In the week prior to the Easter weekend, having completed my Top Band MEPT I was looking at using a similar design for a 40m unit with my 7.00MHz Crystal. Then I received an Invitation to join DL6NL and G6AVK on 14Mhz for a Holiday Weekend activation.
I searched through my Crystals, all I could find was a 5th overtone one which on its Fundamental was out of the top end of the band. With just 3 days to go there was no time to order an Xtal so I replied to Peter DL6NL “thanks for the invitation to the Party but I have nothing to wear.”
Going back to my 40m construction I wondered if I could select the Second harmonic from a Class C stage and get 14 MHz that way. I succeeded in getting a low level signal which was rather scruffy looking on the scope. I considered a Diode Ring Doubler circuit, fine if I was after a low level signal to amplify using several stages, not really simple MEPT stuff.
I moved to a digital Oscillator. Dividing digital signals is easy, doubling is not normally done, there is little information available on any techniques. If we were to put the 7.00MHz signal through enough digital gates we would get a delay of the signal. Apply the delayed signal and direct signal to a pair of XOR Gates and if the delay is correct we will get 2 pulses in one time period.
Alternatively a phase delay can be achieved with an R/C network both methods worked and on the day of the ‘Ball’ I was able to announce that ‘Cinderella has sewn a new Gown and would be coming to the Party’.
I used the Gate Delay method in a bread board ‘lash up’ with a 2n222 Driver and BFY50 PA.
The 50mW TX fared well, conditions were not good but we were all copied (Cinderella and the Two Ugly Sisters) in Nova Scotia and Northern Utah.
I am below G6AVK and DL6NL in this capture into Utah a very thin signal barely visible but I was running 3dB less ERP and of course the antennas and paths are different.
Peter DL6NL announced the switch off time for his activity Noon on Tuesday. At the stroke of Noon my Cinderella TX disintegrated into a pile of components and a single Crystal (slipper).
I embarked on the permanent version, characteristics were different without the long leads I found a combination of Capacitive Phase shift and Gate delay with critical adjustment of the Capacitor gave precise linear results.
I used to despise Veroboard. Costly,Warping Paxolin and lots of stray capacitance between the rows, however it is quick and easy for IC projects and used wisely the capacitance can effects can be minimised.
The signal generation uses two 74HC86 IC’s Selection of 7.00MHz or 14.00MHz is by the positioning of a single link.
Ideally suited as a PA is the 74LS240, having 2 separately enabled banks of Inverters. Band Selection is by a movable link for Enable and another for the Drive to either section. this means the BPF for each band can be permanently connected to its PA.
Separate plug in Band pass filters are used, this including a socket for the crystal means the MEPT is in fact universal and Multi Band.
Link swapping is no problem as the whole MEPT needs re setting when band changing, the FSK shift is doubled when the signal is doubled.
The MEPT is on its test run now using my spare PIC Keyer piggy backed on.
During the tests I tried a new method of Identification conceived to try and overcome the problem of missed identification due to missing parts of Morse Code symbols.
I now call this 5/5 It is 5WPM FSK Morse with a 5Hz shift. If I send my Call Twice and the two letters of my Major Locator Square this forms a Morse Code G in QRSS5 with a filled waveform.
Here it is into Nova Scotia, doing a fine job.