Heavy metal with xors

[zipzap]

sounds good. Tell us if you find something cool. Like unkle k said, a higher cab value should work, since we´re dealing with smaaler sizes when using large pots. I had a 20n cab with a 1meg pot. So you should be getting the same result with a 2uf cab and the 10k pots.
I think that´s still not too big. something like this won´t work with an adsr, where you would need 1000suf and get a lot of leakage, but here it might be ok.
I also noticed the pitch somehow going down with lower voltage, but isn´t the time constant of an rc network independant from voltage?
Maybe it has to do with the squares being messed and the xor being unable to follow.
Another thing is that, by the nature of the xor, higher osc freq does not automatically mean higher freq in the end. By deviding it down some octaves i could imagine that you end up with a pretty even, constant spike output. I just thought maybe a counter and a r2r network could give some interesting textures. Like clocking a sequencer at audio rate.

[Uncle Krunkus]

[Scott Stites]

'Tis. RC time constant is simply R*C - the time it takes a capacitor to charge up to 63% of full charge or discharge to 37% of full charge.

Notice V or I are not a factor in the equation. A lower voltage will mean a lower current...but, well, lower voltage - the cap doesn't have as far to go.

This is what makes expo portamento applied to CV do its thang - same concept. If you play a low A and a high A, it takes the same amount of time to glide as it does if you go from low A to low C. As an aside, linear glide will take a longer time to go to the high A then the low C, but there's usually an op amp or two involved there.

Cheers,
Scott

[fldrvr]

So... after the exors you would still need some more parts (VCA etc.) after this.

If I may be so boneheaded (I am really new to electronics and sound generation/synthesis)... what's a VCA? Just point me in the direction of some info, this thread isn't really for that question....

Hand-trigger? Decay envelope? (sheepish look) what?

Thanks all...

[Mikmo]

A VCA is a Voltage Controlled Amplifier. In a modular synth it is usually controlled by an envelope generator, to crate variation in loudness over time. But it also has other applications.

If you Google "VCA synth DIY" or something like that you should find all the info you could ever want (and then a bit)

[Mikmo]

Uups, forgot half the answer.

You do not need a VCA after the XOR's, i run mine directly into a mixer line in.

But if you have a modular synth you could run the signal after the XOR's into, say first a VCF (Voltage Controlled Filter), then to a VCA controlled by an envelope generator to get Attack Decay Sustain and Release. This would give you a lot of added control over the sound, and allow you to incorporate it with your synths sound.

[Scott Stites]

It's particularly effective when used with a VCA and a VCF. These types of circuits are used a lot in 'pitched noise' circuits. Pitched noise is a particular type of noise that has a general pitch to it - Thomas Henry used the example of the sound of a vacuum cleaner, which is noise, but has pitch as well. Pitched noise is often used in metallic percussive effects - anything from the crash of a cymbal, the bang of a gong to someone beating on a metal trash can.

If you're curious about the effect, I worked with such a circuit a while back and recorded a few samples. On my old site here:

http://mypeoplepc.com/members/scottnoanh/slsdiy/id18.html

Scroll down to:

11 April, 2004

To hear a gong-like effect from such a circuit, download the sample "Bang the Gong" - it contains a gong sound, but also a lot of other filtered effects. On my circuit, I used optocouplers to control each oscillator from a single control circuit. The intial frequency of each oscillator was tuned with a pot, and the control would move all of the oscillators in tandem. I used VTL5C2's, which are slow responding Vactrols (had plenty of them at the time). It's really an effective little circuit to have in a modular - besides percussive sounds, it can create some fairly unique 'outer space' type sounds in conjunction with a VCF and a VCA with very little effort.

Best Regards,
Scott

[zipzap]

yes. and decay envelope means you have a little circuit. It´s hit by a voltage "gate" pulse or by hand (switch). Then the envelope circuit generates a controll voltage. You know everything in a synth is voltage controlled. It starts at high voltage and decays down to 0. The vca will respond to that, because it is a voltage controlled amp. high controll voltage means loud, 0 voltage means no sound.
So with that envelope as a controll voltage source and any sound source at the (audio) input of the vca you get a single note that starts loud and constantly fades away, like a piano note. That happens every time the envelope circuit is triggered - hit by voltage.
It turns the exor noise into a gong.
You can have many types of envelope. google adsr.

[v-un-v]

essentially, the XOR circuit is the same as the psuedo ring-mod out of the MS20

I posted a link to an interesting design some time back that was able to process other waveforms- but I suspect there was some sort of comparator circuit in there to change them into a squarewave.

can't find that link at the moment. I'll try and dig it up in the next couple of days

Does anyone know if the 8038 (?) is still available? I think it was the 8038?. It was a function gen with very low distortion on the sine wave and had a very wide sweep. Four of these with four adsr's all interlinked could reap some real FM damage

[Uncle Krunkus]

You could also get interesting effects by using 1*XOR run by 2*SQR Oscs switching 1/4 of a 4066 which is turning on and off another waveform, such as a sine or triangle. Multiply that by four and feed it with SQR/TRI/SAW/SIN, then mix all four outputs.
By cross connecting the XOR inputs you could make it so that certain combinations exclude others, etc.
Wow, if I wasn't so busy I'd jump straight on my breadboard now!

[zipzap]

now that´s a cool idea. you could have a special module dedicated for thet task, or use this circuit, maybe the output of the first xor, and use a sequential swith with it - 4017, rotary switch for sequence length, pots and cv out if you like (since it´s kind of a waste of circuitry not to do it), and one or two 4066s swithed by the 4017. clock in could be the xor, any square source, a vco, the master clock. Nice also that the 4066 works in both ways.
just one thing: i guess the 4066 should be powered by +- rails instead of + and ground, so it can switch bipolar signals. than the clock must be feed into an comperator to meet the logic levels of a cmos ic powered that way.
I´ve got about 20 modules to go till i´ll be done , this is one of them!
about the 8083. I was looking for it once, i don´t know where i found it, but i remember that the xr2206 was cheaper and better. It can produce sine, tri, square and saw. As Henry and Scott have shown us it can produce other very cool vc-waveforms, can be synced, and can follow 1v/oct. It also has a bipolar AM in for instant ringmod sounds. Better choice maybe. I remember seing the shematic you once posted and i´m sure it just squared up any waveform coming in.
What i´m not sure about is if what you call pseudo ringmod is the same or something different as real ringmod (mathematically).
It´s not achiefed by bipolar am like with an ota or 633, but maybe the result of xor-ing two squarewaves is the same as sending them into a real ringmod?

[Scott Stites]

v-un-v,

ICL8038 is no longer in production by Intersil, the XR8038 looks to still be manufactured by Exar. If you run it like the datasheet instructs you to, it's a very inefficient lump of humus, IMO.

If, however, Thomas Henry had written the datasheet, I think they would have sold loads more. The trick is in how it's powered. If you power it between ground (as the positive supply) and -V as the ground connection, it is quite a sweet little chip. Not as great IMO as the XR2206 for some things, but it does have its place in the grand scheme of things. I can tell you it found its way into Thomas' unreleased new SuperController.

You and Zizap have, in fact, inspired me two-fold. This whole concept is pretty nice, but essentially it's too much of a one-trick-pony for me to prioritize it into my modular until I've had my fill of other things. Don't get me wrong, it's a worthy function, it's just that I have precious little space and I like to get more bang for panel space - that's my Sergian side.

The main limiting factor for me is that it really is pseudo-ring modulation, and is only fit for square wave inputs. Even running other signals into it through the use of a comparator essentially gets you the same thing - square wave pseudo ring modulation. My personal ring modulation tastes run a far greater gamut than pulse waves, in fact I generally prefer the 'smoother' less harmonically rich waves for ring modulation (I.E. triangle and sine).

Fold One - the Lesser Fold

In fact, it's not the best ringmodulator in the world, but, for an app like this, it doesn't necessarily have to be: Why not use the XR2206 idea, cut out the middle man and use the internal modulator on each XR2206 for the ring modulation, with four of them interconnected for general bizarrity? Gone is the Vactrol control, and now you have the advantage of using non-hard edged waves to produce your ring modulation. You could even put the expo control on it (like the XR2206 VCO) but not waste a tempco or LM394 - a couple of 2N3904's and a normal 2K resistor would work fine. The object of such a circuit is not frequency accuracy, but the general tone you get out of it. The expo converter will give you a very wide, more controllable range of frequency than the stock linear control. This would also provide four VCO outputs, which, admittedly would not be terribly accurate, but sometimes that's not a requirement. Put in Bugrand's LFO mod, and you would also have four VCLFO's. Scott's rule number 52 - you can *never* have too many LFOs.

In addition to ring modulation, the module now is quite capable of providing FM effects as well - other than an actual bell, nothing makes a better bell than FM. Let ring modulation give you the metal grunge and glassy glissando, and FM will toll for whomever. For dissonance, ring modulate your bells with other bells, or a pure sine. The original, Vactrol controlled idea will provide only limited FM, and pulse wave FM at that (FM is another area I find myself much more often using non-harmonically rich waves rather than pulses and sawtooths).


Fold 2

Use either XR2206 or 8038 and use the AD MLT04. The MLT04 is essentially four analog ring modulators in a single IC. I've got a couple that I ordered as samples thinking they would be pretty cool to play with, and never did. This would be a very cool app for the MLT04. With that, you could have the four internal VCO's normalled to the inputs of the four ring modulators, and you could also patch external signals into it. I doubt if it would be as nice as an AD633, but the point of this concept has never been accuracy. Save the AD633 for your precision ring modulator.

Either of these ideas takes the concept to a new level - they provide, yet break free of, ring modulation of pulse waves, and provide additional signal outputs as well (as opposed to just pulse waves). Put a dash of Bugbrand in there, and you also get free VCLFO's for your dime. And then, you now have the panoply of FM at your fingertips.

Downside is, the original concept is a very efficient circuit; essentially a two IC job. That, and none of the chips mentioned above are terribly inexpensive (unless you can get samples). Seeing as how I've already got the required parts, it's not a big deal for me to build either one. Fold 2 of the two-fold sounds more appealing to me.

One of these days.

Cheers,
Scott "OOTD" Stites

[Scott Stites]

Oh, yes, nearly forgot - if you want to incorporate an EG, though its not ADSR, it doesn't get much more efficient than this design from the inestimable Ian Fritz:

http://home.earthlink.net/~ijfritz/sy_circ.htm

Scroll down to AD/AR Envelope Generator

For percussive stuff this is more than enough. In fact, I got through two years with only an AD/AR of Ray Wilson's design. They're more useful than one generally assumes, especially if one has a CV mixer and/or a peak/trough function.

Cheers,
Scott

[zipzap]

cool, thats a link i didn´t know. Nice stuff there!
Big post, scott, nothing to add, total agreement.
What you say about that segian side is becoming more and more important as my synth is growing. I loved the am input of the xr2206 when playing with it (and watching the scope), i´m now thinking of getting it into the xr2206-vco design. have a swith for either "normal" function and for the am-ringmod.
I like those multi purpose modules. For me it´s a good way. Especially when going into sound adventures a very capable lfo may be worth a lot in one situation, whereas in another situation...
What i loved about the xor circuit is how easy you get very strange sounds and patterns. If you´re getting started with diy synth and already got a vca and some stuff, build yourself one of these!

[Scott Stites]

Hi Zipzap,

Yes - that's true, my idea was more along the lines of getting audio out and didn't cover yours and Uncle K's discussion - I was still stuck in the audio mode.

Multiple functionality is the lifeblood of my synth. Whenever the opportunity arises, I'll add that one extra connector, switch or function that can allow the module to be used even if it's just a bit part in a patch. Man, those things do come in handy. For example, my 2040 filter not only has three output jacks for the signal, but also three for the mixer section that precedes the input. Then there are three more that provide the inverted output of the same. Because I have mutliple filters, I may not be using that module as a filter, but then it serves time as a three input mixer for either CV's or audio, and also serves as an inverter.

My most extravagent multi-functional module is the one that is *always* in use - my Mixer Model 3, which can handle either audio or CV signals. It's set up so that it can be two four input mixers, or a mixer that can handle up to eight inputs, the mix of up to four inputs can be subtracted from a mix of up to four inputs, it will produce a peak, trough, inverted peak, or inverted trough copy of a the mix of up to four inputs, it has a built-in DC coupled multiplier (AD633 ring modulator), the third control of each four input mixer is normalled so that it can provide a positive offset, the fourth input of each mixer is normalled so that it can provide a negative dc offset, the first control of each mixer is normalled so that it will add the mix of the opposite mixer and the second control of each mixer is normalled so that it will subtract the mix of the opposite mixer. The mixer sections have inverted and non-inverted outputs. The ring modulator inputs are normalled to the mixed outputs of each four input mixer. The peak and trough functions are unique in that they provide the peak or trough functions of up to four signals as they're set by their respective input pots (that took some doing if one thinks about how a normal peak and trough is to function).

Peak and trough are worth their weight in gold, particularly in this function. Picture a normal mixer, and you're mixing a sine wave with a ramp wave. The waveform itself changes in amplitude and shape as you mix in more of one signal or another. With the peak or trough output, the amplitude is always the amplitude of the highest signal, no higher or lower. The effect is like watching one superimpose one signal over the other rather than mixing them together. In this example, a sine wave of a different frequency will 'slide' behind the sawtooth rather than affecting the sawtooth. It's great for many functions - I've used it for rectification, creating hypertriangular waves, mixing EG's (really great for that kind of stuff), and producing some very exotic CV's.

Anyway, this module produces so many functions, it's a rare patch that doesn't use it. It's an adder, subtractor, multiplier, inverter, DC source, and peak/trough processor all built into one 1U wide module. It can perform all of these functions simultaneously. This is model 3, the other models are waiting in the wings, with their own unique Swiss Army Knife functions.

As for the MLT04 - I studied up a bit more on it - forgot the fact it's limited to +/-5V supply voltage, that complicates things a tad, but it's not a deal breaker.

Take care,
Scott

[Mikmo]

So i tried to swap the 10K pots in the joysticks with 1M ordinary pots, and use smaller caps.

With small pots and large caps the output signal is much more noisy, like there's a backdrop of white noise.

The RC constant should be approx. the same, but the noise is different. My lack of serious electronics knowledge prevents me from explaining this, maybe some of you can ?

I like the "large pots / small caps" sound better, so the joysticks are saved for my modular synth.

[blue hell]

[v-un-v]

[Scott Stites]

[Scott Stites]

[zipzap]

That´s what i thought. In my case there are many settings where the output sounds pretty much like white noise.
About the swiss army knife, yes that´s the way to go! I´m thinking of going in that direction for my next mixer. Could even have vcas on each channal and individual outs. only problem is that it needs one more pot and jack for each chanall. A switch could send the vca out to the summing amp - or not. Than it´s just another vca.
I´m currently designing another frame to hold some 16,5cm*50cm of modules.
It´ll be my soundsource frame. The more i built the more functionallity and panel design concerns me. Guess we´ve all made that experiance.
So this will hold 3 oscs, my bus-controll module, noise and a mixer. Propably won´t get much more in there, i´m for good functionality but not for tiny controlls and crowded front panels. Maybe i´ll get a min-max and a ringmod in there. That mixer will have the oscs hard wired + some extra inputs. I´m planning two external inputs to go to my mixing desc as well. They´ll get evelope followers.
So that´s quite complex with many controlls. In a module that big it´s more than ever a good thing to make it flexible, to meet most possible usages. Have to think of a simple yet efficient submix system. A matrix mixer is too big, but each channal could have a little button to send it to maybe two different outs. I wish there were normalled banans.
Even the simpelest 4 chanal mixer can get a lot more interseting if it gets 3 outputs: 1+-2, 3+-4 and 1+-2+-3+-4. Banana jacks are small.
Often it´s only 2 signals you want to mix.
On the other hand i´ve got a mixer that has only one pot, a switch and 3 jacks. 2 inputs. 1 has unity gain, no controll, the other input can either be added or substracted from that fixed input. I find that quite usefull, so this would be an example of an module that is very specialiced. It can only do that one thing, but it´s build in 5 minutes and takes no space at all.
My moog ladder also has a build in cv mixer and a adsr. no cables needed. But there is a cv and a env out. Often i don´t use that filter and am happy to have another mixer.

[v-un-v]

[Uncle Krunkus]

Hey, look what I found!
Does the left hand side of this circuit remind you of something? It's the schem for the Synbal from the early 80's. Just an idea to keep in mind, as most of us who actually get around to playing with this idea will be using a hex trigger chip and a quad XOR.

---

SYNBAL.gif
Description:
Filesize:	76.84 KB
Viewed:	322 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[v-un-v]

heh! that's funny because I kept thinking of exactly this circuit all the way through this thread!

Electronics and Music Maker (E&MM) me thinks

It's nice to see that cmos cluster once more.

[Uncle Krunkus]

I got the schem for the Syntom as well. I'm really impressed with the look of this. I was thinking about re-inventing the DS7/8 but with two types of noise to mix together, and a SV filter this thing looks even more promising. Maybe I'll build one of each.