[sdiy] MIDI output current and MIDI-driven gadgets

[Ben Bradley]

The MIDI standard already 'suggests' a UART be used by setting the
standard serial spec to 8-N-1. The only new suggestion I see here is
to use a faster data rate, which I wholeheartedly endorse.

I might suggest something like RS422 or RS485, which are plenty fast,
but neither of these have the isolation that is needed. Ethernet
hardware is isolated, but the signal would need a more complicated
data format to go through the isolation transformers. OTOH, I'd hope
there are much faster optoisolaters now than when MIDI was spec'd,
though the ones needed at the time were significantly faster than the
jellybean ones back then.

MIDI 2.0 officially came out a few years ago, I downloaded and did a
cursory read-through of the spec, there's much faster data transfer
now but I forget if it's through USB (UGH!) or what.

I vaguely recall in the '80's or '90's a manufacturer made a synth
with a "2x speed" option for its MIDI ports, so that at least when
connected to other units of that brand there was the option to go
twice the standard MIDI speed. No doubt the MIDI association frowned
heavily on that, and may have made them stop doing that.

I actually watched the first half of this video a few days ago, it's
all about the pains of MIDI in the first years. I didn't even watch
the rest to see his solution:
https://www.youtube.com/watch?v=xm3ThKXABJ0

On Fri, 5 Jun 2026 at 02:30, Michael E Caloroso via Synth-diy
<synth-diy at synth-diy.org> wrote:
>
> Because there are too many cooks in the kitchen.
>
> MIDI originally specified inexpensive components which add about $25 to the BOM cost (1983 economy).  It was also the easiest solution to prevent interface problems.
>
> The biggest barrier to acceptance of MIDI was getting the manufacturers in agreement and on the same page.  Adding a UART to a new protocol would invite all kinds of complications.  A UART-based protocol would not only invite disputes on low level programming algorithms, there would be an endless dispute over which UART.  Or which microcontroller or CPU with onboard UARTs.  Not only are UARTs not inexpensive components, but they go obsolete.  The MIDI organization mandated the components that will not go obsolete while mandating the optoisolator but not the specific type (the spec does not mandate 6N138/9 or PC900).
>
> In 1983 at the birth of MIDI, there were about 10 member manufacturers on the MIDI committee,  Today there are over a hundred of them (https://midi.org/partners).  The organization would have to get tenfold more members to agree on any UART-based proposal, which is a very very tall order  Dave Smith had said even getting ten to agree back in 1983 was a major headache that he cares not to repeat.
>
> Ain't gonna happen.
>
> MC
>
> On Thu, Jun 4, 2026 at 9:31 AM Benjamin Tremblay via Synth-diy <synth-diy at synth-diy.org> wrote:
>>
>> Why can’t we just develop a new protocol based on cheap UARTs and passive components that runs I dunno 25x faster and works like patch cords instead of hubs, isn’t gated by some consortium, and is backward compatible if you add a cheap adapter?
>>
> ________________________________________________________
> This is the Synth-diy mailing list
> Submit email to: Synth-diy at synth-diy.org
> View archive at: https://synth-diy.org/pipermail/synth-diy/
> Check your settings at: https://synth-diy.org/mailman/listinfo/synth-diy
> Selling or trading? Use marketplace at synth-diy.org