So now we have it for certain. The reflected signal from an Aircraft will always fall in frequency due to the fact that the reflected signal path decreases in Length as it approaches the TX RX path and increases as it goes away.

The previous Captures have all proved this to be true.

This Capture of G4FCC signal at my QTH on 6m is the exception. What is the Aircraft doing in this instance at the start of the trace?
The clue could lie in the fact that Sywell Airport lies to the north of our signal path and within a few Kilometres of my QTH. The signal is short lived so could well be reflected from an Light Aircraft manoeuvring to land.

G4FFC lies due East of me. The main UK Air Corridor cuts our path approximately midway. He is there on 6m so an ideal chance to experiment.
Right Click on the image and select View Image for full view. Click Go Back to return.

In this Capture a single aircraft crosses our path, WSPR decoded.
Right Click on the image and select View Image for full view. Click Go Back to return.

Early in this transmission a plane crossed the path followed by another which tracked our path closely for a long period before finally crossing and leaving. WSPR again managed to decode.
Right Click on the image and select View Image for full view. Click Go Back to return.

By beaming along the Corridor I tried to find an aircraft this produced a 60Hz shift caused by an approaching plane. WSPR decoded.
Right Click on the image and select View Image for full view. Click Go Back to return.

Tests will continue as time permits.

aircraft scatter,radar,six meters,6m,wspr,reflections,air corridor

With the help of Colin G6VAK’s Grabber I was able to see my WSPR signal on 6m arriving at his QTH, strong enough, but not decoding. I could see my signal was accompanied by Lines descending in frequency and modulated by WSPR. I consider the WSPR decoder is confused by these.

I assumed them to be Aircraft reflections and set out today to duplicate the effect over a shorter range on 70cm with the help of G8EUX.
At my QTH I am well provided with high altitude Aircraft on the North South UK Air Corridor and also the Daventry DTY Navigation Beacon Chain.

G8EUX QTH is 14 Km to the South and Transmitted a slightly drifting WSPR signal which barely showed on the S meter. Under these conditions any reflections were rare but soon we were treated to a steep strong trace, modulated by WSPR which apparently crossed our signal path and went through flutter and zero beat. The steepness meant less time within the Bandwidth of the WSPR signal and it decoded OK.

At first the straight lines always descending in frequency and crossing the transmission frequency may cause confusion. The explanation lies in the fact that we have two stations and one moving Aircraft. As the Aircraft approaches the signal path the distance decreases. After crossing the path the distance increases. The same holds true from which ever way the the Aircraft approaches. The offset depends on the rate of increase or decrease.

The capture shows the 70cm signal with reflection. The map shows the Track of the DTY Navigation Beacon chain to/from the Luton Control Area, the North South Air Corridor and the 70cm path between G8EUX and G3ZJO.

Right Click and select 'View Image' for full view.

G6AVK QTH lies 127 Km distant to my South East and the path is largely parallel to the North South Air Corridor. Consequently the trace of many Aircraft Reflections is less steep and spends time almost parallel to the wanted WSPR signal. Unlike the 70cm trace which carries only a few phase changes from the modulation, the 6m reflections were modulated with an obvious image of the Direct Path signal.
I hope shortly to find a capture for inclusion here.

G6AVK had it stored.

Right Click and select 'View Image' for full view.

The capture shows 2 transmission periods from me with a blank 2 minutes when G6AVK transmits. The second period is the one that has an Aircraft which is on an almost parallel track to the signal path. The reflection from it quickly arrives at the transmission frequency. At this time we can assume that the Aircraft is at the midway point between G3ZJo - G6AVK, flying either North - South or South - North. At this point Doppler Shift is Zero. The plane moves toward one station at the same speed as it moves away from the other, result Zero.

G4FFC at my QTH, some multi reflections here.

Right Click and select 'View Image' for full view.

The Radio Path crosses the UK Jet Motorway.

Right Click and select 'View Image' for full view.

In this one, a parallel tacking aircraft produces a parallel reflection. The decoder in WSPR is confused by this and produces no decode.

Right Click and select 'View Image' for full view.

Interestingly this is Early Radar at work. The first Radar tests used a Fixed Radio Transmitter, [Daventry](G8EUX) and Aircraft [Heyford Bomber] (Thanks BAA or who ever) and a Receiver Station [Watson Watt/Wilkins] (G3ZJO).

Where did this first historic test by Watson Watt take place? Why right below our 70cm Test Aircraft track and only a short distance West of G8EUX.

Right Click and select 'View Image' for full view.

I was going to hang up my soldering iron for the summer but with friends like ZL1BPU a boy like me will get no rest.

Why only 5 tones he insisted, well I took a short cut I used another bit to enable/disable the original 2 bits of WSPR, when the 4 Units were high impedance I set another tone frequency. I then squeezed S/MT Hellschriber into 5 Units, clever aye.

I know you are trying to keep the bandwidth down but with 3 bits you can have 8 tones said Murray. Yes but, no but, well I may do something in the future.

Then I spotted 8 Unit Multiplexers at £0.17 each, well just one more project before the summer.

An 8 unit De-Multiplexer was produced.

I re timed the Software for a 4MHz Crystal for the Microprocessor Chip. The original 3.648MHz Crystal was a cheap one and not too thermally stable. Added to which the the frequency was not a direct multiple of the data timing. The 4MHz unit has proven to be very stable needing very rare compensation for thermal drift.

I included the reset push button for manual timing. However the Grubby Finger has been superseded by technology.

We know how to use GPS time signals, we know how to use MSF/DFS time signal Modules to set lock frequency and time. Despite the new stability I decided that a 20 or 40 minute MSF time re-set would be good.

Using this the shack temperatures can rise and fall without any time sync problems.
I decided upon a KISS and Junk Box solution.

There is one segment that is used only once every 20 and 40 Minutes (twice an Hour) on an LCD MSF Clock. I shine a Bright Blue LED at the Clock Face and detect whether this segment is Black or Reflective. Thus with just 2 chips from the Junk Box, an LM392 Comparator and LM555 used as a one Shot I am provided with a precision pulse.
What is more it works reliably in Daylight, Artificial Light and Darkness.

This photo does not show the segment used.

ZL1BPU extols the virtues of Castle Morse, I described a version which re uses the steps and thus saves bandwidth. Step Morse is now in its second simpler version. It uses a Center Line Dashes are Above the Line Dots are Below. Steps are re used in that they go up 2 and then down again this can cary on above or below the line ad infinitum. Thus all Morse Characters can be catered for using 5 tones.

Does it work? Pierre ON5SL copied my Call Sign and said he found it easy. This is his Grab of my signal.

What about S/MT Hell using the new MEPT. It certainly has a less squashed appearance.

Signals from me and G6AVK into Italy.

The Netherlands PA1SBD. Right click and select View Image for full size.

Finally for now again with G6AVK into Nova Scotia VE1VDM all on 40mWatts of RF. Naturally G6AVK is stronger with his 200mW.

This is all on test and will need housing, one of these days.

My original system for a Multi Mode Beacon with WSPR is achieved in the simplest manner possible i.e.. the addition of just one more tone to MEPT_JT coding.

This give the possibility of S/MT Hellschreiber in a very narrow bandwidth. The Hell has a squashed appearance for sure and could be improved.

QRSS3 Morse takes some time to send, Dashes being 3 times longer than a Dot, have 3 times longer to be effected by QSB, broken up and their information wasted.

There is another system, using equal length characters for both, Castle Mode, this can require up to 12 Tones. Hence I gave thought to perhaps re using the tones in the same character. My new system, Step Mode does this. Using luck my call sign can be represented with 5 tones. I have implemented this. There are plans afoot for a full Morse Character system using 7 tones.

Fine tweaking of the software timing has reduced any long term timing drift to well within MEPT_JT parameters. In reality its stability outstrips my shack computer when it is free running, (no time sync from the Web).

I have changed the name of my coding to avoid confusion. Whilst I am happy to have bits of binary code and bits of chewing gum in the same sentence, I can see that it could cause problems. Equally the system is 5 tone, I have gone for RTTY terminology of units.

It is now 5 Unit Multiomde + WSPR

All of the coding is based on MEPT_JT elements 682.66mS long. A QRSS3 Dot in theory 3 Seconds long is in fact 4*0.68266 Seconds = 2.73064 Seconds a QRSS3 Dash, 9 seconds in theory is 13*0.68266 = 8.87458 Seconds.

S/MT Hell Characters are produced by elements of MEPT_JT code. A 5th element is produce by an extra 3rd Binary bit.

All spaces are multiples of 0.68266 seconds apart from one filler, for accurate 2 Minute slot timing, just prior to the start of the MEPT_JT data and after the 2 step waveforms. The Steps are produced by the 5 element tones spaced precisely 1.46484375Hz apart and 4 elements long. Originally I had a 5-4-3-2-1 countdown with 5 of these. Now just 2-1, as they convey no real information. The top of step 1 is twice the length of step 2.

My WSPR Data starts at 1.36532 Seconds past the even minute, 2*682.66mS, as near to 2 seconds as the elements allow, although the 2 Minute filler can place this where ever it likes.

Paolo IZ1KXQ produces some fine screen shots of my signal.

Don't worry, be happy.

Precision timing is in the eye of the beholder. This is a MANNED Experimental Propagation Transmitter after all.

The dirty finger is deliberate,it was actually dirtier in real life.

My 40mW 5unit Home Built Computerless Transceiverless Multimode Solar Powered MEPT crossed the pond Last night.

Signals into Italy were good too.

I still have critical adjustments to make to the PIC Microprocessor Oscillator to get the timing precise. Over night the timing does get out of synch. It does not help if you make an adjustment in the wrong direction like I did last night just before retiring.

Oh well start again.

Now a report from Denmark OZ1PIF and we are on the WSPR Spots site and the map.

Just wait until I get this thing set up precisely.

We have results despite still being 'work in progress' Paolo IZ1KXQ has sent a report.

The complete MEPT beacon including PIC processor takes just 50mA at 12 Volts.
This is the PIC controller section.

The VCXO.

And the Micro PA stage, my smallest yet, it runs 40mW RF Output.

FRONT

REAR

MEPT_JT set the world on fire. MEPT was a term little known among Radio Amateurs, a little group of enthusiasts world wide, we built our own gear and studied propagation using very low power signals (QRPP).

MEPT_JT soon became WSPR, Weak Signal Reporting System, Whisper, a great name which encompasses the concept and keeps it in the WSJT camp.

A GUI version was soon released making it easier to use and soon a Web Site was set up to list reports and plot results on a world map. This overloaded the servers and the ISP closed it down. No problem, it moved almost seamlessly and goes from strength to strength.

Right from the first day I thought it would be good to have MEPT_JT as an ident on a true Multimode Beacon. A few Hams world wide are working on a computer less system, I needed a further spark from Johan SM6LKM who said yes it can be done.

Yesterday I completed the Multimode concept with a PIC processor controller and a QRPP VCXO transmitter.

Still work in progress I have introduced a 5th bit giving a Manual Time Sync signal and Multimode capability. I have produced a Helschreiber 'font', QRSS3 is programmed in and WSPR is de-coding locally.

Joe Taylor and Murry Greenman have been working on a new mode MEPT_JT. Just released this software produces and receives a 4 bit data stream with a total bandwidth of just over 6Hz.

I got it up and running yesterday morning on 10Mhz, output 100mW to a 6m CB whip with Radials and ATU Output power 100mW.
Here are the reports all in daylight.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

UTC Sync dB DT Freq Message Noise
---------------------------------------------------------------
140200 3 -24 1.8 10.140079 G3ZJO IO92 20 15.1

73,

Johan on5ex

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

C:\k1jt\mept>mept_jt Rx 10.1386

UTC Sync dB DT Freq Message
--------------------------------------------------
232000 4 -27 0.5 10.140172 K7EK CN87 37
232600 4 -26 0.7 10.140172 K7EK CN87 37
233000 1 -28 0.4 10.140173
233400 2 -26 0.9 10.140173 K7EK CN87 37
233600 2 -27 1.3 10.140172 K7EK CN87 37
233800 5 -26 1.7 10.140172 K7EK CN87 37
234000 2 -26 1.1 10.140172 K7EK CN87 37
234200 1 -27 1.5 10.140172 K7EK CN87 37
234400 1 -26 0.9 10.140173 K7EK CN87 37
234600 7 -25 1.3 10.140173 K7EK CN87 37
234800 2 -23 1.8 10.140172 K7EK CN87 37
235000 4 -23 1.1 10.140172 K7EK CN87 37
235200 7 -25 1.5 10.140172 K7EK CN87 37
235400 3 -20 1.0 10.140172 K7EK CN87 37
235600 8 -21 1.4 10.140172 K7EK CN87 37
235800 4 -21 1.7 10.140173 K7EK CN87 37
000000 8 -18 1.1 10.140173 K7EK CN87 37
000200 8 -18 1.5 10.140173 K7EK CN87 37
000600 3 -19 1.4 10.140173 K7EK CN87 37
000800 6 -20 1.8 10.140173 K7EK CN87 37
001000 1 -23 1.1 10.140172 K7EK CN87 37
001200 3 -22 1.6 10.140172 K7EK CN87 37
001400 4 -25 1.0 10.140173 K7EK CN87 37
001600 3 -22 1.4 10.140173 K7EK CN87 37
001800 3 -22 0.8 10.140173 K7EK CN87 37
002000 4 -22 1.2 10.140173 K7EK CN87 37
002200 3 -21 1.7 10.140173 K7EK CN87 37
002400 7 -18 1.1 10.140172 K7EK CN87 37
002600 14 -14 1.4 10.140172 K7EK CN87 37
002800 2 -20 0.8 10.140172 K7EK CN87 37
003000 2 -20 1.3 10.140173 K7EK CN87 37
003200 6 -20 1.7 10.140173 K7EK CN87 37
003400 1 -22 1.1 10.140173 K7EK CN87 37
003600 2 -24 1.5 10.140173 K7EK CN87 37
003800 1 -22 0.9 10.140173 K7EK CN87 37
004000 8 -22 1.3 10.140173 K7EK CN87 37
004200 9 -18 1.7 10.140173 K7EK CN87 37
004400 9 -18 1.1 10.140173 K7EK CN87 37
004600 12 -19 1.5 10.140173 K7EK CN87 37
004800 3 -17 0.9 10.140173 K7EK CN87 37
005000 8 -15 1.4 10.140173 K7EK CN87 37
005200 5 -12 1.8 10.140173 K7EK CN87 37
005400 11 -13 1.1 10.140173 K7EK CN87 37
005600 4 -17 1.6 10.140173 K7EK CN87 37
005800 5 -17 1.0 10.140172 K7EK CN87 37
010000 3 -16 1.4 10.140172 K7EK CN87 37
010200 7 -13 1.8 10.140173 K7EK CN87 37
011000 5 -20 0.4 10.140173 K7EK CN87 37
011200 4 -21 -0.1 10.140173 K7EK CN87 37
011400 6 -20 0.3 10.140173 K7EK CN87 37
011600 3 -26 1.4 10.140172
011800 1 -33 0.0 10.140144
011800 8 -32 0.4 10.140173
044800 2 -26 1.0 10.140130 ZL1BPU RF72 30
045000 3 -26 0.4 10.140130 ZL1BPU RF72 30
045200 1 -25 0.8 10.140130 ZL1BPU RF72 30
045400 1 -25 1.1 10.140130 ZL1BPU RF72 30
050000 1 -25 1.4 10.140131 ZL1BPU RF72 30
050800 2 -23 1.0 10.140131 ZL1BPU RF72 30
051000 7 -25 1.4 10.140130 ZL1BPU RF72 30
051200 2 -24 0.7 10.140130 ZL1BPU RF72 30
051400 5 -25 1.1 10.140130
051600 2 -26 0.1 10.140130 ZL1BPU RF72 30
051800 2 -27 0.4 10.140130 ZL1BPU RF72 30
052000 1 -24 0.8 10.140130 ZL1BPU RF72 30
052200 1 -23 0.3 10.140130 ZL1BPU RF72 30
052400 6 -23 0.7 10.140130 ZL1BPU RF72 30
052600 4 -22 0.1 10.140130 ZL1BPU RF72 30
052800 5 -21 0.5 10.140130 ZL1BPU RF72 30
053000 5 -21 0.8 10.140130 ZL1BPU RF72 30
053200 1 -19 0.3 10.140131 ZL1BPU RF72 30
053400 2 -19 0.7 10.140131 ZL1BPU RF72 30
053600 5 -20 0.1 10.140131 ZL1BPU RF72 30
053800 6 -18 0.4 10.140131 ZL1BPU RF72 30
054000 1 -17 0.8 10.140131 ZL1BPU RF72 30
054200 7 -18 0.3 10.140131 ZL1BPU RF72 30
054400 3 -16 0.7 10.140131 ZL1BPU RF72 30
054600 9 -20 0.1 10.140130 ZL1BPU RF72 30
054800 2 -20 0.4 10.140130 ZL1BPU RF72 30
055000 6 -21 0.8 10.140131 ZL1BPU RF72 30
055200 2 -20 0.3 10.140131 ZL1BPU RF72 30
055400 4 -22 0.7 10.140131 ZL1BPU RF72 30
055600 2 -20 1.1 10.140131 ZL1BPU RF72 30
055800 3 -22 0.4 10.140131 ZL1BPU RF72 30
060000 6 -23 0.8 10.140131 ZL1BPU RF72 30
060200 2 -19 0.3 10.140131 ZL1BPU RF72 30
060400 9 -20 0.7 10.140130 ZL1BPU RF72 30
060600 4 -21 1.1 10.140130 ZL1BPU RF72 30
060800 4 -21 0.4 10.140131 ZL1BPU RF72 30
061200 1 -23 0.3 10.140131 ZL1BPU RF72 30
061400 2 -24 0.7 10.140131 ZL1BPU RF72 30
061600 1 -23 1.1 10.140131 ZL1BPU RF72 30
061800 1 -24 0.4 10.140131 ZL1BPU RF72 30
062000 3 -25 0.8 10.140131 ZL1BPU RF72 30
062200 2 -28 0.3 10.140131
065200 1 -23 0.9 10.140130 ZL1BPU RF72 30
065400 2 -24 0.3 10.140130 ZL1BPU RF72 30
065800 1 -26 0.1 10.140130 ZL1BPU RF72 30
093000 1 -29 1.0 10.140132
114600 2 -27 1.1 10.140132 ZL1BPU RF72 30
114800 4 -25 0.4 10.140132 ZL1BPU RF72 30
115000 1 -23 0.8 10.140132 ZL1BPU RF72 30
115200 1 -25 1.2 10.140132 ZL1BPU RF72 30
132800 3 -20 0.1 10.140173 K7EK CN87 37
133000 2 -25 0.4 10.140172 K7EK CN87 37
133200 8 -22 0.8 10.140172 K7EK CN87 37
133400 1 -26 0.3 10.140172 K7EK CN87 37
133600 3 -26 0.7 10.140172 K7EK CN87 37
133800 5 -24 0.8 10.140172
134000 8 -22 0.7 10.140172 K7EK CN87 37
134200 8 -24 1.1 10.140172 K7EK CN87 37
134400 1 -21 0.6 10.140172 K7EK CN87 37
134600 6 -22 1.0 10.140172 K7EK CN87 37
134800 9 -18 0.4 10.140172 K7EK CN87 37
135000 8 -20 0.8 10.140172 K7EK CN87 37
135200 3 -19 1.3 10.140172 K7EK CN87 37
135400 5 -19 0.7 10.140172 K7EK CN87 37
135600 14 -17 1.1 10.140172 K7EK CN87 37
140000 3 -23 0.9 10.140172 K7EK CN87 37
140200 1 -22 0.3 10.140172 K7EK CN87 37
140400 4 -18 2.0 10.140091 G3ZJO IO92 20
140400 7 -19 0.7 10.140172 K7EK CN87 37
140600 12 -14 1.1 10.140172 K7EK CN87 37
140800 4 -20 0.5 10.140172 K7EK CN87 37
141000 2 -18 1.0 10.140172 K7EK CN87 37
141200 8 -17 0.4 10.140172 K7EK CN87 37
141400 2 -22 2.1 10.140090 G3ZJO IO92 20
141400 6 -19 0.8 10.140172 K7EK CN87 37
141600 1 -25 1.8 10.140090 G3ZJO IO92 20
141600 4 -21 1.2 10.140172 K7EK CN87 37
141800 8 -23 1.4 10.140090 G3ZJO IO92 20
141800 1 -24 0.6 10.140172 K7EK CN87 37
142000 1 -23 2.1 10.140091 G3ZJO IO92 20
142200 1 -27 1.7 10.140090
142200 5 -23 0.4 10.140172
142400 4 -25 1.4 10.140090 G3ZJO IO92 20
142600 5 -25 1.3 10.140172 K7EK CN87 37
143000 1 -28 1.1 10.140172 K7EK CN87 37
143200 5 -28 0.4 10.140172
143400 1 -27 1.4 10.140172 K7EK CN87 37
143800 3 -25 0.3 10.140172 K7EK CN87 37
144000 2 -27 0.7 10.140172
144200 1 -26 0.7 10.140172 K7EK CN87 37
144400 4 -22 1.1 10.140172 K7EK CN87 37
144600 1 -25 1.5 10.140172 K7EK CN87 37
145200 1 -28 2.5 10.140090
145800 2 -28 2.1 10.140091
150000 1 -19 2.2 10.140091 G3ZJO IO92 20
150200 4 -19 1.8 10.140091 G3ZJO IO92 20
150400 3 -19 2.5 10.140091 G3ZJO IO92 20
150600 6 -18 2.1 10.140090 G3ZJO IO92 20
150800 2 -24 2.7 10.140090 G3ZJO IO92 20
151000 3 -23 2.4 10.140091 G3ZJO IO92 20
151200 3 -22 2.0 10.140091 G3ZJO IO92 20
151400 1 -23 2.5 10.140091 G3ZJO IO92 20
151400 1 -28 1.3 10.140172 K7EK CN87 37
151800 1 -29 1.1 10.140172
152200 1 -28 0.6 10.140172
152800 2 -25 1.4 10.140173 K7EK CN87 37
153000 2 -24 1.8 10.140173 K7EK CN87 37
153200 1 -27 1.2 10.140172
153400 1 -27 1.1 10.140172 K7EK CN87 37
153600 2 -25 0.5 10.140172 K7EK CN87 37
154000 1 -26 1.4 10.140172 K7EK CN87 37
154800 1 -32 1.1 10.140172
160200 2 -24 0.6 10.140172
160400 2 -24 2.5 10.140100
160600 2 -25 1.7 10.140099 G3ZJO IO92 20
160600 6 -25 0.4 10.140172 K7EK CN87 37
161200 3 -27 0.7 10.140172 K7EK CN87 37
161600 1 -25 0.5 10.140172 K7EK CN87 37
161800 1 -24 0.9 10.140173 K7EK CN87 37
162000 3 -23 0.4 10.140172
162200 4 -24 1.3 10.140172 K7EK CN87 37
162400 1 -26 3.1 10.140092 G3ZJO IO92 20
162400 2 -25 1.7 10.140172 K7EK CN87 37
162600 3 -22 1.1 10.140172 K7EK CN87 37
162800 2 -27 2.4 10.140091
162800 2 -25 1.5 10.140172 K7EK CN87 37
163000 2 -26 2.7 10.140091
163000 4 -25 0.5 10.140172 K7EK CN87 37
163200 1 -25 2.9 10.140091
163400 2 -26 2.1 10.140091 G3ZJO IO92 20
163400 3 -27 1.4 10.140172 K7EK CN87 37
164000 1 -25 2.8 10.140091 G3ZJO IO92 20
164000 1 -25 0.4 10.140172 K7EK CN87 37
164200 3 -24 2.5 10.140091 G3ZJO IO92 20
164200 3 -24 0.8 10.140172 K7EK CN87 37
164400 1 -24 1.2 10.140172
164600 4 -25 3.2 10.140090
164600 3 -25 1.2 10.140172 K7EK CN87 37
165000 3 -26 1.8 10.140090 G3ZJO IO92 20
165000 3 -24 0.3 10.140172 K7EK CN87 37
165200 2 -26 0.7 10.140172 K7EK CN87 37
165400 1 -24 2.2 10.140090 G3ZJO IO92 20
165400 5 -24 1.1 10.140172 K7EK CN87 37
165600 6 -18 1.8 10.140090 G3ZJO IO92 20
165600 3 -24 0.5 10.140172 K7EK CN87 37
165800 2 -26 2.5 10.140090 G3ZJO IO92 20
170000 1 -22 2.2 10.140090 G3ZJO IO92 20
170200 3 -19 1.9 10.140090 G3ZJO IO92 20
171000 1 -27 0.4 10.140172 K7EK CN87 37
171400 3 -21 2.9 10.140090 G3ZJO IO92 20
171600 2 -21 2.5 10.140090 G3ZJO IO92 20
171800 2 -26 2.1 10.140090 G3ZJO IO92 20
171800 3 -27 1.1 10.140172 K7EK CN87 37
172400 2 -25 0.3 10.140172 K7EK CN87 37
172600 3 -27 0.7 10.140172 K7EK CN87 37
172800 6 -28 1.8 10.140172 K7EK CN87 37
173000 1 -29 1.2 10.140172 K7EK CN87 37
173200 1 -29 1.6 10.140172
173600 1 -30 2.8 10.140089
174000 4 -27 0.4 10.140172 K7EK CN87 37

-- 73, Joe, K1JT

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

C:\Program Files\WSJT6>mept_jt Rx 10.1386
UTC Sync dB DT Freq Message Noise
---------------------------------------------------------------
181400 8 -23 2.7 10.140093 G3ZJO IO92 20 18.8
181600 2 -24 3.8 10.140093 G3ZJO IO92 20 18.9
181800 1 -17 2.9 10.140093 G3ZJO IO92 20 19.2
182000 4 -18 3.0 10.140093 G3ZJO IO92 20 19.2
182200 4 -19 3.1 10.140093 G3ZJO IO92 20 18.9
182400 5 -23 3.2 10.140093 G3ZJO IO92 20 18.8

73 de LA5VNA Steinar

A nice mode aye?

My Diversity MEPT uses two oscillators on the same frequency each with Frequency Shift Keying 180 degrees out of phase. This results in a box appearance when receiving both signals, resembling a string of Paper Dolls (Origami).

The two oscillators are locked in frequency by exploitation of the cross coupling inherent in the construction of CMOS Gates.

The 'G3ZJO Lock' is described on my companion blog. Radioman

The reception reports of MEPT Beacons come in from around the world. The first thought that goes through the mind is, what would the signal have been like using a different antenna. The comparison between a Vertical or Horizontal antenna is particularly interesting in Propagation Study.

This Diversity MEPT is the result of a long train of thoughts on the ways to achieve identifiable dual polarisation from a transmitter.

I used some of the existing techniques I have used in other Beacons and circuits I have tried before. Building identical channels for each transmitter.

I had problems with the Red LED's as regularly used as Varicap diodes, When I switched off my Bench Lamp the frequency shifted intolerably. This gave the excuse to try the Black LED's which are not effected at all. Yes of course they are second hand, TV InfraRed remote devices, they work fine as varicaps.

Another free, second hand component is the PIC timing Crystal a 4.43Mhz TV item from the Philips Video 2000 Recorder. The ident is selected by a push switch, yes Video 2000 front panel switch.

The Mail on Sunday has a write up about the involvement of Radio Ham Robert Rowlands who 'stumbled' on a gang of bank robbers 'on the job'.

There was so much mystery tied up in this story that it has inspired a film The Bank Job.

Click on the link below to hear the recording made by Mr Rowlands.

http://mfile.akamai.com/11297/wmv/anm1.download.akamai.com/11297/video/2008/Feb08/Bank.asx?obj=1

Now we have the measurement method sorted I can confidently say that my new QRP MEPT is 26.3mWatts Output.

First a realisation. The other day I moaned about the Media elsewhere.
You know, how if you know anything about the subject being discussed it is often riddled with errors.
I have been aware of this for years so why did I get caught out by a 2004 Shortwave Magazine article about QRP power measurement.

The piece described an adapter, two SO239 sockets with 50 Ohm resistive load, diode and capacitor. It gave the formula for deriving Power from Voltage across 50 Ohms, this is the method used for any Practical Wireless QRP contests, it read.
There was also a chart of Voltage and Power from 250 mW up to 20 Watts.

I accepted the article without any further thought, why oh why did I cut that page out and put it in a folder for future reference. Bah #!#!

Only when I saw a power measurement from voltage by G0UPL with my figures fresh in my mind did I puzzle over how there was such a vast difference, not explained by the Oscilloscope Peak to Peak value from G0UPL and mine derived from Rectified Peak value.

That article gives correct values of RMS Power in the chart. Use the formula given and you end up with Peak Power. There is no mention of the published circuit reading Peak value and no mention of RMS.

For each MEPT I have built I used that SWM page for reference, for QRP tests the same, my 10mW is really 5mW by everyone else's standard.

G3ZJO you should have known better.

Once again we are testing JT2 and also JT65 on 80 meters when the propagation is really poor between UK and South West France a path of 720Km.

The picture is self explanatory. Right Click and select View Image for full picture.

At 25mW level on this test taking QSB into consideration results were comparable usng both modes. Perfect decodes.

Most reports I see regarding the experimental mode JT2 seem to mention the Sound Card in the computer being critical. My own tests with F6GUU have revolved around some strange problems which we have not come to any conclusions about.
One hundred percent perfect test sessions can be interspersed with fifty percent failure sessions at other times. Seemingly fine signals in the clear just do not decode. Lack of local stations able to help in the tests cause uncertainty as to whether problems lie at one end or the other or both.
Consistently, failed decode sessions at my end indicate 0 Sync for each RX window.

The mode still lacks some refinements as stated in a reply to me from Joe Taylor.

The JT2 decoder in WSJT 5.9.8 was cobbled together very quickly. I
wanted to get something working so that on-the-air tests could be made.
Many short-cuts were taken, and potentially important features left
out. (There is no AFC action, for example.) I am sure that on-the-air
performance can be improved considerably. I hope to be able to spend
some time on it again, in coming months.

We would just love to be able to identify positively the signals and conditions that will cause a decode fail. To this end I pushed the limits in some local tests.
Using un-measurably low power into a Dummy Load on TX and an Inefficient Antenna or a Dummy load on RX I tested weak and strong signal performance.

Test 1

Lower signal limits were reached when the RX noise level was set to -10dB 50 Ohm Load and a signal which read -11dB.

100% decodes every time.

Test 2

Next a correct setting of 0dB RX noise 50 Ohm load and a signal at -7dB.

100% reliable decodes.

Test 3

Next 0dB RX on an antenna -8dB signal giving the Icom IC706 some real signals causes lots of visible spurii.

100% reliable decodes.

Test 4

Badly adjusted RX level of +8dB and signal level of -5dB seemed to be about the upper limits. Just look at the mixer products and the PSK31 stations nearby, if you want to stress a Digi system use an IC706.

100% reliable decodes.

No signs of any critical sound card input levels here. Vast maladjustments still produced reliable results. The test reproduced a true on air path as near as possible.

Thanks to the Walrus and Southgate ARC for this picture.

Now look 'ere my man it's not my fault you've got a hole in your boat.

In an effort to strain JT2 Mode just a little more we have chosen to test on 80m when the path is predictably poor.
There was no other activity audible around the frequency other modes were failing.

I started the test with 10 Watts reducing to 5 Watts then 1 Watt the last 2 decodes are -12dB and -11dB are at the 1 watt level.

The above capture is of course the French end, e-mailed back in the UK within minutes.

First a copy of F6GUU with adjacent QRM.

Next a non de-code, the RTTY station has jumped clean on top of F6GUU.

From some previous tests on 40m under QRM conditions we were beginning to think that the AGC in the RX was dropping the RX gain to below the Decode Capability of JT2. So producing a limiting factor for the mode.
The tests were at a time when propagation between France and UK was failing, so giving extremely low comparative signal levels.

A night time test on 80m, 3.590 USB Dial, among strong RTTY contest stations has proved far more promising. A reliable contact was made using JT2 despite the complete failure of Chip64 with which we used 3dB more power at each end.

The use of the DSP on my Kenwood TS870, setting the RX bandwidth to just the JT2 window of 1.2KHz certainly reduced the effect of the strong QRM on the AGC.

These are my decodes of F6GUU.

Now F6GUU decodes of me.

There is very little activity on JT2 Mode to be found on the HF bands. With a bandwidth of only 8.75 Hz it is attractive for Low Signal work.

I have been testing with F6GUU in Mansle Charente. It is a similar mode to JT65M which we have used for Meteor Scatter contacts.
After some rapid success we had a period of problems from the F6GUU station.

Perseverance is the thing, we now have good decodes both ways and can continue with some meaningful tests on 7.034 MHz USB dial setting and 10.137 MHz USB on the dial.

First we have my signal as decoded by F6GUU.

I was using 5 Watts on 7.036 MHz to 1/2 wave Doublet Antenna.

Next F6GUU as received by me, he is also running 5 watts to an End Fed Wire Antenna.

Just a couple of Web Sites if you are interested in Vintage computers.

http://nascom.wetpaint.com

From the days when you soldered dozens of components onto the printed circuit board and loaded programs from audio cassette tape.

My Latest UK Radio Site with Links to other Radio Interest Sites.

http://www.g3zjo.talktalk.net

A dynamic site that keeps up to date with the experimental side of the hobby by blog linking.