The addition of PI4 to the U3 TX by Hans Summers suite of modes has lead to my testing it on 28MHz.
PI4 is not really a suitable mode for QRP but some users asked for it and Hans is always up for a challenge. Where those that asked for it are I don’t know, but myself and EA2KV are using it on 28.303MHz time shared on the same frequency.
I have now been running the mode for over a month and there has been one solicited report from Greece. I always disregard solicited reports on established modes from any results but this is different the frequency has to be publicised.
I have also been running other tests concurrently awaiting unsolicited reports. These are possible because the signals are in a window where hundreds of highly skilled expensively equipped Hams are monitoring worldwide. WSPR remains the mode for propagation tests. It knocked QRSS off its perch when it was introduced. I have been running a comparison test on 28MHz for over 5 years and the report numbers are in excess of 100,000 WSPR to 1 QRSS.
For the test to be fair I TX each mode from the same transmitter and for the same period of time. WSPR carries Callsign, Locator and Power in 2 minutes, DFCW5 takes slightly over 2 minutes to send my Callsign only.
I have recently added FSKCW7 to the list, in just under 2 minutes it manages just a ‘G’. That’s fine, many years ago I grabbed a ‘G’ as my international identifier on QRSS, to keep it legal I also ident with the full Callsign. That has been running for over a month with the PI4 with no reports despite the fact I often see it on Grabbers.
Now I have just gone and publicised it so reports no longer count 😀
Above. My little signal holding its own among signals into VE, some running 10 times more power. It took nearly 2 minutes to convey a not very readable ‘G’, in the same 10 minute time frame full WSPR decodes were made in the USA.
2014-12-07 14:48 G3ZJO 28.126190 -26 0 IO92ng +20 0.100 KB9AMG EN52tx
In order for QRSS to perform a useful service in the digital age it needs to be several times slower and the Grabbers need to have matching FFT times. The message time then gets to be so long that as a communication mode it is boring. However with an MEPT and Grabber set up it is ideal, after the first frame you can view and admire your own signal when ever you like.
Here are the photos of the U3 on test showing the really nice blue screen display and with my hand to reference the true size.
It really is a little beauty. Currently I am setting up the 20MHz clock oscillator and checking free running with no GPS, frequency stability. It is very good, I have no Xtal heat sink fitted and see no problems. The dirty finger nail is ‘an in joke’ here since I did a picture of poking a dirty finger on a switch and the photo didn’t do it justice.
Rear view showing the 600m LPF, those scruffy ‘almost impossible to wind’ inductors don’t look too bad.
I thought my LPF’s were small but the U2 / U3 ones (green) are tiny, the one on the right is the even more impossible to wind 2200m LPF, a little fatty, but it fits the board with a little overlap.
The U2 circuit board layout gave problems of interference to the GPS, my test lead for the U501 is shorter than the one I used on the U2 and I forgot to put ferrite beads on the wires at either end, so far I have seen no problems.
The Ultimate v2.02 chip was released and I had the chance to Beta test it.
Current release is v2.02a if you have a U2 with v2.01 please, please shell out the mere £3 and update ASAP. Other band users will love you for it.
Good news:-
Gone are the transients due the software zero setting the DDS before every small wanted frequency shift. This made the signal socially unacceptable, the Gentlemen operators of the past would have turned in their graves to have heard the clicks which sounded on air like someone banging a spade with a hammer.
Sorted is the linearity of the frequency shift which went awry as the band frequency got higher and caused WSPR to produce non decodeable signals. The linearity correction has also made Slow Hell Mode clean and readable.
Gone is the broad band interference which wiped out Radio 4FM on 93.3MHz when the U2 was TX’ing on 7MHz.
Gone are the bands of interference which wiped out 6m (50MHz) when the U2 was TX’ing on 28MHz.
These were the major bugs which kept me off the air with the U2.
WSPR-15 Mode has been added, also some ergonomic improvements in menu settings and cosmetic updates.
Results have been good on air, Power Output using just the one supplied Output FET at 474KHz is 250mW, I tried WSPR15 on that band and it works fine one report from 127Km was -40dB.
The Power drops as the frequency rises to give 109mW on 18MHz, at 28MHz efficiency seem to go back up a bit and 190mW is measured. I consider 28MHz to be the upper limit of operation.
All modes were found to work fine across the bands.
Bad News:-
The GPS facility which worked perfectly I though on v2.01 with my Trimble GPS Module totally fails on v2.02. Actually on re checking with v2.01 the time is always one second faster than UTC.
Using v2.02 the system time gallops on in 2 second leaps and soon is faster than real time. The 1PPS Frequecy control does not work.
Some GPS users had problems with some GPS Modules on V2.01 and some parameters were changed in an attempt to accommodate those. As far as I can tell only 2 Modules now work, which worked previously as well, and the Trimble lost out.
The GPS facility was the clinching factor on buying the kit for me and its a real let down to have it fail.
The utilisation in code of GPS Data is something I have never done, however I do note that the $GNRMC Position Data used in the U2 contains the UTC Time of the Position Fix. It is known that this time can be 1 or 2 seconds late. If one sets the system time from this Data then clocks that with 1PPS the next Data read may well be different, giving rise to erratic time keeping symptoms experienced by me and others.
It would appear that different GPS Modules have various accuracy levels and I would guess from my limited study that the better accuracy ones have a shorter PPS pulse time. No doubt that given a particular set of parameters a compromise can be achieved. I do hope a setting for diverse modules can be attained.
From my study I have concluded that the correct way to deal with GPS time is to read the $ZDA Data which is the correct UTC Time then clock that with the 1 PPS, when that Data is read again the two will agree and time galloping in 2 second leaps would be impossible.
The U2 generates noise on RX because when it is not TX’ing, the DDS is not disabled, it actually is working away on 6.25MHz doing calibration tasks, this signal is output to the PA and the PA power is still applied, Key Down being by gating the FET Source to Ground. Signal leaks out via the Drain Load to the Antenna socket.
This situation is disturbing to me, I have never had such a noisy sprog ridden shack, several times during testing I was caught off guard and confused by the spurious signals both on RX and TX.
There are some odd effects also reported by another Beta tester of pick up from the U2 via its own leads, I couldn’t improve this using decoupling, earthing changes or anything. This can manifest itself with 50Hz and 100Hz sidebands on WSPR when monitoring your own signal.
To be fair the crud, S2 on 28MHz, is not devastating but as a QRP RX station as well as TX I take great care to get as low a noise floor as possible. I complain bitterly about my neighbors Plasma TV’s the U2 competes from within.
Finally:-
The U2 has earned a place in the shack and Ultimately, pun intended, as a portable MEPT with GPS. Hence it also earned a box and it is off the bench.
It is interesting to see the Enclosures used by various assemblers of the kit. They have been of all types, ranging from large and roomy with a bit of spare room for a few sandwiches and a flask as a picnic portable to solid Die Cast Boxes which would fend off a WW2 Tank attack and some small, compact jobs too.
I went for small, compact, light weight, Hand Held with the 5Volt Regulator inside, I wanted to avoid having the LCD display fixed through the front panel and worked on avoiding having its fixing screws, which are in an odd 3 screw configuration showing. The Data Entry buttons can be quite awkward if not ergonomically arranged I mounted them for modern two thumbs use and it works out well.
The DDS has an LED indicator I didn’t add another panel LED with extra current consumption, I have a spy hole in the front panel which shows the DDS one, a light pipe will be a nice addition for that.
Band changing requires a swap of the plug in LPF. On the bench I found this quite awkward to get at, in the box the PCB is mounted so as to present the LPF at the top right hand corner, open the front and it and the PA transistors are looking at you, it is easy to hook a couple of grubby finger nails underneath to remove and change.
Input and Output, Power, GPS, and Antenna Sockets are on the rear.
Here is the U2 Boxed on Video sending WSPR on 28MHz.
A nice visual report arrived by e-mail yesterday from Ray W2XC, he was pleased to be seeing my 200mW 28MHz MEPT on WSPR regularly and I was pleased to see his screen shot.
Now this is not exceptional, what I do find odd is the lack of reports from QRSS stations and Grabbers of the same 200mW TX when it sends QRSS. The contrast is daily reports from America, Europe, Africa, Asia and Australasia on WSPR, zero reports for years on end from anywhere for QRSS.
OK so most activity is on 10MHz, but stations and Grabbers do go on to 28MHz regularly, QRSS needs to be strong enough to be visible but just look at the waterfall display of my signal on the screen shot, (2nd signal up in the second column from the right), on the scrotty little non optimised display which is no more than an indicator for WSPR stations.
PA0SBD sent me some very nice Argo Grabs today of my 28.322730 MHz FSK QRSS3 MEPT.
Peter said:-
Below “all” screenshots where you where visible this afternoon > 16h00 UTC until now, 19:00 UTC.
The antenna is a 30mb horizontal dipole @ 5 m from the ground. [some quick fading].
The distance between us is 533km.
The 2 minute gap in the signal below is not QSB it is the period when the TX QSY’d to 28.126073 MHz to send WSPR.
Since the switch on of the new Dual Mode MEPT The WSPR Spots have been romping in from around the EU, Portugal to the South Iceland to the North, Slovenia to the East. I like WSPR, stations who report have no choice and no operator skill or effort is required.
I still like the visual modes but the reports are much thinner on the ground. Later hopefully many Italian stations will be monitoring the 28.322MHz window for Beacons and no doubt others worldwide will take a look. For now it is the keen, skilled enthusiasts who are reliable.
Paolo IZ1KXQ has sent a daily report of my 300mW signal from Sarzana.
Signals, strength and QSB are very similar on both days, below 20th May.
Today 21st May the signal lasted a little longer before the QSB took my signal into the noise.
It takes just under 8 minutes to send my Call and Locator at QRSS3 so for full copy of all data that period of signal is required with no fading or interference. Using WSPR Call, Locator and Power are sent in 2 minutes. Extensive tests on 500kHz have indicated that despite the theory that QRSS would have the advantage at low signal strengths, WSPR has the edge.
For details of the 28.322Mhz Italian Beacons see Rete QRSS/QRPp a 28.322 MHz
Just in time for the improved conditions on 10 Meters I have put my thoughts into action. I have been a bad boy for long enough, running WSPR along with QRSS, thus causing multiple suicides worldwide among the Luddite Hams who can’t conceive radio as anything more than an up and down Morse key with a valve.
I have even been told that there is no room on the 500kHz band for my 6Hz wide signal on a band that should be used as a Morse Museum. Mausoleum more like, it is not the opinion of the licensing authority, thank goodness.
With just a little concession to less KISS, I have used a AD9834 DDS chip to do my frequency control. A modified bit of PIC programming calls for a QSY from the 28MHz QRP Beacon slot to TX the WSPR data in the Whisper window. The PIC, DDS Control, QRSS and WSPR Generation and Mixer to 28 MHz are all on the same board.
QSY imminent, the Red LED does not show up too well in the photo.
WSPR in progress, the Green LED is even less obvious. The pictures speak for themselves, no need for me to give the frequencies I am using.
One beauty of the new system particularly on 10 Meters is that I can put the Computer on and receive WSPR between my TX periods. Similarly I can monitor for QRP Beacons for 2 minutes in every 10 minute cycle.
The very first transmission, 350mW of RF brought a Spot from CT1JTQ. Followed by Italy, DL, OY, OZ and EA the conditions can only get better, we should be in for some fun.
It goes without saying that reports are very welcome particularly on the QRSS ident.
For details of the QRP Beacon slot mentioned see Rete QRSS/QRPp a 28.322 MHz This Italian group started with the use of readily available Crystals for the frequency.
A change of mode on 500kHz brought in some reports.
The Transmitter is the 4MHz Crystal divided by 8, modulated by a 16F628 PIC keyer. The PA is another Complementary Pair Class D, the transistors again ex Computer memory BFY50 and a CV No. PNP Silicon. I have replace the Inverter PA with this design, as used in my WSPR MEPT. Inverters being CMOS are limited to a 5Volt or so supply, I find them fine for HF where tens of milliwatts is used. On 500KHz you have to start to push them, the added current produces heat in the Voltage regulator but not much in the chip. This is totally wasteful, if I am going to burn my hands I want to do it on the RF device no the PSU.
I have built this TX with the Altoids tin in mind, each unit is on a board cut to fit the noble QRP enclosure. I had the Dispensers at the local chemists smiling the other day when I bought more mints and told them I only wanted the tin really, “they have a worldwide following you know”. “Yes sir, fine, keep taking the pills won’t you”.
It would be impossible to put the whole thing in 1 tin, so we have 3 modules, Keyer – RF Generator – PA. Ah! Happiness is a warm Altoids Tin.
The PA does not quite run as much power as the WSPR one, maybe it is the drive waveform, I have not worried about it as 500mW RF into 50 Ohms is just right.
The first session brought a nice report from G3XBM, interesting as we still have not managed to get a WSPR signal to him. There must be an answer to that somewhere.
Next G3CWB in IO91rm, BTW its downside up, read the bottom line.
Super signals at M0BMU, he is my standard by which I judge my performance, the first, and only for a while, station to receive my QRppppp when I first started with a dubious antenna.
G4GWT is at a fair distance in IO83qo 188.9 km or 117.4 mls. Captured at 1245utc on 24th January.
There may be more reports to come today, if so I will add them later.
A report was received from Mark GM-SWL, he sent some grabs which are pretty surely my signal although not enough information was clear for a positive report. Also after dusk Rik ON7YD copied my call several times.
Thanks to all reporters, unlike WSPR, you had a choice and put in some effort to spot the signals from IO92.
The ability to go anywhere with an MEPT is increasingly more desirable, unless you want to sit Rock Bound on 30m or use DDS.
Many of us have a Multiband TX/RX which can run QRP but for some modes, (WSPR), a computer is needed in TX only mode which makes a mockery of the term QRP as applied to Station power consumption.
We have produced PIC based MEPTs for Multi Modes and WSPR.
Some have recorded WSPR audio on MP3 and re played that for Computer-less operation.
I think Computerless is a well understood term in all walks of life. Well this uses a lot less Computer than most. This uses a small PIC processor with no external memory, sound card, hard drive, video card, monitor, keyboard or mouse and I swear it has never had relations with Microsoft, or Monica Lewinsky.
For the purists I shall say PC Less in future, if I remember.
I have now produced the KISS WSPR Organ. Sorry Whisper refuseniks for the name, it does QRSS just fine.
The name stems from my original idea of using a musical tone generator to produce the WSPR tone frequencies. Talking to an Organ engineer yesterday morning soon proved that approach to be far too complex and my simpler KISS idea was implemented.
The Output can be applied to any HF/VHF SSB Transceiver or better, a simple home brew SSB TX.
The PIC will provide QRSS Morse, Hell, Patterns, WSPR whatever, plus PTT TX/RX switching for any pre determined periods
There is an advantage over the generate on frequency MEPT in that nothing is running on signal frequency so the generator can be left running during receive periods.
Two Xtals on the same frequency use a single 74HC86 for Oscillator, Mixer and Buffer. One Xtal is pulled 1.5KHz HF and the PIC modulates this Oscillator also.The modulated 1.5KHz signal from the Mixer is selected, fed to the Buffer and feeds the TX Data Audio Input via a Set Level pot.
Just one chip and thats it. And it works, this from G6AVK at 127Km on 50MHz.
And on WSPR it is performing well no drifts in some 10Hours use, again thanks to G6AVK for the reports. The severe Aircraft Scatter has caused a bad decode but we are used to that on the direct path.
Sorry the power level shown is erroneous due to the experimental nature of the project, on 6m this is less important than on HF for any database information. The level stated when accurate is actual Power into 50 Ohms at the Antenna Socket not ERP. The use of beam antennas make any ERP estimate impossible (was the RX station on the side or back of the beam).
The input will be from your favourite PIC modulator with suitable level adjustment and shaping. The LED I show used as a Varicap can be either way up depending upon your normal methods. I use it reverse biassed, the opposite of how its drawn:-)
I added an RC Audio LPF to the output and a divider and pot for easy deviation level setting.
I use my blog as a reference for myself. Today I looked for a picture and description of my 6m si570 MEPT test set up here and found I mentioned it but posted nothing.
To save confusion I have removed the term DDS althogh they are generally getting called DDS now which isn’t strictly correct.
So here we go.
Here we see the s1570, SL612 Mixer, 16F628PIC Modulator, 2n3866 equiv PA.
In total 8 active devices 5 IC’s and 3 Transistors for 100mWatts of WSPR on 50MHz
Radio - LF/MF to Microwaves 