Now I know this can be done...

[thesimulacre]

...This Synare Percussion Synthesizer I am severely under-using is in need of triggering. And not just pads (those are long gone), or buttons (those are taking up a hand that could be used to adjust resonance), but audio-level signals somehow causing a mere closing of a switch.

I have a hunch that a reed switch (actually 12 of them) is what I should be looking for, with maybe relays? If hits were triggered by piezo it would in fact already be done, but it's actually more simple than that. There is one hot lead coming from the underside of this beast that can make a circuit with any one of 12 paths to the 4 oscillators at a time. Monophonic, see? I have a bank of 12 momentary switches that I kind of slung together as a substitute for the ungainly and crumbling rubber pad apparatus, but how would I go about substituting an audio transient from my soundcards outputs to "push the buttons for me." I think I am writing this backwards, and fairly certain the answers are somewhere on this very page...

....so what would it take to turn 12 (or four with three frequency ranges each) +4 outputs from the patchbay into physical closings of 12 momentary switches?

There are no schematics, not even zool.

[RefD]

so, they will work with audio as a trigger?

(i'm assuming they don't need a clean gate or pulse.)

maybe piezo pickups mounted on plywood with $1 mouse pads glued to the opposite sides for playing surfaces?

sorry, i'm not familiar enough with the Synare...just thinking out loud.

[thesimulacre]

no no... that's the thing. They won't work with audio as a trigger. That would be too easy. If I were to hook a piezo to this thing it would instantly trigger and the buzzer would buzz like crazy from the current flowing through it. I just need to figure out a component that can translate an audio signal to a physical completion of the circuit at one of 12 points. Is this what relays do I wonder?

It's kind of hard to explain.. but thank you for trying.

[RefD]

there are a few different Synares.

knowing which one would help us help you.

is it one of these?



if not, could you post a pic of yours and maybe some additional info?

[thesimulacre]

I dunno, the link was broken. I had a Synare 03 ( http://usera.imagecave.com/thesimulacre/DSC01938.JPG ) which I sold to a boy named Brion Patrick Hill who never payed me and then changed his info to read France, not NY. Stupid-a__ thieves. Anyway, the one I am modifying is a Synare 01:



From which I have already removed the original trigger pads (just contacts) and to which I added a "breakout cable" of sorts:




I hope the visuals help. The red wire is hot and WAS wired through 12 switches to all of those returns at the same time... though only one could trigger the oscillator at a time. See what I mean?

[sparky]

james-- here's a ghetto idea that would at least be easy to try. lets for a second assume that you can convert your audio triggers to something fairly clean independent of this, like...you have a computer. so like if you had logic you could take the drum tracks, strip silence/audio->score, tweak the midi notes, and assign the midi notes to a esx-24 playing like...half-cycle long squarish waves. you could just edit down one sample in the sample editor and assign it as a oneshot with no transpose. i'm sure in PT or all the other programs you can do the same thing but i don't know how to use them. So then whatever you audio is is ultimately showing up split up on your patch bay outputs as individual pulses.

now, take some crappy little speakers from radio shack or a toy. screw them to a board with some offsetting bushings at like .5 the xmax of the speakers. man you are going to need some amplification channels for this too but lets assume you have a couple spare channels of hifi amplification from like an old stereo, at least you could use to see if this works.

take a piece of tinfoil, glue it to the board. glue another to the speaker's cone. solder a bit of wire to each of the tinfoil pads. This is your switch, so take these leads back to the synare.

Now, when you drive the pulse audio from the computer into the speaker, it will push the speaker out towards the board, making the tinfoil bits touch and triggering the synare. its like a real ghetto audio-triggered momentary switch.

if you think about it, a relay is basically the same thing as a speaker's motor. voice coil = coil, etc. the difference is reusing a shitty speaker is cheap/free and you already know that a speaker's motor structure, etc is designed to operate at audio speeds and you can reuse handy study surplus as the driving electronics.

i'm sure there's like a good way to do this but this seems like a low-risk way to try it.

[thesimulacre]

Boy howdy did that ever get the gears turning! Some serious McGuyvering there. I think the square waves are gonna be the way to go, but I will have to generate them inside of the computer and send them out the ADAT so that I can retain my analog outputs. This means that I will have 8 (sadly, not 12) unbalanced signals for triggering. Man am I ever glad nobody wanted to buy that Hosa patchbay!

So I have (finally) done some research on reed switches, and it seems that for tidyness-sake I will most likely go this route... assuming the voltage being passed by them (coming from the Synare) is not too high. Wait, wait... the leads required are not adding up. A reed relay is what I need, right? So anyway, I just need to measure the constant voltage & current coming from the hot lead on the synare, and match this to the part of the relay that is switched; then create a signal and measure an appropriate voltage (maybe -6dB) from the ADATs analog outputs... matching this to the part of the relay that does the switching.

Phew!

It looks like I'm gonna be using that Usine program after all. I can't think of anything else that could sequence squarewaves with MIDI, send out of multiple outputs while indside a separate sequencer software, AND save presets for; because I really don't want to set this up every time! When I get the voltages and figure out the symbols maybe I'll draw up a little diagram for all the visual-spacial types like myself, for instance.

Thanks for the push!

[thesimulacre]

OK after breaking out the multimeter I have successfully generated more mystery. It appears that all 12 leads are sending +14.55V and the solitary red wire is the ground. Duh. Silly me for assuming the red wire was hot. On a non-sarcastic level I guess it was silly to assume that the only common wire wasn't a ground.

Especially when the darned thing works no matter what I ground it to. Meaning, in the process of switching the meter over to measure the voltage coming from my interface output, I got my operations out of order and connected pad one to the 'tip' of a cable coming from my soundcard via patchbay. *triggered!*

How about ring? Same. Sleeve? Same. T,R,&S from the headphone out? Same. Completely different headphone out on a different mixer on the other side of the desk? Same. Anything metal and grounded in the entire room? Same.

I think I need to consult Forrest Mims again before someone responds and I feel like a complete jacktard. But even without a 'why' it seems like this will be easier somehow. Easier, harder, easier, still not sequencing the analog drum machine.... AAAHH!

[RefD]

this is the Synare in question?

[thesimulacre]

Yeah, wow. That is exactly the Synare in question. And that's my desk.

I love your avatar, btw...

[sparky]

james I don't know shit about electronics, like the Forrest Mims has never made it off my Wishlist into the Shopping Cart, but two things come to mind:

1) paia might sell a widget for this problem in kit form? I haven't read their catalog in like 100 years but i remember they had all kinds of midi->cv/gate kits and such.

2) my old roommate was working on this art project once a couple years ago, and in a way, it had similar requirements to what you want to do. He used a device that was like a relay, but was solid state. it was cheap and fast and didn't have breakable parts. i googled "solid state relay" and something came up, but...i feel like it had a weirder name. i have a vague fear that a) relays will be super expensive, idk about these reed ones but I remember looking for relays for this other project i wanted to do and they were really expensive. and b) if something has to move, does that make it too slow? Like lets say it takes 10ms to close. Thats going to be audible delay. OTOH maybe these reed things are super fast, i don' t know. anyway might be another thing to google.

on the s/w end idk what prog you are using but in Logic you can definitely use EXS-24 to do a multi-output square wave setup w/ presets, etc.

[The Scum]

Without reverse engineering what's going on inside the thing, it's hard to say what's going on. They could have a signal-conditioning input that triggers on any little line glitch. I'd expect a sample & hold, too. If you understood that stage of the circuit, you could more effectively interface with it...and perhaps go beyond just triggering it to get some pitch control, too.

If you just want to go the brute-force way, I'd look into a MIDI to trigger converter. Paia had one, and I think theres a MIDIbox version that does it, also.

To get from a trigger to a relay closure, you'll probably need a transistor switch to get the relay to close...otherwise the relay coil will draw more current than the TTL outputs will want to drive. Depending on the quality of the trigger output, you might also need a one-shot or pulse stretcher to make pulses that will switch the relays.

Don't forget that you need a diode in series with the relay coil, so that it doesn't destructuvely kick kurrent tnto the driving circuit when switched off.

[ashcat_lt]

Does this help?

[thesimulacre]

Hey Lord Byron! I didn't know you were still around. Thanks for stirring the 'ol pot of brain stew. Funny, the thought of using a transistor as a switch crossed my mind at the same time as Mr. Mimms, because that is the chapter I went over about 15 times trying to understand. The thing is, while drawing it out and setting up a Reason template with 8 squarewaves coming out of 8 separate channels I thought about skipping the middleman. Arduino came to mind, and while scouring their site looking for a way to transform midi directly to voltages I was reminded that I had seen that somewhere....

Voila!


and not only that, but *cue beam of light from the heavens*


I was starting to suspect I was getting into a whole lot of man-hours here, and this sort of confirms that. There goes all of my DIY motivation, right there. Especially since it affords the opportunity of selecting a MIDI channel (say 11) from within any project and just hammering away, instead of opening an instance of Reason, that template, connecting it via ReWire, routing those outputs to the ADAT, re-patching the ADATs outs to Synare instead of summing, and possibly experiencing more latency than the Doepfer route.

Speaking of latency sparky, the published response time on these particular relays is <2ms, which is close to that of my current DAW, but still within the threshold of generally unnoticeable. I guess if I wanted to go all out I could replace them with the relays from NTE I was looking at which hover around 300&#956;s.

The boards should fit inside the Synare itself with nothing but a MIDI cable protruding! Assuming I can tap power internally somewhere. Otherwise, two cables... Shoot, maybe they will fit inside the enclosure for the bank-of-buttons and dual joysticks that was my plan A! Best of both worlds?


This brings up questions about the possibility of modifying the A and D of the volume envelope and Freq and Rez of the filter using only pots and maybe some capacitor changes. But maybe there is a DIY synth forum out there that would be a little more keen.

Enough caffeine.

[ashcat_lt]

How 'bout this? Not my drawing.