[sdiy] Idea - Triangle wave DCO core

[Tom Wiltshire]

On 14 Mar 2012, at 21:16, ASSI wrote:

> On Wednesday 14 March 2012, 13:46:08, jays at aracnet.com wrote:
>> I assume that we would run an integrator into a comparator which would
>> generate a signal to reset the integrator via the processor.
> 
> Not in a triangle VCO.
> 
>> Normally you'd run that into an interrupt line on the processor.
> 
> Why would you do that, even on a sawtooth core?  Do the reset directly and 
> use a capture unit to measure the time between resets if you need that.

I can think of two reasons:

1) Oscillator sync. If you have a comparator going back to the processor, you can boost the charging current beyond what it should be, then give a reset pulse every time the comparator flips as well as when the timer times out. This gives hard sync sounds with only one oscillator.

2) Amplitude compensation. If you've got a comparator fed back to the processor, it can set up the timer and then test a given charging current to see if the comparator flips or the timer times out first. For stable amplitude, you want the biggest charging current that you can use that still lets the timer time out first (e.g. just below the comparator threshold). If you do this across the range, you can calibrate the DCO output for temperature and component tolerances.

There are other ways, of course. There's more that one way to skin a DCO. For (1) you can feed the comparator output directly to the integrator as another reset pulse. I did this with a simple diode-OR and it worked very nicely.

Tom