen and all should be well.

Here's another shot from PA2OHH after I have altered the TX frequency, you can now see my call around 14.0140.135 (ish):

And this one is an image Onno sent to me as a report of my sigs:

As a quick experiment, I decided to test the characteristics of the low pass filter on the output of the QRSS beacon, here's what I find using the tracking generator in the new Spectrum Analyser thingamabob:

So the filter is passing up to 10.something MHz with little or no attenuation and the first harmonic is going to be almost 50dB down - all good!

All in all, not bad, egh?

Friday, 14 June 2013

QRSSing Again


You may remember way back when I did some mucking about with QRSS (that's really low power slow morse):

I've found another kit from Hans Summers:

This time it's an Arduino shield - so I can write and fiddle with the software to drive this myself.

Here's the kit as delivered:

it's very simple to make:

and here is the finished thing sat on top of it's Arduino host:

So, a bit of simple software to get the board to send my callsign in FSK QRSS mode, and we now need to align the beacon to be somewhere between 10.140.000 and 10.140.100 - that's quite a narrow band to aim at.

The easiest way seems to be to use the Argo software - it's a kind of spectrum monitoring thingamabob where you can set the offsets associated with the radio connected. So I tuned by FT5000 to 10.139.000 USB (because the carrier offset in USB is 1000Hz) and then entered that as the offset in the Argo software. I then adjusted the beacon to be at what looks like 10.140.060:

I dont really understand why I can see more faded images of the beacon in other parts of the spectrum, but the main beacon seems to be around 10.140.060/065. You can see the end of an "X" and then the start of my call again with a "G" then a "0".

Looking at the output of the beacon on my new Spectrum Analyser thingy, we see this:

The Analyser cant go to an accurate enough frequency, but it confirms that the RF peak is at 10.140 - so I am certainly in the right ballpark and my maths isn't out of bonk.

It looks good and clean on the 'scope too:

Might just take this back with me to A71 land and see what we can see. In the meantime I'll connect this to an antenna and see if anyone can spot me!

Fun, egh?

Tuesday, 11 June 2013

I'd Get Annoyed - If I Actually Cared


Been doing a bit of JT65ing over the past few days, using the callsign A71A from QARS:

This is my friend Ali, A71BZ, I've been trying to get over to him and others the benefits and usage of low power digital modes. It seems, however, when you live in the desert, have oodles of space for antennas and can run power without anyone even noticing, SSB has great appeal....

This is the simple setup I have been using for JT65, on the left of the FT-2000 is a gigawatt rated ATU, in front of that is the remote control for a 2K amplifier from Acom, and above the ATU is a bandpass filter box of some kind. This stuff is used when they have two stations running in the same room at the same time with antennas not very far apart. How well this works is rather dubious as the preamp in the FT-2000 is clearly destroyed - running in the setting IPO results in far stronger signals that using Amp-1 or Amp-2. All is clearly not well inside the radio.

Anyhow, my 20W on JT65 is leading to signals all over Europe and as far the other direction as Japan and China. Something is working.

Now, here's the problem. When I am in QSO with someone in JT65, I really don't expect other stations to start calling me on top of the station I'm working while I am mid QSO- there's a fixed set of exchanges in JT65... as any JT65 operator will know. I understand that the call is probably a new country for most in that mode but....

Do I need to say more - this applies as much to digital modes as any other mode:

I kind of though that the digital boys were better behaved - how wrong I was.

Not done much from the apartment yet:

Being on the 21st floor is likely to have advantages RF wise - it's not so clever when the wind is extreme and there's sandstorms outside. Time will tell...

Bad news though, egh?