Wired the output of the FIFO to the AD9761 dual DAC.
wrote a test program that makes an IQ sequential file with 2 sine waves of different frequency for I and Q:
Added a 4046 PLL, could not find any 74HC[T]4046 in the box, found an old CD4046, using it at its extreme of 4.1 MHz (officially goes to
2???) Had a fight with the AD9761 datasheet terminology: Inputs are 'latched at the positive edge of the clock'. A latch and a D flipflop in my experience are 2 very different things, I would expect a latch to pass on the signal when its control input is high (or low), and a D flipflop to trigger on a positive (or negative) edge, but a latch that triggers on a positive edge?Comments? This right?
Anyways that caused some circuit changes, needed an inverter, only had an old
7402 quad NOR in the box, (not LS of HC[T], YES a good ol'd TTL!!! ) It worked after 40 years of deep sleep, replaced the diode OR gate with it (I take back what I wrote in previous post about diode OR gates, it was not fast enough with the Raspberry flat cable as load) so now I had a NOR on board, used that).All works, the test program version 2 (has loop function):
All this makes an IQ file with 2 audio tones if you clock it at 4.096 MHz... It shows 100% channel separation and correct working of FIFO wait function (I can tune the frequency way down and the output IQ goes down with it). Good output amplitude, connected an earphone to listen to it. In this picture the scope is not triggered, and the 2 sine waves have different frequency (are in no way locked, see C source), was difficult to take a picture of that:
So now I can use these IQ signals or more preferably ones recorded at 2048 Msamples /second with rtl_sdr, to modulate a carrier, this will be done with an AD8346.
Probably the first test will be to record a FM station as IQ samples, and then replay the IQ samples into the AD8346 QAM modulator, and see if out comes usable audio after demodulation (with rtl_fm).
IF that works then the 'Anything Transmitter' 'universal replay attacker' or whatever will have been proven.
It is a bit challenging to see how these IQ signals will recreate for example the spectrum of an FM modulated carrier.
Fun Fun Fun...