FM data line switch grid question

[davelybob]

So, I never did get anything done to my buddy's PSS-140 I posted about previously besides the standard data line on/off switches and the buzzer & pot mod on the rhythm chip (props to the dude who posted THAT, it's easily as cool as the data line mod, I rigged it up with a rocker switch and a momentary so that you could use the momentary to pop it on and scramble the beats, or switch the scrambling fully on and use the momentary to cut it out, AND it's a great addition to the data lines as well because not only does it spazz out the beats while it is on, but using the momentary you can use it to advance whatever loop the datalines are locked into to a new different loop- HIGHLY recommended).  Anywhos, he came through town and I gave it to him and it was like watching a little kid on xmas, he was well happy with it and I told him once I find the time to figure out some more cool shit on mine I will add the mods to his at a later date.  He was really hyped about the possibility of the video out so once I find that, I'll just send him a pic of those points to do himself since it looks like an easy enough mod, I just didn't know WHERE they were hiding on that particular board and ran out of time to search.

I've got a couple more FM Yamahas in the stockpile waiting to be bent, and since  I am absolutely in love with these tiny little 8 switch DIP switches, instead of the patch bay I had been considering (although I will likely add one to connect multiple keyboards together) I think I will have a little breakout box on the side with an 8 switch DIP to shut off the data lines, and then 8 more 7 switch DIPs to connect each CPU point to each FM chip point (1->2,3,4,5,6,7,8; 2->1,3,4,5,6,7,8; etc.).  Now I don't want a bunch of redundant switches or for the switches to connect more than the points they are supposed to, so I was going to put diodes on either side of each connection.  I know you guys have crossed different lines on the same machines and between machines before, but has anybody found using diodes on them to be effective in getting a range of different results?  Which way do the diodes need to be directed, going from the CPU to the FM chip, right?  It's not a two-way data stream is it?  I'm not sure at all how that would work, but if so I would imagine they would mess up normal operation so in that case I would need to leave them off of the "regular" 8 switch DIP (kind of thinking out loud here).  If it is a two way data stream, then theoretically I could get still more varied results by adding 8 more 7 switch DIPs going in the opposite direction (from FM to CPU) as well, or maybe even putting diodes on the 8 switch DIP controlling the main line cuts but having 2 of them to break each connection just one way at a time.  Or maybe I am overthinking this, I don't know- has anybody else gotten this OCD with the data lines before?  Was it worth it?  I'm sure a little trial and error would easily solve this, but right now my commute is killing me and my wife who already had several chronic illnesses came down with muzzafruzzing EPILEPSY and is having a hard time contributing to the upkeep of ye olde domicile, so build time is at a premium, and I'd like to have as much of the design work figured out before I break out the soldering iron as possible, especially since we're looking at possibly wiring connections for up to 128 switches and 256 diodes.  Goddamn, that sounds utterly insane just reading it.  Anyway, I won't get around to this for a while since I'm in the middle of a 4ms Stereo Panneur build and have an SA-38 to trick out after that and a couple more projects for buddies before I revisit my FM boards, so thought I would throw this out there, maybe one of you lucky young bucks with more time to kill than I and a board waiting might want to try it out and maybe report back, I'd be eternally grateful.

Word to your moms.