Re: NBTV/SSTV-625 Up Converter MkIII
Posted: Fri Nov 22, 2019 11:47 pm
Steve, your problem is: time. There should be a decoupling of time for the FM-demodulation and the storage into the frame memory. That is also more or less the problem that I encountered one year (?) ago.
Let me separate the two processes:
1. There is a process that measures the amount of time between two zero crossings of the (limited) input signal. The result is converted into a "grey value number". This is a process that is synchronous to the input signal wave form. The latest grey value number is for some time the "voltage" of the video signal. This is until the next zero crossing of the input signal occurs.
2. The second process is a process that runs on an internal clock. This clock is coupled to the sync pulses. That is the sync pulses of the previous SSTV-line, or even better a fly wheel frequency derived from several previous sync pulses.
This process calculates the current position in the frame memory and stores there the latest demodulated grey value number, given by process 1.
And then there is a third process:
3. A crystal stable clock calculates the addresses from where a grey value is read, converted to analogue and transferred to your 625 line monitor.
I don't know if you need absolutely three microcontrollers to perform these three processes, or that you can combine two processes in one uC. I can imagine that the processes 1 and 2 can be combined in one uC, as process 1 runs on a external event (the zero crossings) with the aid of an internal timer, while process 2 will run on the interrupt of a second internal timer. And both processes are relatively slow, while process 3 is much faster.
Does this help? Good luck.!
Let me separate the two processes:
1. There is a process that measures the amount of time between two zero crossings of the (limited) input signal. The result is converted into a "grey value number". This is a process that is synchronous to the input signal wave form. The latest grey value number is for some time the "voltage" of the video signal. This is until the next zero crossing of the input signal occurs.
2. The second process is a process that runs on an internal clock. This clock is coupled to the sync pulses. That is the sync pulses of the previous SSTV-line, or even better a fly wheel frequency derived from several previous sync pulses.
This process calculates the current position in the frame memory and stores there the latest demodulated grey value number, given by process 1.
And then there is a third process:
3. A crystal stable clock calculates the addresses from where a grey value is read, converted to analogue and transferred to your 625 line monitor.
I don't know if you need absolutely three microcontrollers to perform these three processes, or that you can combine two processes in one uC. I can imagine that the processes 1 and 2 can be combined in one uC, as process 1 runs on a external event (the zero crossings) with the aid of an internal timer, while process 2 will run on the interrupt of a second internal timer. And both processes are relatively slow, while process 3 is much faster.
Does this help? Good luck.!