[sdiy] (OT) Doepfer MIDI-CV oddity

[ChristianH]

No, I don't think delay is a problem here, not in a way that real-time
behavior would suffer.
We have single-digit microseconds here, at least with the PC900 (which
is the 'official' MIDI receive coupler, the standard explicitly shows it
in the suggested circuit).
I think the reason for problems with 3+ stacked THRUs is that the
couplers may have different delays for rising and falling edges (it is
with the PC900), so the serial bit timing gets messed up more and more.

It's quite different if there is a software THRU involved - that one is
quite certain to introduce delay and jitter. However, as far I have seen,
those are pretty rare. (But hey - you have perfectly re-generated bit
timing, isn't that something...)

Chris



On Wed, 19 Jan 2011 08:44:06 -0600 "Jerry Gray-Eskue"
<jerryge at cableone.net> wrote:

> Oh I forgot to mention that rebuilding the pulse train Also adds a delay in 
> the data stream which can be yet another problem for real time stuff like 
> MIDI...
> 
> ----- Original Message ----- 
> From: "Jerry Gray-Eskue" <jerryge at cableone.net>
> To: "Ullrich Peter" <Peter.Ullrich at kapsch.net>; "Colin f" 
> <colin at colinfraser.com>; <synth-diy at dropmix.xs4all.nl>
> Sent: Wednesday, January 19, 2011 8:39 AM
> Subject: Re: [sdiy] (OT) Doepfer MIDI-CV oddity
> 
> 
> >> The problems get bigger and bigger the more MIDI gadgets are connected in 
> >> series as
> >> the MIDI pulses get more and more distortions in each thru stage.
> >
> > This is true of any communication link that uses an "analog" repeater, if 
> > you use a device that regenerates the digital pulse train using a system 
> > clock you get a correction of the distortion at that stage.
> >
> > ----- Original Message ----- 
> > From: "Ullrich Peter" <Peter.Ullrich at kapsch.net>
> > To: "Colin f" <colin at colinfraser.com>; <synth-diy at dropmix.xs4all.nl>
> > Sent: Wednesday, January 19, 2011 6:17 AM
> > Subject: Re: [sdiy] (OT) Doepfer MIDI-CV oddity
> >
> >
> >> Hi!
> >>
> >> And the 6N139 switching times can further be optimized by used a bias 
> >> resistor on the
> >> base of the opto-receiver transistor.  Without this receiver the 6N139 is 
> >> quite slow.
> >> I normally use a 15k resistor from pin 7 to GND.
> >>
> >> The maplin MIDI thrubox has this base resistor in it and works very fine. 
> >> I had the same
> >> circuit without the resistor in my own thrubox and had problems.
> >>
> >> The problems get bigger and bigger the more MIDI gadgets are connected in 
> >> series as
> >> the MIDI pulses get more and more distortions in each thru stage.
> >>
> >> Ciao
> >> Peter
> >>
> >> -----Original Message-----
> >> From: synth-diy-bounces at dropmix.xs4all.nl 
> >> [mailto:synth-diy-bounces at dropmix.xs4all.nl] On Behalf Of Colin f
> >> Sent: Wednesday, January 19, 2011 12:58 PM
> >> To: synth-diy at dropmix.xs4all.nl
> >> Subject: Re: [sdiy] (OT) Doepfer MIDI-CV oddity
> >>
> >>
> >>> As an aside, this is the kind of thing that makes life hard for
> >>> manufacturers-
> >>
> >> Yep, that's why I'm so well acquainted with the range of 6N139s...
> >>
> >> Another new factor in the MIDI equation is that some PC interfaces drive 
> >> the
> >> line with a lower voltage supply.
> >> MIDI is a current loop, so in theory you just need to reduce the series
> >> resistors and 3.3v pull up on pin 4 should be no problem.
> >> But if the forward voltage drop on your opto is higher than usual, you 
> >> might
> >> find the margins of operation sooner than you'd like.
> >>
> >> Cheers,
> >> Colin f