2014-10-09 03:20 AM
2014-10-10 07:26 AM
Attempting to interrupt at 200 KHz seems to be a pretty poor choice here.
Consider if you can DMA the output from the ADC to the DAC directly.2014-10-10 08:47 AM
Consider if you can DMA the output from the ADC to the DAC directly.
That becomes difficult if one wants to insert a signal processing stage between source and sink. Real DSP's would be better suited for this purpose.
2014-10-16 01:13 AM
2014-10-16 12:50 PM
That becomes difficult if one wants to insert a signal processing stage between source and sink.
Indeed, but it would significantly reduce the latency, and variability in that latency. It was thrown for the rhetorical entertainment. The code this was culled from, and other examples, did actually accumulate more data, in a ping/pong fashion, and pipe data to the DAC(s) in a lock/step fashion against a hard timebase. Of course if you're jugging a single sample at 200 KHz, in an interrupt, with non-determinable processing speed, yeah you'll get a bunch of uncontrollable jitter.