acl wrote:Finished board
Steve Anderson wrote:Harry, yes those D-As you sent a few of to me should be fine, the specs read OK so no reason not to use them, I'll try a few this end too...
n a blinding flash of pure inspiration or perhaps a case of crass error-prone stupidity, I yesterday concluded this could all be done with just one processor!
Basically omit IC102! Though I'm not 100% convinced - yet. So I may go quiet as I dwell on this, there's a lot of 'hand-waving' and 'fag-packet sketching' to do yet. The rest of it should be much the same as already posted. But there are bound to be a few changes, hopefully minor. This may be like teaching a dog to walk on its hind legs, time will tell.
My saving grace may be that I'm running the read processor at twice the speed of the previous versions.
The biggest hurdle is interleaving the read output, which cannot be interrupted, with the non-synchronous SSTV input, More to come.
Steve A.
The pitfalls of designing as-you-go! If I'd presented this as a fete accompli none of this would have happened, but hey! It wouldn't be as much fun!
Anything like this in the future will use PIC24 (16-bit) devices, it's time to move on from 8-bit. This basically results in four times the throughput, minimum.
e.g. A 16-bit x 16-bit signed multiplication with a 32-bit result on a PIC18 takes 40 cycles, (2.5us), whereas a PIC24 takes maybe four, (250ns or 0.25us).
Who would have thought back in the '50s we'd be talking about these sorts of speeds for SSTV?
acl wrote:Yes Harry I tend to agree with you there. I worked on nuclear sites where the last thing they wanted in their critical control systems was microprocessors that could latch up at any time . Many that did had analogue back up systems. Anyway I have filled in a PTGI form (permission to give in) with the crystals I purchased on Ebay. Tried loading O/P and checking for open collector outputs, Taking the unused pin to zero and still not luck. I have ordered some more with the makers data sheet attached.
Regards Chris Lewis
Harry Dalek wrote:..may be i will build an oscillator with a crystal.
Harry Dalek wrote:I have been looking at the 2 little 1024 ics..
Steve Anderson wrote:Harry Dalek wrote:I have been looking at the 2 little 1024 ics..
Make sure you get the correct devices, 23LC1024s (RAM), not 25LC1024s (EEPROMs), it's an easy mistake to make! And they're both made by the same company, Microchip. What doesn't help is they're both memory and 8-pin devices...but not interchangeable.
Steve A.
Harry Dalek wrote:So am i right in saying the 23lc1024's all come from the same place as i see different sellers.
Return to Cop McDonald's FM SSTV System
Users browsing this forum: No registered users and 2 guests