Radio - HF to Microwaves

So far observation at G3ZJO of straight line reflections from Aircraft have concentrated on the departing path as planes travel North up the UK and turn toward the USA. This has produced the now classical approx 60Hz negative Doppler shift.

Today an accidental alignment of beams at G6AVK and G3ZJO has produced AS which approaches the path between the two stations then vanishes. I surmise that these are reflections from Aircraft on the Thames Estuary approach from Europe which may well vanish when past the front of the AVK beam.

Looking at the capture at 13.30 GMT there is a straight line AS reflection which decoded, it has a 53Hz shift but this time positive Doppler. This is the first time positive straight line Doppler shift has been observed here. Again I assume from an Aircraft on the EU Thames approach heading toward G6AVK and G3ZJO at 53x540/50.294436 = 569KmH

The Modular DDS MEPT was completed ahead of the start of the Sporadic E season, I have put it on On Air Test on 28MHZ (10m). Reports on the WSPR ident have come from Portugal, Italy, Norway, Slovenia, Iceland, Germany, Spain and Greece. A very pleasing start.

Always extra pleasing is a capture or three of the Visual modes.

First in was Paolo IZ1KXQ in Italy.

Next up a capture of the WSPR display but from France F6GUU a Country we don't see many reports from.

Edit : And another from Pat F6IRF in JN35AU on the 7th of June.

Third Spain from David EA1FAQ.

The MEPT is running 300mW, the Antenna is a 6m CB type Vertical at ground level. I knew 10m was going to be good fun. this is the first time I have run a QRP MEPT on the band.

A great quote from an Alaska station today, "its so exciting I may not be able to hold my water". I know just how he feels.

Today was superb for visuals of the Airliners in the UK South / North Air Corridor. The clear blue sky made it easy to observe Aircraft travelling North 'over my back garden' and South 'over the front'.

Here is just one Plane of the constant stream which passed my QTH and my 6m Vertical Dipole today.

Right Click and select View Image for full view.

It would be interesting to find the true distance at which this one passed, I estimate 1 Mile. I think Aircraft in the Corridor can be anything from Overhead to 20 Miles away.

My last blog entry covered an experiment to prove that it is possible to get a straight line reflection from an Aircraft travelling away from both the originating, TX, station and the receiving station.

The Aircraft needs to be, I believe on the same bearing as that of the path between the two stations.

At my QTH this is merely a waiting game, there are so many aircraft on the S/N corridor that finally one will fit the equation.

Previous tests with G6AVK did not achieve the desired result. I think that, while his omni directional signal reached me fine, by the time it got to Aircraft sufficiently North of me it had run out of 'steam'.
The addition at G6AVK of a mast and rotatable 6m Beam means that strong Aircraft reflections from the North are now possible.

I pointed my beam in line with the S/N UK Air Corridor G6AVK beamed toward me.
Note the trace at 19.32. There is the direct signal from AVK. A long descending trace from an Aircraft, which has passed me and is heading North, on a track at a tangent to the path between AVK and ZJO. Third a parallel trace, ‘classic’ 61Hz shift equating to an air speed of c.600Km Hour.

This produced 2 decodes of WSPR on the two frequencies.

2009-05-29 19:32 G6AVK 50.294482 -18 1 JO01ho 5 G3ZJO IO92ng 127 306
2009-05-29 19:32 G6AVK 50.294421 -25 0 JO01ho 5 G3ZJO IO92ng 127 306

I think we can now say that this is the definitive explanation of the phenomena. Repeatable by experiment, using 3 different signal sources.

Whilst looking at the destruction that Aircraft Scatter (AS) can do to WSPR signals I received a puzzling series of signals decoded by WSPR from G8EUX.

Right Click on image and select View Image to see full size.

The grab above shows a straight line signal 59Hz below the original signal which itself appears to be effected by AS.
This lead to G8EUX testing his TX for spurious emissions and me saying it cannot be an RX odd response due to the fact that the second signal has AS. To add to my identification as a Scattered signal I never see odd decodes from G8EUX even when signals are much stronger.

After some thought I came up with an explanation.
The straight line is AS and the other related curve is from another aircraft reflecting the fundamental signal. Being a straight line reflection means the Aircraft is moving away from both stations on a constant path relative to both stations. The Doppler shift of 59Hz thereby gives us the Aircraft speed. Which works out at 633KmHr.

One example of this phenomenon does not prove anything so I began to consider a method of proof.
Beacon GB3BAA on 50MHz lies to the South East of me, not parallel to the Controlled UK Airspace South North routes but knowing the flight paths I decided to give it a try.
Pointing my beam North away from GB3BAA I set Spectran so that the beacon was just at the top of the display, only the Morse Ident Keying, lower sidebands can be seen.
I was most interested in the minus 50 to 60Hz Doppler shift region. In a short time an example of 60Hz straight line shift occurred. There are short periods of no AS then just like buses several come at once and many with a straight line portion with a Doppler shift of 50 to 70Hz.

Right Click on image and select View Image to see full size.

So what are we seeing? One explanation I can put forward is.
As the AS signal descends in frequency we are witnessing Aircraft travelling North away from both GB3BAA and G3ZJO and at an angle having already crossed the direct line signal path. As the Aircraft turn to cross the Atlantic there is a period when the trigonometry produces the straight line. I do not show the signal path from GB3BAA to G3ZJO. Its not critical, just think of GB3BAA as being down London way.

Right Click on image and select View Image to see full size.

Equally we could suspect that the Aircraft involved could be turning earlier to take the route across via Ireland, there is an example in the animated plot doing just this, I consider the angle to be too large for this to be the case.

The modular approach to the DDS MEPT has meant I can try out different methods. I used a Class A PA strip to get a good clean 250mW on 50MHz.
I thought I would try the same 2N3866 Class A strip idea for a Broad Band HF PA, to get a bit more power on 24 and 28 MHz bands. I needed a little more drive so I added a 2 transistor BC184 Driver Amplifier was added too.

Right click and select View Image for full view.

Compared to the 74HC240 PA version total efficiency is low, only 12.5%. Total consumption including DDS being 4Watts, not my idea of true QRP but once you throw that Xtal away you are on a slippery slope. I quake in fear at some of the temperatures that silicon gets run at so I added a 'Smoke Stack' to the output transistor heat sink. Here is another view.

To the right of the 'Smoke Stack' is the band specific LPF at the left hand side is the band specific input filter. Both modules are easily changed over, just 3 solder links on each for band swapping.
Here we see the DDS, mixer and PIC WSPR, Hell and QRSS modulator.

I have run on air tests on 24MHz and 28MHz, unfortunately with no reports as yet. Future band improvements should change that soon.

Three stations G8EUX, G6AVK and G3ZJO ran WSPR for more than 12Hours on 6m, 50.293MHz dial on 2 May 2009.
Apart from Spots from G8EUX of G3ZJO at 15Km there were no other de-codes. G3ZJO was RX only, the spots of G8EUX by G3ZJO have been done many times before.
G3ZJO had the advantage of continuous beaconing, multi modes, 5ele Rotatable Beam and the grabber at G6AVK activated.

Previous tests using 100mW of Slowfeld produced very useful copy at G6AVK of G3ZJO both from the direct beam heading and indirectly by beaming across the North South Air Corridor. The same beam headings were used for the 2 May tests.

G3ZJO also ran a 3 band co-linear antenna for a short time this produced the occasional burst of Aircraft Scatter at G6AVK. Peter, G8EUX produced a little WSPR and me, G3ZJO produced a short bit of QRSS.

G3ZJO could produce constant signals on the Grabber at G6AVK with 100mW to the 5ele Beam,(ERP nearer to 1Watt, I will quote RF out at the TX across 50 Ohms)constant but rarely usable due to multipath.

The most copyable signal that I observed in the test period was the Hell Ident in the following capture. 5 Unit Hell G 3 Z J swoops downward with Doppler shift from an Aircraft reflection and the O is back 'on the line'.

Several versions of QRSS Morse and WSPR proved useless under these conditions, 5 Unit Hellschriber produced some results. Previous activity on 6m by many stations has proved to be disappointing due to multipath signals. The results using Slowfeld which was produced to exploit Aircraft Scatter provided superb communications.

Following my recent Portable Trip when I found RF was causing the Touch Pad to Freeze. I have got round to curing the problem and giving the /P set up a test. To save writing it up twice this is what I posted on the Samsung user forum.

My Sammy has behaved faultlessly at home, used to provide coding and decoding of Digital Signals in my Amateur Radio station even when running quite high power levels.

On a recent trip to the coast I ran Sammy, Transceiver and all Domestics, (TV, Broadcast Radio, CD player Lighting, Alarm and Water Pump) on my 50 Watt Solar Panel for 4 days and Sammy came home fully charged by the Sun.

One problem showed itself on Sammy, due to proximity to the Transmitting Antenna it was impossible to run more than a few hundred milliWatts. The Touch Pad froze in the RF Field. In the absence of any other mouse it was impossible to lower the RF level once activated as the operations needs a working Touch Pad. Although most frustrating, the low power signals I could run actually made it to China, something I have not done from home :-)

The installation used the same leads, Audio In/Out and USB to Serial, both fully transformer and opto Isolated as used at home. It was difficult to run any tests whilst away from home although coiling both leads close to Sammy allowed a little more RF power to be used.

As soon as I good get round to it I made up a shorter Audio lead and wound around 10 turns onto a Maplin QT26D Ferrite Ring. Likewise the USB Lead was passed through another QT26D with about 10 turns in all. The rings are as close as possible to the plugs that go into Sammy.
On test with the same Antenna and Portable installation on the drive at home the Touch Pad behaved impeccably at high power levels and at 10MHz and 50MHz.

I have not seen this problem on any Laptop Computers before. Due to the very low RF field involved it could explain any touch Pad freeze ups when Sammies are used in certain locations and with external leads.

For several years I have had an appeal on my website for a capable Ham to try Slowfeld with me.
I see from archive material that in 2007 I was appealing on the Packet Radio Network and making test transmissions anyway, just in case.
At around the same time some QRSS Knights were doing it, I was not aware of the Knights.

My previous posts show the hopeless signals on QRSS and WSPR over the 50MHz path between here and G6AVK.
Quite typical of WSPR on 6m the signals are there but the scatter is so great that decodes are rare. My thoughts turned back to Slowfeld.

What you need is that capable Ham at the other end of the path, none better than G6AVK, I put forward the idea ready for his return from a hard day at work (with overtime). While his dinner cooked the software was installed and the tests commenced.

My first transmission produced readable Hellschriber at his end.

Next I took a TX session from him.

Finally I watched my signal on the G6AVK Grabber and turned the beam so that only signals scattered from aircraft were arriving, the TX power was only mWatts.

The results are astounding when QRSS3 FSCW fails and also WSPR then good old Hell comes to the fore.

Many thanks are due to G3PTT for the software which I have tried for several years to put to use. A superb mix of modern computer methods of signal analysis and an ancient visual data mode.

I built a new module for the DDS MEPT.

This time using a 27MHz Xtal, a triple tuned filter and conventional Class A amplifiers the output is 100mW.
On its first on air test signals were received by G6AVK at 179Km. The path crosses the London Luton, London Stanstead and London Heathrow air corridors so reflections were many and sometimes quite bizarre.

The use of Class A amplifiers puts up the current consumption degrading its status as true QRP, maybe I can brew tea with the wasted power.

A recent trip to Milford on Sea found a couple of unexpected problems with the Sammy Netbook when he is out and about. RF managed to get into the touchpad somehow and it went silly with very little RF power about. This problem never showed up at home, I am hoping that shorter audio leads maybe with a ferrite filter will sort that one out.

All in all it was a successful trial. The complete station including Sammy ran on Solar Power. The Wind was not really significant in the provision of Current.

Running less than 1 Watt of WSPR on 10MHz, (I couldn't run any more power due to the RF problem), I was pleased to be copied in China.

Hiding round the side of the 'van proved to avoid a little of the RF getting at the Sammy.

I like the KISS approach to QRP and much prefer a crystal oscillator based rig. The possible need for a go anywhere MEPT when it comes to future activity on 28MHz spurred me to try a modular system using my si570 DDS for my Softrock SDR.

ABOVE The Softrock si570 DDS Module.

It is not possible to FSK the Softrock si570 oscillator without a computer connected, no good for me, the computer consumption will far outstrip that of the transmitter.
What it does do is generate any frequency up to 260MHz in 1Hz steps without the computer connected. The computer is used merely to set the frequency via USB it is then disconnected.

I am pretty bored with plain old QRSS FSK Morse anyway, so often it is impossible to identify ones own MEPT in the presence of QSB on a grabber despite it being on the correct frequency, no wonder reports do not come flooding in.
Far more identifiable is a simple unique ident or a whole load of different modes which no one else is duplicating and the addition of a burst of WSPR ident every so often.

ABOVE The Modular KISS DDS on 28.300MHz.

The solution to applying modulation is to modulate the Oscillator of a Mixer, I used the SA612. In this way the setting up of the deviation for accurate WSPR steps is done only once.

ABOVE The SA612 Mixer/Modulator Module.

Just one Mixer/Modulator Module is used to cover 1.8MHz to 144Mhz with a c. 9MHz IF Frequency. Any old Xtal frequency will do in the range as the Output Frequency is merely I.F. Frequency + DDS Frequency. For 1.8 and 3.5 MHz it is - DDS due to the lower frequency limits of the si570. 7.0 and 10 MHz would benefit from a 3-4 MHz XTal, VHF could use 33MHz the Crystal is plug in and so fully flexible.
I have tested the Mixer over the full 1.8MHz to 144MHz range, only the 28MHz HiQ Filter/Preamp has been constructed as I write.

ABOVE The HiQ Filter/Preamp Module for 28MHz.

One HiQ Filter/Preamp Module will be required per band. The transistors in the Preamp stage are so cheap they may as well be duplicated for each band. Thus for VHF bands higher frequency devices can be used.

Regarding the Power Amplifier. A Broad Band PA could be used, this would require a BPF for each band and some form of switching. Just like the Preamp the devices are so cheap that including the PA device on the BPF Module is a reasonable choice thus avoiding any changeover other than plugging in the Module.

ABOVE PA/BPF Module 28MHz.

Spot the PA device, it is so small, a surface mount 74HC240 for 50mW Output into 50 Ohms, you can't tell it is there. It is the same arrangement that I have run on 10MHz 24/7 for a year.

ABOVE The precision timed PIC Modulator Module.

This Module produces 5 WSPR based frequency shifts c. 1.4Hz steps to produce any modulation mode that can be achieved with 5 units. I still like this system as the 5th unit is 'free' produced by a simple little trick I used. The De-Multiplexer Enable puts the 4 outputs into an all high state, another set level pot gives me another bit (unit) of information to play with.

ABOVE The Modules connected together on test on 28MHz.

This system allows me to experiment with any individual module and try new ideas for an on going MEPT project for all bands up to 2m. Modules will slide into a standard slotted plastic box for use as an MEPT for a particular band.
There are no plans for a posh enclosure, that would put a halt on its experimental status.
No Digital read out, that would use current that donates nothing to the signal. The 'compromise' of using a stand alone DDS has meant more than a doubling of the MEPT power consumption
No switches and knobs etc. Modules are connected together as needed. Now where can I get some nice little sockets to fit the square gold pins I use on the boards, I have run out.

I thought I would try some 2 meter WSPR. The Barometric pressure has been high for several days the other morning was quite misty, maybe a chance of some DX.

On VHF above 6 meters there are a couple of problems, any Rig may be easily be 500HZ off calibration and there still does not seem to be an established dial frequency for WSPR mode.

WSPR incidentally according to our wonderful national magazine Rad Comm which arrived today, 20th March 2009 is a 'brand new mode', 'rather difficult to get going' and 'not capable of any QSO mode communication'. I know it takes time for articles to get published but I am sure glad I don't rely on our National Society to keep me abreast of the latest in Ham Radio.

Anyway, back to frequencies. I carefully calibrated 2 rigs against GB3VHF and G8EUX calibrated his with a 'posh' bit of test gear. We were then both pleased when our RX'd signals reported 2 to 5Hz difference.
The French stations tend to use a dial frequency of 144.48800, a table on the web says 144.48850 after a short period of operation the concensus of opinion settled on 144.48860, there are other frequencies published too.

We had better get these frequencies settled if we don't there will an awful lot of wasted time and effort.

We still need to fix and confirm dial frequencies for 23cm and 70cm.
At this time the frequencies in use are:-

Band Dial freq USB (MHz) Tx freq (MHz)

160m 1.836600 .......... 1.838000 - 1.838200

80m 3.592600 ........... 3.594000 - 3.594200

60m 5.287200 ........... 5.288600 - 5.288800

40m 7.038600 ........... 7.040000 - 7.040200

30m 10.138700 ......... 10.140100 - 10.140300

20m 14.095600 ......... 14.097000 - 14.097200

17m 18.104600 ......... 18.106000 - 18.106200

15m 21.094600 ......... 21.096000 - 21.096200

12m 24.924600 ......... 24.926000 - 24.926200

10m 28.124600 ......... 28.126000 - 28.126200

6m 50.293000 ......... 50.294400 - 50.294600

2m 144.48860 ........ 144.490000 - 144.491000

The Zurich 30 Amp Regulated Power Supply was marketed under several other names, including MFJ. I have had mine for some 12 or 15 years now I recon. It is a 'real' PSU with a chunky transformer.

When it started to play up I was on air, I saw 10 Amp or more pulses of current indicated on the Ammeter. I was operating on PSK31 with a friend. When I mentioned it he said 'does that not worry you it could damage your rig'.
No, no worries here, this was the practical demonstration of one more reason for the system of 12 Volt supply I use in the shack. The PSU float charges two 40 Amp Hour Lead Acid Batteries individually via Schottky Diode isolators. A fault like this did not apply 15 to 17 Volts to my expensive radios, it might just do the batteries a little good by giving then a few de-sulphating pulses.

'Surely' he said 'your power supply is protected against Over Voltage'.
Yes it is, and Over Current, but this fault was obviosly not related to over voltage, the voltage involved must be a 'legal' one. Any voltage between 0 and 15 is 'legal' it being a variable voltage supply with a front panel adjustment pot.
This should have been a good clue to the cause of the problem, I didnt find it instantly, adjusting the voltage set pot and wriggling it made no difference to the pulses of extra current so I ruled it out at first. Finally I substituted a large preset pot and the problem was cured. I suspect that the connection of the earthy end of the potentiometer track was intermittent possibly under the riveted tag.
I ordered three 10K Ohm pots from different sources and one turned out to be very similar physically to the original.

The original has an indent at the center 13.8 Volt point, the replacement does not. No problem I don't use that point anyway, remember the Shottky Diodes, their foward volt drop means I set the output above 13.80. The original also has a splined shaft the new one is plain so I provided a different ponter knob.

Whilst inside I soldered the wire wrap connections on the control PCB.

Also the wire wrap connections to the load share resistors, these connections do get warm, I am much happier to see some solder on these after several years of use and possible corrosion of contact.
The Regulator Transistor connections on these boards had also had many hot /cold cycles so the joints were all re soldered.

Finally back in use on the rack, Battery Terminal Voltages set to 13.8 Volts.