Multi Space Filter

[StephenGiles]

Initial ideas are thus:
http://img.photobucket.com/albums/v517/uncle_boko/multispacefilter1.gif
This is basically an extension of Zero's recent FX-25 project containing the dual filter idea with additional envelope generator facilities. The filters are from the Dod Fx-25 while we are on 13600 kick, and the envelope generators are adapted from the EH Space Drum to work with a 13600. There are start and stop frequency controls and an envelope level control for each so the options for different swept sounds (in theory) are limitless.

IC1 and 5 are LM 324 at the moment because this chip was specified in the Space Drum circuit, and ICs 2,3 and 4 are LM13600/13700.

Please note I have only built the envelope generator. Experimentation is required on signal level to each filter, envelope interface with filter current inputs and envelope level control. I hope to have some time this weekend to breadboard this beast. It might also be interesting to use the Bi Filter Follower filter as one of the filters with an opto which should provide a conpletely different response from the enveope generator.

Meanwhile I would welcome input on this.
Stephen

[Steve Newton]

Have you actually "driven" anything with the Env Gen? To see how well it works?

And what had to be adapted/changed for it to drive a 13600?

Also what is the envelope level control for? If you can set where the filter starts and where it sweeps to doesn't that make a level control redundant?

As you said there would appear to be a lot of posibilities in this. Maybe even more so if the filter sections were adaped to be full state variable jobs with Lo/band/hi and notch outputs. Isn't that just adding an opamp mixer at the input of the filter? And tapping the relevant  stages for outputs?

Would be really interesting to hear this in action.

[lovekraft0]

Stephen, yet another challenging idea  - that's another page on my to-do list, thanks! Can you think of any reason why you couldn't (within reason, at any rate - I'm not looking for a vocoder filter bank or anything) continue to add filters, perhaps even 5 or 6? I'm thinking setting the respective sensitivities so that some filters only react to very loud inputs while others sweep even at much lower levels might open doors to some really unusual timbral envelopes that could be controlled by playing dynamics. Now I have to go hunt down some cheap 13600s!

[Coriolis]

I'm thinking setting the respective sensitivities so that some filters only react to very loud inputs while others sweep even at much lower levels might open doors to some really unusual timbral envelopes that could be controlled by playing dynamics.

Now there's a great idea, me thinks! I'd like to try that with all-pass filters to get dynamic phasing, hopefully something a bit more playable than a Small Stone, but not as subtle as a Phase 45...

BTW: Isn't the 13700 a better option than the 13600? Seeing that the former is in production, and that they're similar in operation (they are, right?).

C

[StephenGiles]

Stephen, yet another challenging idea  - that's another page on my to-do list, thanks! Can you think of any reason why you couldn't (within reason, at any rate - I'm not looking for a vocoder filter bank or anything) continue to add filters, perhaps even 5 or 6? I'm thinking setting the respective sensitivities so that some filters only react to very loud inputs while others sweep even at much lower levels might open doors to some really unusual timbral envelopes that could be controlled by playing dynamics. Now I have to go hunt down some cheap 13600s!

Now that could be interesting! Perhaps this evening I'll post some info about the envelope generator. Steve - yes I have driven the filters from the Bi Filter Follower with success, was on breadboard but now pulled out to make room for Space Filter work. I simply exchanged a 13600 for the CA3094 on breadboard so it was just a matter of getting the pin connections right.
Stephen

[Mark Hammer]

Also what is the envelope level control for? If you can set where the filter starts and where it sweeps to doesn't that make a level control redundant?

You can turn the volume up to halfway on something slowly or quickly.  You can also turn the volume up full on something either quickly or slowly.  Big difference.  The Stop and Start controls are analogous to the Attack and Decay controls on a synth, and simply indicate how long it takes for the transition to take place.  The Level control indicates how MUCH transition will take place.

[StephenGiles]

  OK, here is the "how it works" blurb for the envelope generator (formerly in the EH Space Drum with a CA 3094):
IC1a amplifies the incoming guitar signal and its output is sent to the VCF input via a resistor to be decided upon. IC1a output is also buffered and further amplified by IC1b, the gain of which is selected by the Sensitivity control. The voltage at IC1b output is converted into a current suitable for triggering the Sweep Generator by the 20k resistor into the current pin of IC2a .

The sweep generator is formed by IC2a, IC1c and their associated components. With the system at rest, no current flows through the 20k resistor into IC2a pin 1. This results in the transconductance amplifier portion of IC2a being effectively disconnected from its darlington emitter follower output stage, hence the voltage at IC2a pin 8 follows pin 5, which is at the voltage set by the Stop Frequency pot and buffered by IC1c.(bugger me surely this can be put to good use............sample & hold  )

When a note is played on the guitar, IC2a is energised by the current into its pin 1 and becomes a voltage follower, forcing IC2a pin 8 to equal the IC2a pin 3 voltage, set by the Start Frequency pot. Soon after each note is plucked, current  into IC2a pin 1 ceases to flow  (why.......anyone?) however, and the voltage on IC2a pin 5 decays through the Sweep Rate CR network to the voltage set by the Stop Frequency control, at a rate set by the Sweep Rate control - which is virtually independent of the natural decay of the note played, and will therefore provide a smooth filter sweep with very minimal jitter even at long sweep settings.

So maybe Mark in this instance the Stop and Start controls set "where" the transition occurrs, and the Sweep Rate control sets how long each transition takes to get back where it started. I think one needs to either listen to a Space Drum or better, a Microsynth to get a feel for the way in which the various settings of the Start and Stop controls change the sweep.
Stephen

[moosapotamus]

Or, maybe... the Start Frequency and Stop Frequency pots set the frequency at which the sweep starts and stops, and the Sweep Rate pot sets how fast the sweep goes from the start freq to the stop freq? Low start and hi stop would give a sweep up... hi start and low stop would give a sweep down, yes/no/maybe?

If so, seems like the resulting attack/decay would be fixed and symetrical, maybe? In other words, you can't have a fast attack with a slow decay, because the Sweep Rate pot controls the overall rate of the entire sweep, from the start freq to the stop freq... again, maybe? That might have a very unique sound/feel compared to what you get from a typical attack/decay envelope.

Oh... or, maybe the Sensitivity pot sets how fast (or how close) you get to the start freq depending on how hard you play? That would give you control over the attack. Then the Sweep Rate would set the decay, um... maybe?

~ Charlie

[StephenGiles]

That's interesting Charlie, what I remember of the Space Drum what that when it was struck, there was only a very short attack phase because, I suppose, the intention was to provide drummers with a gizmo that will either sweep up or down immediately he strikes it.

Now, in my recent breadboarding to which I referred on my previous post, I don't recall a noticable attack phase, and......I forgot to mention this, the peak of the sweep is proportional to the strength of the note played, in case you couldn't tell (my name is tricky Dicky and I'm doin' very well - ah dear old Ian Dury), but it decays at the rate set by the Sweep Rate control, thus eliminating the usual jitter found in autowah circuits.

What I suggest is that you make the envelope generator, and look at the 13600 output (which I think should have a 4k7 resistor to ground at the darlington output) on a volt meter, so that you can see the needle sweep for yourself.
Stephen

[Steve Newton]

So maybe Mark in this instance the Stop and Start controls set "where" the transition occurrs, and the Sweep Rate control sets how long each transition takes to get back where it started. I think one needs to either listen to a Space Drum or better, a Microsynth to get a feel for the way in which the various settings of the Start and Stop controls change the sweep.

That's how I'm reading the schem. Start and Stop set the extent of "travel" of the output and Sweep Rate the speed of travel. Which I why I can't see the point of the envelope level control which would surely just scale down the width of the envelope sweep?

The ability to set where a filter sweeps from and to with greater accuracy than juggling cut off and sensitivity pots on normal autowah/filter pedals would be great!

[StephenGiles]

I think you could be right there Steve, but I hope to test this all out over the weekend - my wife has a school reunion tomorrow which gives me maybe 5 hours, although already dwindling from "a little shopping to be done"! and her monthly sing on Sunday with the North London Sacred Harp group which gives me 6 hours - and our neighbours are away............yipeeeeeeeeee volume up to 11!!

[StephenGiles]

Any further views on this? I would hate to drag anyone away from his booster or fuzz box

[liquids]

I just found this, and I think this is much of what I'm looking for, actually, given this post:
http://www.diystompboxes.com/smfforum/index.php?topic=73702.0

So it's kind of exciting, but I have little to no idea how to dissect the circuit and isolate the filter section or the voltage control section, other than enthusiasm.

I'd like to figure out how to hook the start/stop frequency up to something like tim's Q&D VCF  http://folkurban.com/Site/QDVCF-714.html  or MS-20 Sallen-Key LP VCF http://folkurban.com/Site/690_25337.gif?1173243628249  for example...

but I basically have no idea how to do that.   

So for starters, does the above circuit work?  It's got dual filters, no?  The FX-25 filter used here is bandpass, correct?  I'm looking for a LP filter.  But even so, for starts, what could I eliminate if I breadboard it, for just one filter sweep?   Lots of interest, but lots of questions...   

[liquids]

Likewise -- Stephen, maybe you are the guy that is smart enough, armed with this info, to rip the wonderful filter section out of the Micro Synth and mod it for 13600 or 13700 OTA's as an independent circuit?   

Or anyone else?  This could really be phenomenal for guitarist who want a synthy sweep but realize that "triggered" and controlled filter sweep makes more sense than following the guitars dynamics for really synthy sounds...it's an ASDR for the filter sweep if nothing else, given that the guitars ASDR is pretty one-trick pony, and all forms of compression/distorting it give you cool tones but sacrifice the whole ASDR thing in general.

This would conceivably blow the doors off the digi-synth type pedals that I think sound cool but are missing a lot in adjustability, tracking, tone...but would also give the option to run your own tones through it (think of all the synth-wave shaper/octave etc circuits here that could be combined with this) for maximum impact and customizability for out there tones!

Of course that music from outer space guitar synth thing keeps calling my name...if only that EH microsynth weren't so darn big...

[Mark Hammer]

The FX-25 is actually a state-variable filter (or at least one type of same), and the bandpass output is tapped.  I haven't tried it myself, but apparently it can provide a lowpass output as well.

[StephenGiles]

What I want is the initial peak of the sweep to follow the peak of the note and the decay to be independently of the note.

[liquids]

What I want is the initial peak of the sweep to follow the peak of the note and the decay to be independently of the note.

not to go backwards...and maybe this would be even MORE difficult...but  houw about this:

An an oscilator that is triggered by the guitar, without care over frequency (of the input or the oscilator).   Modify this oscilator with a tweakable ASDR.

Having the filter sweep the input's audio according to that ASDR.   

It seems that defining what point it switches from following the guitars peak to going into auto-sweep would be hairy...

Is this even possible?

[StephenGiles]

I looked at some keyboard controllers and what seems to happen there is when a key is pressed, it makes contact with a resistor chain connected between a high and a low voltage (say) the voltage at the point of contact is passed to a sample & hold circuit from which the held voltage is fed to a VCO.

in simple terms, if the guitar input drives a buffer/amplifier, where the instantaneous peak output voltage is sampled and passed to a voltage follower charging a capacitor to that same voltage, and then allowed to discharge at a rate independent of the guitar note, in theory it should work. Please give it a try, I need to have a bit of a clear up before I can devote space to a breadboard - it's as bad as that!!

[StephenGiles]

http://www.modular.fonik.de/files/EFMforum/orgEFMfiles/kbd2e.pdf

Look at this keyboard controller for instance, from point A you have some of the building blocks required - trigger/gate generator and sample & hold.

Think what would happen if

1.  The voltage output of buffer/amplifier (IC1b)  was also fed to this trigger/gate generator and sample & hold

2    The top end of the Start Frequency pot was connected to the sample & hold output instead of simply connected high. 

Perhaps when energised, the sweep generator built around IC2a and IC1c might sweep up to the sampled peak voltage extracted from the guitar, and then sweep down to the voltage set by the Stop Frequency at a rate set by the sweep rate, which of course is wholly isolated from anything happening in at the output of IC1b.

And perhaps the VCF would sweep up quickly to a point proportional to the strength of the note struck and then sweep down smoothly according to the settings of the Stop Frequency and Sweep Rate pots.

[StephenGiles]