latest hairbrain half baked idea... could use some thoughts.. a fuzz phaser

[haveyouseenhim]

How about going the totally impractical route and using a spinning shade on a bulb.

[deadastronaut]

4017...and get 8 'shades' of fuzziness.

choppy or smooth...

different filters/eq from 1 to 8... 1 being the highest...to 4-5 being the lowest...working its way back to 1...for a full sweep...

i need more coffee...

[pinkjimiphoton]

well i was thinking 4 differently voiced fuzzes. but sweeping between different input caps into one fuzz may be easier i guess... thinking a ramp up/down led and ldr can replace pots in it to make it smoother..

maybe can switch in a 555 for a square wave on/off effect.

thanks for the ideas guys... would love to collaborate on a forum project... hint hint..

[duck_arse]

the 4520 dual binary up/down counter, will give you 2 x 16 steps of fade/brightness/led drive/something. one counts up as the other counts down. any help?

[Mark Hammer]

well i was thinking 4 differently voiced fuzzes. but sweeping between different input caps into one fuzz may be easier i guess... thinking a ramp up/down led and ldr can replace pots in it to make it smoother..

maybe can switch in a 555 for a square wave on/off effect.

thanks for the ideas guys... would love to collaborate on a forum project... hint hint..

Check out the Roland Double Beat fuzz wah.  The fuzz section has a couple of different passive networks on the output, to produce very different sounds.  I whipped up one with a 6-position rotary, to extend the sorts of filtering applied.  Switching between those filtered outputs gets a lot of variety.

[pinkjimiphoton]

the 4520 dual binary up/down counter, will give you 2 x 16 steps of fade/brightness/led drive/something. one counts up as the other counts down. any help?

it may... still way above my paygrade to understand without having the thing in front of me. fuzzes are easy. this stuff (for me) is not.

well i was thinking 4 differently voiced fuzzes. but sweeping between different input caps into one fuzz may be easier i guess... thinking a ramp up/down led and ldr can replace pots in it to make it smoother..

maybe can switch in a 555 for a square wave on/off effect.

thanks for the ideas guys... would love to collaborate on a forum project... hint hint..

Check out the Roland Double Beat fuzz wah.  The fuzz section has a couple of different passive networks on the output, to produce very different sounds.  I whipped up one with a 6-position rotary, to extend the sorts of filtering applied.  Switching between those filtered outputs gets a lot of variety.

is it a scoop, like on the shin-ei pedals mark?

something like that may be perfect.. i would imagine to get it to "phase" tho, it would take at least 4 stages, right?

so how to pan between the equivalent of 4 pots on 4 different voiced filters becomes the next step maybe?

larry, could that chip be made to pulse on and off, too, alternatingly?

sorry for all the questions guys, and i appreciate the answers... need to try and see a way to implement this on the cheap, with readily available parts if at all possible in the easiest way possible.

i bet crossfading between different eq curves on the input of the fuzz would sound very phasey. or more quacky if the fuzz came first...

so thinking audio wise,  have maybe  envelope control > buffer> "? the thing in the middle?" fuzz (or even better, two fuzzes, if we're gonna sweep something like that, i bet it would sound amazing, like moving icebergs in stereo at real slow modulation speeds) and an output buffer.

does that sound right?

maybe a toy electric motor and take the stop out of a wah pot so it just spins?

i'm thinking if it were a stereo fuzz output,  two filters per fuzz would probably work to create a four stage phase shift?

or would we need 8 in total? i wish i could describe the sound i'm hearing in my head.. it's phasey and wah-y but it's fuzz, yet somehow still phasey and wah-y. i think it definitely needs an opto stage to get that watery sound you get.

thanks for putting up with me guys. if we can figure out a way to do this, i'll be grateful.. to be able to electronically build the sound in your head would be a truly awesome thing.

[mth5044]

4017...and get 8 'shades' of fuzziness.

choppy or smooth...

How do you make the transistion between the outputs smooth?

[deadastronaut]

@#$% knows..

just thinking aloud..whilst half asleep.

[Mark Hammer]

4017...and get 8 'shades' of fuzziness.

choppy or smooth...

How do you make the transistion between the outputs smooth?

Well, here is perhaps where it might get kinda weird or complicated.

One possible route to go is to use a quadrature LFO, where you have four triangle or possibly ramp outputs, with output B 90 degrees out from A, C is 180 degrees out and D 270 degrees out.  The outputs control the gates of JFETs set up as part of an attenuator network.  That is, you have four different filtered versions of the basic fuzz, each with its own resistor/JFET attenuator (not all that different from the input of an Orange Squeezer), and feeding a 4-input mixer.  Each of the differently filtered outputs becomes prominent for its own litte chunk of time, while the others fade in and fade out, like a revolving door where you can "see" everybody going around, but one of the people really sticks out more at any given time.  As always, there's a relevant scan here from POLYPHONY, courtesy of Thomas Henry: http://hammer.ampage.org/files/quad-shep.pdf

The downside is that it limits you to morphing over "only" 4 versions of your fuzz.

[mth5044]

Mark - I was just looking at Quad LFO's, but finding a schematic is tough! We must have been having similar thoughts at similar times. Thanks for linking to that scan - but I can't get it to load on neither my computer or my phone. Does it work for anyone else?

I was having a look at phase shifters using inductors, but my brain exploded.

As far using using 4 LFOs at 90 degrees separation - I hope we can get a schematic! The LFOs can drive LED's that could blend in the four fuzz circuits. I can't imagine more than 4 fuzz's initially! Might as well use a phaser and a fuzz in one if you want a bunch of them, no?

Anyway, if the period of the LFO was set to 4 seconds (for ease of this demonstration), the fuzz levels would be like this (numbers 0, 50 and 100 are % volume):

Time:         0     1     2     3     4(0)
Fuzz   1  -  100  50    0    50   100
        2   -  50   100  50   0      50
        3   -  50    0    50  100    50
        4   -   0    50  100  50     0

If fuzz 1 - 4 were set up with the LFOs at 0, 90, 180 and 270 phases. Or something like that. So at each 25% of the period, one fuzz would be full on, two would be at 50% volume and one would be off. That would cycle through each 25% of the sweep until we end back up again at the beginning (0/4).

That seems really cool.

[Mark Hammer]

Mark - I was just looking at Quad LFO's, but finding a schematic is tough! We must have been having similar thoughts at similar times. Thanks for linking to that scan - but I can't get it to load on neither my computer or my phone. Does it work for anyone else?

Works (i.e., loads up) for me....and not because I have any special privileges.  You can simly do a google image search for "quadrature oscillator" and wads of designs will pop up.  OR...you can look here:  http://www.birthofasynth.com/Thomas_Henry/Pages/QFG.html

[mth5044]

Thanks, that worked well. My that's a big circuit.

[Mark Hammer]

The scan I linked to earlier is also a Thomas Henry circuit, and is decidely more complex than a simple 2 op-amp LFO, but is also simpler than the birthofasynth circuit (which is intended to cover a wider range, as well as being more flexible and suited to synths).

[mth5044]

Well, here's my idea. U1 = 13700, U2 and 3 = OPA2134 (possibly other op amps?)



The top left is the quad LFO that is based on a schematic posted by frijitz on electro-music.com forum. Sine and cosine outputs that are synthed and inverted to get the 0, 90, 180 and 270 phase shifts of the same LFO. The rate is controlled by the amps the 13700 sees. The four outputs drive four independant bases of transistors that fade LED's in and out. Perhaps the LED's could be driven from phase outputs? I'm not sure what kind of power they put out. That would make it slightly more simple.

The audio section is just the AMZ splitter four times to four different fuzzs. Between the splitter and the fuzz input is an LDR much like Taylor's Tap Trem to control the volume. As the LFO makes its phase cycles, the LDRs open up in sequence which allows signal to the fuzzes. The fuzz circuits are then mixed simply by mixing resistors. Could be buffered mixing, not sure. You'd also have to make sure that the fuzz circuit it self does not invert the phase or that could lead to mixing problems.

This is not tested, breadboarded, or anything. Just a hodgepodge of ideas I had while at my desk at work. Feel free to fix it up.

@#$% a duck it's upside down.

[pinkjimiphoton]

4017...and get 8 'shades' of fuzziness.

choppy or smooth...

How do you make the transistion between the outputs smooth?

part of the question i've been asking. i would love to see a choice... smooth transitions should sound phasey, abrupt like a sample and hold effect i'd imagine.

4017...and get 8 'shades' of fuzziness.

choppy or smooth...

How do you make the transistion between the outputs smooth?

Well, here is perhaps where it might get kinda weird or complicated.

One possible route to go is to use a quadrature LFO, where you have four triangle or possibly ramp outputs, with output B 90 degrees out from A, C is 180 degrees out and D 270 degrees out.  The outputs control the gates of JFETs set up as part of an attenuator network.  That is, you have four different filtered versions of the basic fuzz, each with its own resistor/JFET attenuator (not all that different from the input of an Orange Squeezer), and feeding a 4-input mixer.  Each of the differently filtered outputs becomes prominent for its own litte chunk of time, while the others fade in and fade out, like a revolving door where you can "see" everybody going around, but one of the people really sticks out more at any given time.  As always, there's a relevant scan here from POLYPHONY, courtesy of Thomas Henry: http://hammer.ampage.org/files/quad-shep.pdf

mark, this is just about nailing it i think... and of course the scan of polyphony is gonna be a must read.

i hope i can comprehend a little of it.

Mark - I was just looking at Quad LFO's, but finding a schematic is tough! We must have been having similar thoughts at similar times. Thanks for linking to that scan - but I can't get it to load on neither my computer or my phone. Does it work for anyone else?

I was having a look at phase shifters using inductors, but my brain exploded.

As far using using 4 LFOs at 90 degrees separation - I hope we can get a schematic! The LFOs can drive LED's that could blend in the four fuzz circuits. I can't imagine more than 4 fuzz's initially! Might as well use a phaser and a fuzz in one if you want a bunch of them, no?

Anyway, if the period of the LFO was set to 4 seconds (for ease of this demonstration), the fuzz levels would be like this (numbers 0, 50 and 100 are % volume):

Time:         0     1     2     3     4(0)
Fuzz   1  -  100  50    0    50   100
         2   -  50   100  50   0      50
         3   -  50    0    50  100    50
         4   -   0    50  100  50     0

If fuzz 1 - 4 were set up with the LFOs at 0, 90, 180 and 270 phases. Or something like that. So at each 25% of the period, one fuzz would be full on, two would be at 50% volume and one would be off. That would cycle through each 25% of the sweep until we end back up again at the beginning (0/4).

That seems really cool.

this sounds like it should work.... very cool!!!!

Well, here's my idea. U1 = 13700, U2 and 3 = OPA2134 (possibly other op amps?)



The top left is the quad LFO that is based on a schematic posted by frijitz on electro-music.com forum. Sine and cosine outputs that are synthed and inverted to get the 0, 90, 180 and 270 phase shifts of the same LFO. The rate is controlled by the amps the 13700 sees. The four outputs drive four independant bases of transistors that fade LED's in and out. Perhaps the LED's could be driven from phase outputs? I'm not sure what kind of power they put out. That would make it slightly more simple.

The audio section is just the AMZ splitter four times to four different fuzzs. Between the splitter and the fuzz input is an LDR much like Taylor's Tap Trem to control the volume. As the LFO makes its phase cycles, the LDRs open up in sequence which allows signal to the fuzzes. The fuzz circuits are then mixed simply by mixing resistors. Could be buffered mixing, not sure. You'd also have to make sure that the fuzz circuit it self does not invert the phase or that could lead to mixing problems.

This is not tested, breadboarded, or anything. Just a hodgepodge of ideas I had while at my desk at work. Feel free to fix it up.

@#$% a duck it's upside down.

i am gonna take a look at this and see if i can make it fly somehow.. thanks!!! i have the parts i think... worth a breadboard, right?
soon as i'm done with the project i'm knee deep in, i'll take a crack at it.

thank you all!!

[Blitz Krieg]

maybe something along these lines?

Also, if you think about how pads work, you can put a fixed series arm and several shunt arms to ground. You can switch in shunt arms in parallel on all stages at once to change the padding constant per stage, so you can change the clipping and sustain on the fly if you want - even modulate it with a variable resistance. Imagine a six stage with the pads connected to LDR's to ground. The LDR's are clustered around a light bulb that is either LFO'ed or envelope controlled, so that the sustain/tone compromise varies within a note...

It's interesting to think that as you hit a note, all the stages in a muiltistage clipper are driven into clip. As the note decays, the stages nearest the input come out of clip one by one until only the last is clipped, and it's only then that the tone changes. You're free to mess with the EQ between levels, of course.

http://www.diystompboxes.com/smfforum/index.php?topic=14915.0

[R.G.]

A morpher is one of my personal favorite hairbrained ideas. I've spent many hours designing this while I was trapped in some meeting or waiting room with nothing else to do.

Getting a smooth fade for more than two signals gets complicated. The joystick is a great choice, so is the rotating lampshade and light/LDR setup. I came up with two more complex ways - remember, anything worth doing is worth overdoing. 

Getting a fade in and out of two of N channels is tricky. I postulated, but never programmed, that a uC could use ultrasonic PWM to PWM in/out signals on multiple channels, switching them directly. I know this would work, as it's a variant of the switched-capacitor filter. But tricky to do well.

The other involved voltage controlled amplifiers.  Imagine a voltage controlled attenuator where 0V is "off" and 5V is unity gain. These are common in the synth world, and exist in the pedal world in many tremolos. The Pulsar tremolo in particular uses a resistor + NPN transistor as a voltage controlled attenuator. Some JFETs will do this as well.

To get a voltage controlled fade in/out, all you have to do is make the control voltage fade up and down. No problem - we have capacitors! So a triple or quad - or octal! - morpher can be done with some logic to make 0-5V signals for the controls, and a resistor-capacitor per channel to slow the fade. If you use a CMOS one-of-N decoded counter like the 4017, then only one channel is on at a time. When you clock the counter, the "1" enable moves to the next one, and turns off the current one, leaving the others off. And we can take care of how fast the "move to the next one" happens by manually ("pedally"?) clocking the counter, or using a digital oscillator to tick it every so often.

But we'd also like to vary how fast the fades happen. And better yet, we'd like to vary how fast they fade up and how fast they fade down independently. Well, OK, I would. That gets tricky. I designed several approaches to this, including NPN-PNP opposing current mirrors with master control currents and .... AGHH! It got complicated. Then I remembered that OTAs are current output devices, and that they output a current proportional to both the Iabc current, and the difference of the + and - input voltage. Ding!  One OTA per modulating voltage, with the logic on/off signals fed to controls for how fast up/down, and the whole mess fed by a master ramp speed control in the bias pin for all the OTAs.

But I'm wandering.

To get the phasing effect, put a FIXED phase shift stage after each fuzz. Then as you fade between fuzzes, the phase changes whether or not your fuzz was helping.

[pinkjimiphoton]

bottom line is the moneymaker....

so maybe a fuzz built with a cmos chip like a 40 or 69 or a quad transistor array (or two) followed by a stationary phaser stage with like, maybe a quad transconductance opamp to keep the parts low?

then maybe use the led/ldr trick to make an lfor sweep between the stages?

sorry rg... i thought i kinda understood, but this is still above my paygrade.

wouldn't it sound more like a filter sample hold effect in a case like this?  i am sooooooooooooooooo lost...

gonna lurk and hopefully learn.

[bilo01]

Interesting thought. Don't know if anything like this exisis and i sure won't be any help on the technical aspects because i don't know much about this great hobby here....Yet!!!
I just mean it's addictive lol.
So you mean a central pedal that interferes with all of the 4 fuzzes. Let's call it the Quadro. A central mixer-pedal with which you can adjust en select the output level of each fuzz.
Isn't that a A, B, C, D-selector? Or am i thinking too simple ( as usual lol )
Will

[pinkjimiphoton]

hi will,
nah, more like a phase shifted fuzz sorta. but with the sweeping modulation sweeping between different sounds.

ayyyyyy my head hurts...