[sdiy] Help-Blacet midi-din sync chip problems!

[Theo]

Not 100% sure about the value, but from memory the DINsync needs 
a minimum pulse of 2 msec.
The easy way to overcome the dutycycle problem
(or bad interrupt/poll handling by Roland programmers)
is to convert the outgoing 50% duty cycle to a fixed 2 msec pulse.
Something like a 4093 Schmitt Trigger gate ,a cap and a resistor would 
do the job nicely.

Theo



From: Rob B <cyborgzero at home.com>

> Problem is, this 50% duty cycles screws
> *some* 808s and 606s up pretty badly,
> and they begin to slip.
> 
> I propose a scheme by which the duty
> cycle can be set somehow. The 808s and
> 606s tend to work better on about a 30%
> duty cycle. I used a pwm circuit to get
> the Xbase to sync a 606 properly, funny
> thing is that it synced a 202 and a 303
> just fine with the 50% duty cycle of the
> Xbase.
> 
> 
> Rob
> 
> ----- Original Message -----
> From: "Colin Fraser"
> <colin at colinfraser.com>
> To: "'David D'" <shockadd at hotmail.com>;
> <synth-diy at dropmix.xs4all.nl>
> Sent: Sunday, February 10, 2002 6:51 AM
> Subject: RE: [sdiy] Help-Blacet midi-din
> sync chip problems!
> 
> 
> > > -----Original Message-----
> > > From:
> owner-synth-diy at dropmix.xs4all.nl
> > >
> [mailto:owner-synth-diy at dropmix.xs4all.n
> l] On Behalf Of David D
> > > Sent: 09 February 2002 9:11
> > > To: synth-diy at dropmix.xs4all.nl
> > > Subject: [sdiy] Help-Blacet midi-din
> sync chip problems!
> > >
> > >    So I got the midi to sync
> convertor chip from Blacet and
> > > have constructed
> > > the midi to 24ppqn circuit as
> described in figure three of the data
> > > sheet/schematics that Blacet has on
> their web site and sends
> > > out with the
> > > chip. Ive followed the normal
> conventions for creating such a
> > > circuit and
> > > have at least twelve years
> experience with such-
> > > so I am not concerned about my
> ability but I just cannot seem
> > > to get that
> > > circuit to output a nice stable
> 24ppqn 1ms long pulse! If I sub
> > > a 1meg resistor for the 100k that
> the schemo calls for it is
> > > pretty close to
> > > a 1ms pulse but it is still not very
> stable- it lags and
> > > rushes at a very
> > > high rate.
> >
> > For the Blacet chip I would have used
> a 74LS123 to stretch the pulse -
> > it should be much more reliable.
> > I woudl also recommend at least a 2ms
> pulse.
> > In my own MIDI to SYNC IC (
> http://www.colinfraser.com/m2d/m2d.htm )
> , I
> > used the internal timer to generate
> the pulse.
> > There are a number of devices that I
> have found don't sync well with
> > only a 1ms pulse.
> >
> > Something that has concerned me
> recently is that the internal clocks of
> > the 303, 606 and 808, probably others,
> generate a sync clock with a 50%
> > duty cycle.
> > At a tempo of 120 BPM, this would give
> a pulse length of about 10ms
> > (20ms period * 24 ppqn * 120bpm = 60
> seconds)
> > The 303 uses the trailing edge of the
> clock pulse as a timing reference
> > for turning the gate off (each untied
> note is 3.5 clocks).
> > This means a 303 synced using a fixed
> length pulse of 1 or 2ms, will not
> > have the same gate length as if it
> were clocked internally.
> > 8ms @ 120bpm isn't much of a
> difference, but if you subscribe to
> > Synth-DIY, you have to be pedantic
> about these things ;-)
> >
> > So, an ideal MIDI to SYNC convertor
> should track the tempo, and maintain
> > a 50% duty cycle if you want all
> synced devices to behave exactly as
> > they would with their internal clock.
> > I'm looking at doing a new SYNC chip,
> based on an AVR, with 50% duty
> > clock output, 48ppqn output,
> arpeggiator output etc.
> >
> > Colin f
> >
> >
>