[sdiy] Midi powered?

[Dave Magnuson]

I've seen quite a few devices with 3 connections on the MIDI port.  I'm at
work (so I don't have schematics handy) but I beleive the third
(ground)connection is only present at the outputs, but not at inputs (or
did I get that backwards?)  I think it's intended to ground the shield of
the cable, and it's only connected on one end to avoid ground loops.  

Perhaps the devices that use MIDI for power rely on this ground, which may
not be present on all devices.

Dave Magnuson



At 07:55 AM 5/1/03 -0700, Jim Robertson wrote:
>
>Not sure about the 3 connection MIDI hookup.  Never
>seen that before.  But without digging into the MIDI
>spec.....
>
>Standard MIDI connections should be operating as a
>two-wire "current loop".  When the connection is idle
>(no data), the receiving device would see a steady +5V
>level.  During data transmission, the start bit, stop
>bit, and data bits with a value of "1" are represented
>by a 0V level.  
>
>I believe that devices powered from the MIDI line are
>simply filtering the "data" input itself (via a
>capacitor) to provide a steady +5V to power a circuit
>with very modest consumption.  
>
>Since the exact MIDI output circuitry used by any
>given manufacturer may vary somewhat in its design, it
>would follow that each device will have a different
>tolerance for powering external devices.  Most MIDI
>inputs use optoisolators so the MIDI output isn't
>necessarily expected to drive much more than a couple
>LEDs.  
>
>If there are devices that use 3 connections (i.e. one
>of them is a power pin for external devices), then you
>could also run into problems since some MIDI cables
>only use 2 of the pins.  This is because only 2 pins
>are "required" for MIDI transmission.
>
>I would carefully check the spec sheets of your
>devices.  It is oversimplifying the matter to say that
>older devices draw more current than modern devices. 
>There are too many other factors.  Though I realize
>that Colin is generalizing here, and for the most part
>the statement is probably accurate.  I also agree that
>there would likely be a resistor on that data line (to
>protect the output transistor/gate) and it will limit
>the available output current.
>
>Jim, US Air Force 
>Ramstein AB, Germany
>Usually just a lurker.


Resonant Frequency:
resfreq at hoohahrecords.com
http://www.hoohahrecords.com/resfreq/index.html
-------------- next part --------------

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.476 / Virus Database: 273 - Release Date: 4/24/03