[sdiy] MIDI optoisolator candidates?

[Steve Lenham]

From: "anthony":


> I'm working on Ray Wilson's MIDI-to-CV converter and dug through the HUGE 
> pile of optoisolators that I've scrounged and came up with ZERO 6N138's. I 
> did have some with Darlington outputs, but not the photodiode/Darlington 
> combo like the 6N138. HOWEVER I do have a number of H11N3's which has a 
> Schmitt trigger output.
>
> I had wondered if this might even work out better than the 6N138 since 
> what follows the 6N138 directly is a Schmitt trigger. What do you all 
> think?

Crucial thing for MIDI is the _speed_ of the isolator rather than its exact 
architecture. Lots of standard optos are simply too slow - by the time they 
drift on/off, the next data bit (or the one after that, or...) has arrived.

The old favourites (6N136, 6N138 - also PC900 and CNY17-2) are rated for 
1Mbps so handle 31.25k with ease. A quick Google shows your H11N3s are 
specced for 5Mbps, so should be great for MIDI.

I got bitten by this a couple of weeks back when I dusted off a MIDI 
interface I made years ago, which I thought I'd tested at the time. The 
H11AV10 I'd used for the opto (because I happened to have one) was far too 
slow, resulting in most MIDI bytes appearing as all zeroes :-( Transplanting 
a 6N136 put things right.

HTH,

Steve L.