[sdiy] Din Sync Master Implementation

[maxrest at gmx.de]

I got it sorted out - finally:

The 909 needs some clock ticks before a reset on the start line to work
properly. Now everything starts without a lag.

Find reference timings here for those who will need it in the future:

http://e-rm.de/data/dinsync_timing/mc+_Powerup_Start.png
http://e-rm.de/data/dinsync_timing/mc+_Stop_Start.png

Combined with the signal scheme provided in my first message, seamless start, pause, continue and stop is now possible without a glitch.

Thank you a lot Richie and Maarten for giving me new thoughts on the
subject!

Best,

Max


> On 2014-08-18 15:38, maxrest at gmx.de wrote:
>>
>> But would you agree with me that muting the pulses shouldn't do any harm
>> to the machine?
>
> It won't harm the machine, but whether it will throw some algorithm
> inside is another thing.  Why don't you try both options:  1. Stopping
> the clock pulses, and 2. De-asserting the run signal but keeping the
> clock pulses going.  The latter is what most DINSYNC outputs tend to
> do.  This gives the receiving device a "preview" of the tempo clock
> before the run signal goes active to start playback.
>
>> I mean they don't implement a PLL or something similar
>> that needs to settle and that might mess everything up..
>
> I haven't disassembled the TR-909 ROM code so I really don't know how
> it is done.
>
> For the internal TEMPO clock that is software generated anyway they
> could just run the timer or NCO at 48ppqn speed instead of 24ppqn to
> get the extra shuffle resolution.  But I have no idea how this is done
> for DINSYNC inputs.  I wouldn't be suprised if it acts on the rising
> and falling clock edges.
>
> Maybe you could try setting something up with a couple of debounced
> toggle switches.  One for RUN/STOP and one for TEMPO clock and then
> try manually toggling them very slowly and listening for instruments
> triggering on the rising and falling edges of each clock with
> different shuffle settings.  Then you can see if it only acts on
> rising clocks, acts on rising or falling clocks, or does something
> else instead like a PLL.
>
> -Richie,
>