[sdiy] MIDI notes filter which is also a MIDI converter for Junos ?

[rsdio at sounds.wa.com]

On Dec 29, 2013, at 13:14, Neil Johnson wrote:
> Brian,
>
>> Second, powering a circuit from the MIDI cable is a hack that was  
>> never
>> sanctioned by the MIDI Manufacturer's Association (to my knowledge  
>> - if it
>> has been added to the spec, please let me know).
>
> They are aware of the issue, and have promised some clarification, but
> it seems to be taking some time.  I raised this with the MMA some time
> ago:
>
> http://www.midi.org/cgi-bin/forum/forum.pl/forum.pl?m-1171180161/


Thanks for providing this link, Neil. I really hope that the MMA  
doesn't throw a monkey wrench in the works. In any case, I'll be very  
interested to see what comes of your questions on that thread!

I think it's very interesting that the/an Administrator pointed out  
how MIDI Solutions uses opto-isolators on all MIDI-powered devices  
with multiple MIDI IN jacks.

I agree with your comment, Neil, that providing power is not a simple  
task. It's especially not simple to retrofit into an existing  
specification! Not only does providing power over one line of a  
current loop data link seem fishy, but using the cable shield to pass  
power supply current would defeat the purpose of the shielding. I  
can't guarantee it would be a problem everywhere, but I can imagine  
audio studios filled with MIDI cables transmitting the switching  
power supply frequencies of the MIDI devices. Didn't the AES work for  
decades to get pin 1 problems fixed on balanced audio interconnects?


Looking carefully at the MIDI spec, specifically the example THRU/OUT  
circuit: when providing the required 5 mA, pin 4 is at 3.9 V and pin  
5 is at 1.1 V, for a differential of only 2.8 V maximum. In other  
words, pin 4 only provides 5 V when no current is being drawn at all.  
The more current your MIDI-powered device uses, the lower the input  
voltage. I really don't understand why the MMA wants to condone MIDI- 
powered devices unless they recommend something very thorough.

In my opinion, since you can't get the full 5 V from a MIDI cable  
while pulling any current anyway, all sanctioned MIDI-powered devices  
should be required to have their own charge pump regulators that can  
generate the 5 V or 3.3 V or whatever is needed by their specific  
circuit. Then it's a matter of pulling as little current as necessary  
to generate whatever voltage is needed. There should be no  
requirement that the sending device be powered by 5 V through a 220  
Ohm resistor, since that's a very bad power source. Even with these  
restrictions, it's still a mess to change a current loop data link  
into a power source after the fact.

But, I guess they're not changing the spec so much and telling how  
NOT to do it wrong (implying that what's left is "right").

Then again, I'd rather that the MMA modify the MIDI spec to add a  
filtering network on the cable shield to ground, which would reduce  
the potential for noise - and would just happen to make power over  
MIDI impossible.

Brian Willoughby
Sound Consulting