[LAD] [semi-OT] midi snakes using CAT5?

Gene Heskett gene.heskett at verizon.net
Mon Nov 2 00:27:52 UTC 2009

On Sunday 01 November 2009, Jörn Nettingsmeier wrote:
>Arnold Krille wrote:
>> Hi,
>> On Sunday 01 November 2009 16:47:44 Jörn Nettingsmeier wrote:
>>> i'm playing with my shiny new BCF2K, and i'm going to use it some
>>> distance from my machine, so i'm going to try a midi link instead of
>>> USB. what is the maximum length of midi chain that you have used without
>>> problems? i read somewhere that no more than 15 meters are recommended,
>>> which strikes me as pretty short even if it's unbalanced.
>>> my idea is to abuse a cat5 cable as a triple midi loom - do you think
>>> that could work? pinout would be as follows:
>>> 1	midi 1 signal
>>> 2	midi 1 ground
>>> 3	midi 2 signal
>>> 4	midi 2 ground
>>> 5	midi 3 signal
>>> 6	midi 6 ground
>>> 7	common +5V
>>> 8	common +5V
>> If you want to make use of the drilling of the pins in cat5, it should be
>> 1+2 Midi 1
>> 3+6 Midi 2
>> 4+5 Midi 3
>> 7+8 Midi 4
>sure, i just numbered the wires, your count is based on rj45, but i
>meant the same.
>> Why do you want to carry 5V? According to my docs, Midi is just Signal+
>> and Signal- and shield. Which, when carried over Cat5, should be all on
>> the shield.
>any midi has to carry 5v iirc. and midi is not balanced, so there is no
>signal+ and signal-. would be cool if there was - then 100m would be no
>problem :)

Preface, I am a C.E.T., and a broadcast engineer with 47 years time in the 
field.  Yeah, that and my 75 years age makes me an old fart. ;-)

Midi, by spec, is a current source, just enough to run the opto-isolator in 
the next piece of gear.  The rise and fall times become degraded by cable 
capacitance, eventually leading to timing errors.  AFAIK, no one has actually 
tried to run it using a low capacitance cable to extend the range.  In cat5, 
the twisted pair is probably only a slightly higher impedance than the 
microphone cables I have bought labeled as very expensive midi cable just 
because they had the right connectors on the end.  So it might go a little 
further than the store bought cables, but not by very much.

Another effect in the 4 twisted pair cat5 is crosstalk if using each pair as 
a different midi port, but I don't think that would show up before the signal 
lags ate the signals lunch.

Just for grins, I once changed the crystal in an rs232 interface so that it 
could be run at midi speeds, then hooked it up to a couple of different 
keyboards.  The polarity inherent in the opto-isolator inputs seemed to 
protect the input and I ran it that way for a couple years with no damages to 
either keyboard, or to the rs-232 ports hardware.

From an engineering standpoint, midi, designed as an idiot proof interface, 
has served us very well indeed.  But with cat5 all over the place and cheap 
as dirt, I am amazed that the gear makers haven't switched interfaces. Today, 
the average midi circuit is badly overloaded because its so slow. I've heard 
music that because of the note update rate, sounded like Floyd Cramer's 
piano, and it wasn't written that way...  Gigahertz ethernet can be run 
several hundred feet, with 200-2000x the data rate midi can manage.  Think of 
the possibilities that could open up.  Every instrument in a 200 piece 
orchestral construction getting its note on in the same millisecond.  That 
might be too perfect, too mechanical, but you get the idea.

Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
The NRA is offering FREE Associate memberships to anyone who wants them.

"I will make no bargains with terrorist hardware."
-- Peter da Silva

More information about the Linux-audio-dev mailing list