Tuesday, January 29, 2013

I've Been Busy...

Seeing as how I have a fair amount of time on my hands, I've decided to redesign my modules. Yes, almost all of them.
List of mechanical changes:
-made knobs smaller, two sizes
-met Eurorack size standards
-made all modules either 4, 8, or 12hp (except sequencer, still 84hp, which is the typical width for Euro modules)
-made switches smaller (except on sequencer)
-made all jacks 1/8" (except on Out module)

New modules:
Filter: this is a very complex VCMMF (voltage-controlled multi-mode filter). It uses both lowpass and highpass 12dB filters and special audio routing to produce interesting timbres. Each filter has switchable resonance (yes, these can be totally non-resonant) and feedback amount, as well as attenuated voltage control. There is also a master CV input (control both filters with same CV) and single audio input. Outputs include lowpass, switchable highpass/bandpass, and a crossfader between the LP and HP/BP outputs. BP can also be BR, depending on how the filters are interacting.
M-Mults: 6 sets of 6 mults, and a set of 3 mults. If you need more... eek.
Tone: passive lowpass into a highpass, with a crossfaded diode clipper. Merely gives you some tonal control.
MTS: 3 sets of 4 mults. In case you do need more.
Out: a simplified version of Pittsburgh Modular's Outs module, has left and right inputs (if nothing is in the right input, it uses left signal), TRS 1/4" and two TS 1/4" outputs. Bidirectional, if you want.
Shaper II: this will be a Buchla-based complex waveshaper, assuming I don't build it into the redesigned Oscillator. Speaking of which...

Redesigned modules:
Oscillator: I'm completely redesigning my oscillator. I don't like the old one. What you see here is probably not what it will be, but it will be inspired by Make Noise's DPO.
Ringmod: All I did here was add input attenuators.
VCA: given the new space using smaller parts, I put two VCAs in one 8hp module :) Same thing electrically.
SNH/ASNH: I simply combined my Sample and Hold and Arbitrary Sample and Hold modules in one.
CV: Yes, revision 4 of my CV Tools module, but I think this one is good. 2 inverters, 1/4-1/2-3/4 voltage divider, 4-in averager, two comparators, REAL offset (created using a slew limiter into a comparator referencing 8V), and a 2-in summer. I think it has enough stuff in it now :)
Mixer: didn't add much here, just made each atteunator have a separate output so you can use them just as attenuators. Yes, I based this idea off of Pittsburgh Modular's mixer.
Sequencer: Didn't change much here, just added a summer circuit with one input being the master CV out. Input a CV, and the pitch will be transposed. For example, if the sequencer is outputting 2V and you add a 3V input, it will output 5V.

Modules to come:
LFO: Simply a low frequency oscillator. Probably will have square, sine, triangle, switchable saw/ramp (or maybe variable waveform saw, depending on the design I go for. Maybe both!), and maybe built-in S&H using either another waveform or noise. Still working on it, but very close.
ADSR: What synth is complete without an envelope generator? I'm thinking of using a modified Roland 100m ADSR (module A-140, no relation to Doepfer), modified in such a way of course that it's totally my own. Kinda hard to do, but I think I know how to keep the snappiness of it while also letting it have up to 20s of time between sections. Should be fun...
Logic: I was thinking that a few logic gates would be useful, but definitely a last priority.
FM processor: interesting module which will allow you to input any two signals and it will use one input to modulate the other. Not really sure how that will work, though.
Formant-y Filter: this will be 4 individually tunable high resonance bandpass filters with an attenuated master CV control and master manual control. When swept, it will sound either like multiple wah pedals or kind of like a formant filter. Either way it'll be interesting.
TipTop Audio uZeus: Yes, I'll be using a premade PSU once I get stuff built. I like the uZeus' flexibility (not speaking just literally here), and given that I can probably make ribbon cables for bus>board connections, I'd say I'm pretty much set with the uZeus.

So yeah, I'm having all sorts of fun over here. And, I know that each module works because I test them before posting about them. Should be interesting to actually put them all together and play it!

Saturday, January 26, 2013

New CV Processor Module!

Vy hello there. I came up with a neat idea which has probably been done before! As has probably everything on this blog...

No matter, this is useful. It's a 4-function Control Voltage processor! My CV processor has gone through three revisions now. The first was bad because it had an offset section, but no control over it. The second was bad because who really needs a comparator and a clipper? So now I present you with revision 3: beefier, more useful, and somewhat unique!
It has: 2 inverters, 1 4-input Averager, 1 3-way voltage divider, and 1 2-input voltage summer. The inverters do just that: invert a positive signal to negative, and vice versa. The Averager is quite simple: input 2-4 voltages, and the output will be the mathematical average of the 4. The Divider has separate outputs for 1/4, 1/2, and 3/4 of the input voltage. The Summer adds two voltages together.

Maybe I can find a way to have a comparator in there. The divider section does take up more room than it needs, but I was thinking of labeling each output too. Hmm...

Beware, for my mind is fleeting; you'll see a lot of new modules in the coming days. It'll be like NAMM! But just one company doing one thing! And actually after WNAMM! Hey, maybe by SNAMM 2013 I'll be able to actually go there as a company, eh? Be all professional n stuff, y'know. All the while blogging about exactly how I make everything...

Tuesday, January 22, 2013

Analogue Bitcrusher

There are those who say that bitcrushers are digital-only effects. These people are fools, I say.

Though it is true that bits are digital, and removing bits is also digital, you can get the exact same effect using... *drumroll* Sample and Hold!! Yes folks, the standard sample and hold function on pretty much any good analogue synth (Arp 2600, Moog Modular, ANY modular, etc.) can be used as an analogue bitcrusher-like effect, given you drive the clock signal into the audio range. Here's how:

First, grab yourself a voltage controlled switch and a capacitor. If you don't have a VCS, don't worry: transistors can be switches, too :)

Pretty simple, you've got an input, a FET, a clock signal, and a capacitor. The 1M resistor and diode are there purely for stability, and not actually required (in fact, leave them out for some interesting effects). How it works is pretty simple, too:

When the clock signal is low, the FET turns off, and the signal passes through to the capacitor, which given its voltage storing power actually holds the voltage at that point. When the clock signal goes high, the FET turns on and the signal is passed normally.

Due to this, however, if you have a 50% duty cycle clock signal, only half of the time will your signal be held (again, interesting effect). Thus, you need a clock signal with a duty cycle of 10% or less to get a sample and hold effect and keep it this simple. This can be done with, of course, the 555.

So far, we just have a standard Sample and Hold circuit. Now, it's time to bitcrush!!

There are three main parts to a bitcrusher: downsampling, jitter, and dither. Downsampling can be done simply by changing the frequency of your clock signal. To get bitcrushing sounds, you'll need to operate the clock in the audio range. This will create a waveform very much like a normal sample and hold function.

Jitter is actually very simple, and if you listen to modern dubstep (or brostep) you have heard this effect in talking growl basses. These sounds are formed using a bitcrusher, with the samplerate jittered with the filter's LFO; that is, the samplerate (frequency of the clock signal) is modulated by an LFO. Simple enough.

Dither is generally just smoothing the signal. This can be done one of three ways, the first being the most common: noise mix, portamento, and filtering. Adding noise adds a certain bit of normalled grit to the sound, portamento smooths out the edges of the output sample and hold, and filtering smooths the corners of the output sample and hold.

And there you have it, a fully functional analogue bitcrusher. Pfft, who needs digital bitcrushing? LOL

Monday, January 21, 2013

Neat Synth Tricks using a 555

I should probably start by saying that this isn't so much about audio purposes, but how a 555 can be used for all sorts of neat stuff for controlling analogue synths and their sound.

First, an oscillator:

How it works is pretty simple: the cap C1 (can be any value, as can all the components shown) gets charged and discharged in a motion similar to that of a saw wave. Physics lesson over.
Not that that was much of a physics lesson. ANYWAY, so what can this be used for? The saw is the same frequency as the square wave, so it can't be used for fat analogue bass sounds, or -- or, can it?
The neat thing about a 555 is that it can be used as a frequency divider. Pull up the datasheet for a 555 and you'll see. If you input this oscillator's square wave into a monostable 555 setup (which outputs pulses on trigger) whose timing cycle (frequency) is set just right, you can make a divide by two, three or more circuit. Actual values would vary depending on your oscillators. Output the monostable's saw wave (same deal as above) and you have a saw below the first oscillator.
More typically, you'd output the above oscillator's saw wave, then use the lower square wave to fatten up the sound. Think Moog bass, but about $999 cheaper. More or less. This is musical, too: say you are playing an A at 440 Hz, the other oscillator will be at 220Hz with a divide by two circuit. That is an exact A, one octave lower. You can use this for sub-oscillators, if that's your thing.
Speaking of sub-oscillators, if you make one I recommend a switching system. You can feed your signal to the timing resistor into a multi-way switch. Each switch point has a certain size resistor on it. Switching between these resistors (after you calculate their values) will allow you to create a switchable-octave sub-oscillator. Not many instruments have this, so it would be an interesting feature.
Neatly enough, this same monostable circuit can be used as your trigger signal from your analogue keyboard. First, set up a comparator circuit to compare the keyboard signal to ground. When you play a note, the voltage will go high, thus creating a gate signal. If your gate/trigger needs are met with this, great! If not...
You can then input this gate signal into a monostable 555, and get a trigger output! Now you have analogue CV (keyboard voltage via the standard resistor chain), gate (comparator), and trigger signals. And, given that we know that those are the three typical signals in an analogue synth, we now have full control.
If you've been reading this blog for long, you also know that I use 555's as normal audio oscillators. This is great if you need saw, triangle, and square/pulse waves, but what if you need more interesting waves?
Yep, a 555 can help you there, too. Used in typical oscillator fashion (as above), it can be used as a timer (which is actually the intended use of a 555). Here is where we delve into lifehacking and such fun.
Grab two 555 oscillators and an N-channel JFET (quite common). Run the first oscillator into the Gate input of the FET. Then, wire the Drain output of the FET after the timing resistor of the second oscillator. Wire Source to ground. Voila, you suddenly have a (rather shoddy) FM oscillator! The first is modulating the second. Be sure to monitor the FET. They like to get hot when you externally switch them a lot.
Or, secondarily, you can take the same two oscillators and wire the outputs in series along a wire, to the output. I guarantee you will get some neat sounds. This is the same premise as my 'HardMix' module.
Heck, combine these two to get just totally freaky. And, well, noisy. Oh, and of course, feel free to use the saw outputs in these situations as well. Very interesting stuff.
I know these are somewhat basic 555 functions, but I've been playing around with them lately. THEY'RE FUN, OKAY??

Sunday, January 13, 2013

XA-80 Patch Notes

Thanks to Marc Doty, I'm going to make a few changes to my XA-80 plan. I'll also tell you a bit about its progress so far.

As for things I'm not probably totally scrapping:-Brilliance: I was misunderstanding the brilliance control on the CS-80. As it turns out, that's actually a secondary resonant low pass filter after the two voices mix together. So, I'll be adding that.
-IL and AL controls: I was right about how these affect the amount and voltage start point of the filter ADR, however I will still not be using IL. AL is the attenuator I was talking about.

2nd Voice Detuning, Octave/Footing Controls, Keyboard Controls, Touch Response, Footswitch/pedal Controls, and Presets are still being omitted. That seems like a large list...

Things I'm adding:
-MIDI control: yes, I'm making it MIDI controlled (In and Thru). I'll explain why and how later.
-CV outputs: Pitch CV, Gate, and Trigger will have outputs on the rear panel.
-Stereo audio out: just in case. This may also result in a panner, probably VC via the LFO.

External audio input is still in place.

So far, the instrument's plan is going well. And by well, I mean I'm slowly building it virtually. So far it has all the oscillators (except the noise oscillators), filters, amplifiers, attenuators, the ringmod, and current I/O planned.

However, there might be a few changes. I've been speaking with a longtime friend of Nyle Steiner (yes, the guy whose name is plastered on the Steiner-Parker modular, and more recently their filter) about the instrument and things I can do to make it affordable to build yet still playable and somewhat simple. He taught me two things: how to do the duophonic keyboard, and how to make a resonant filter.

As you probably know, I totally suck at building resonant filters. I also am bad at normal filters, so it's really no surprise. However, when he was telling me about the original Steiner filter, I noticed something about it: resonance was induced by increasing the gain on the non-inverting amplifier at the end of the signal chain! Though, when I tried it, I got weird phasing and stuff so I apparently need to work on it, but if this does in fact work then resonance will be a snap!

Now, about the keyboard... it's digital. OMG WHAT DUDE IT'S SUPPOSED TO BE FULLY ANALOGUE AND STU-- Yeah, I know. However, it requires a micro-computer (or a crap load of transistors n stuff) to accurately track two keypresses and send them to the oscillators and such. The XA-80 will actually use a MIDI keybed (with MIDI stuff still attached) and will feed a polyphonic MIDI-CV converter, which will then send out gate, trigger, and polyphonic pitch CVs to the rest of the (fully analogue) synthesizer. So really this will be a digitally played analogue synth. That's ok, right?

However, by using a MIDI keyboard, that means it's MIDI enabled, thus allowing for MIDI input, thus meaning MIDI sequencing is entirely possible!! However, it will only be for pitch, gate, and trigger. I'm not making the whole front panel MIDI-controlled. I'm not Moog :)

So, yeah, that's about it for now. Woot!

Monday, January 7, 2013

Xanu XA-80

Well hello there!

So, as you may know, I have a pseudo company called Xanu Guitars and Music. It's not totally legit, mainly because I don't have the tools to start it up. Sounds like I should head on over to Kickstarter.com, eh?

Anyway, along with guitars and audio equipment, Xanu does simple synths (the ones I blog about, like the Triple Fiver). Well, in my journey to building the KereMAX, I have started a new project: the XA-80!

If you don't know, the Yamaha CS-80 was essentially a total badass. It was beautiful-sounding, thick, warm, and analogue (and used by Vangelis! And heard in Blade Runner! Which Vangelis did music for! With his CS-80!), not to mention weighed like 200+lbs and was a pain to move and keep in tune. Oh, and nowadays it runs for around $14,000.

Enter the XA-80. It started as an idea to sort of 'recreate' the CS-80's controls and structure and its general sound, but on a more doable level for me and with some slight moderisations and helpful additions. And, before the thought even crosses your mind, no, I can't just sit down with components and faithfully recreate it. This will end up being a cheap hardware emulation that is somewhat similar to the CS-80 but actually a totally different instrument.

What's inside the XA-80:
-Control: the XA-80 will have a 4-octave duo-phonic (maybe) keyboard and an adjustable-range ribbon controller, both adjusting pitch. I'll also have an overall pitch adjustment.
-Modulation: I'm not THAT awesome, so for modulation the XA-80 will only have exponential portamento and tremolo affecting the primary VCAs.
-Voices: 2
-Oscillators: It will have 4 individually-tuned ultra-stable oscillators outputting saw and square waves and noise, but I'm thinking everything dealing with pulse width will have to be kinda scrapped. Either that or I develop/find a totally new oscillator that has square waves with VC adjustable pulse width and saw waves. Either way, it's two oscillators per voice.
-Filters: the XA-80 will feature two 12dB filters, one high-pass and one low-pass, for each voice. They may or may not be resonant, it all depends on if I can actually make a resonant filter or not. It will have an ADSR for control of the LPF (which is an improvement over the original CS-80).
-VCAs: the XA-80 will have two, like the original, except the second one will simply be an amplifier, not voltage-controlled. The first will have input and output levels, a sine wave attenuator (from the oscillator), and an ADSR, just like on the original VCA sections. The second is just an amplifier with a level control, like the CS-80's Output control.
-LFO: this multi-waveform oscillator was labelled sub-oscillator on the CS-80. Really, it was an LFO with attenuated outputs to the VCA, LPF, VCO, and had a separate speed control. The XA-80 will have the same thing, but wider ranges.
-Ringmod: This I'm completely changing as the original was really more of an FM oscillator (but still way cool). The XA-80's will be a transformer-and-diode-based ringmod with two attenuated inputs and an attenuated output. It will use the two individual voices as the inputs. I may also put a crossfader on it to cross between the RM and the normal voices.
-Voice mixer: just like the original, a crossfader.

As for things I'm probably totally scrapping:
-Brilliance: Just because Yamaha can make an analog velocity sensitive keyboard affect levels, doesn't mean I can. Just set the ADSR's D a little higher, same general effect :)
-IL and AL controls: they would affect the level and voltage of the filter's ADR, but I will simply have an attenuator on the ADSR.
-2nd voice detuning: given that I'm using individually-tuned oscillators, this is kinda pointless.
-Octave/Footing controls: I'm using very wide-range oscillators, so these are also pointless.
-Presets: Yeah, ok, so I'm scrapping one of the main features of why the CS-80 was awesome, but you know what? I'm not about to even think about how to do analogue presets. You want presets, go digital or expensive :) Besides, I feel a synthesizer should be played and used as a synthesizer, not an instrument that plays like a piano but with different sounds.

Things I'm adding:
-External Input: the XA-80 will have no external control (not like the CS-80's was very useful anyway), but it will have an audio input which uses the 2nd voice's VCF and VCA.
-Other previously mentioned items.

So overall I'm doing what I can, but there's really no way I can faithfully recreate the mighty CS-80. As you can see, though, I'm recreating a lot of what the CS-80 was, as well as adding some useful features, so I just really hope it turns out well. Obviously I'll blog about my progress.

Oh, and if you are a CS-80 or perfectionist fan, please don't hate on me for leaving out important features. I'm simply not capable of doing them yet. This is a totally different instrument anyway, it'll just be kinda similar.

Thursday, January 3, 2013

4017 Sequencer Test Build

So my LEDs and 4017s came in today. Thus, I also built my sequencer. As I had assumed, it had a few problems from the original design, so after Googling '4017 LED' and learning a few tricks I figured it out!

... Well, I was going to show you, but Blogger keeps not letting me upload pics. Basically all I did was clean up my original idea, make it make more sense, and ground the hold input. Yeah, it sucks, but this thing won't have a hold feature, which was going to be a neat element of it.

I also need to work on the CV attenuators, and fix the output voltage (it's 0-2V). And figure out the external reset. For ease I could just remove that, too, but then what fun is that? Then it's just a simple sequencer that isn't really fun or special at all! Well, actually it still is, and if I can amplify the output voltage I should be okay.

Ta-ta for now. Wait, who actually says ta-ta? Screw it, I'm bringing it back!

Wednesday, January 2, 2013

Saw-Core Oscillator n Stuff

Well, let's not get ahead of ourselves. First off, happy new year! I feel this is a good post to kick off the new year with. Also, did you see that I changed the blog page a bit? I just did a few small edits, nothing huge. Let's hold a contest: if you can name everything I changed (5 things), you'll get... I dunno. A Heart Container? haha

So, as I said I would do before the ACIII post, I built a saw-core oscillator. I didn't use the little idea I had posted, though I was going to. Then I came up with something else:
It's a little messy, but what really matters is the output. It outputs a near-perfect saw wave (ok, really a ramp, but the sound is similar), as well as a pulse output:
Obviously, changing the pulse width will make a more triangular wave.

But, here's where the 'n Stuff' part of the title comes in: I have a new module!!

As you probably know, I have a ton of fun playing with unpowered OpAmps. Maybe too much. But, today I used that, an NPN, and a few resistors to come up with what I'm calling an Arbitrary Sample and Hold.
Yes, yet another 555. That's the clock signal. The input is the random line going up, and you can also insert your own clock signal (can be any waveform, which creates neat effects). The module has a control for clock speed (which I might make voltage-controlled), a jack for inserting your own clock, and the bottom jack is actually the internal clock's output, should you need a synchronised output and/or square wave. The two middle jacks are input and output.

Now, you probably want to know what it does. Well, here ya go:
This is with the saw oscillator as an input and the 555 clock. The original input signal passes when the clock signal is low, at least in the positive voltage area. In the negative voltage area, it's the opposite effect. This is because during those times when the original signal passes, there is no secondary voltage to compare anything to. When the clock signal is high, something magical happens and the circuit sums the two voltages together. I seriously don't know enough about electronics to understand why this happens, but it's fancy.

As you can probably tell, this isn't really a summer, sample and hold, or even a gater. Or maybe it's a gated summer, I don't know. But, it is fancy, and so it's a new module now. I can see this being nifty for both audio and CV signals.

I just took another look, and if you don't have an input, it acts as a ramp oscillator:
However, I'm pretty sure those voltages are, well, probably realistic, but scary. An OpAmp can amplify 1000x, you know. 

Oh, and there's some more 'n Stuff' stuff.

So there's this dubstep producer, Bolshevik, maybe you've heard of him, who produces really interesting stuff. It's obviously dubsteppy, but has some major hip-hop and techno vibe. Only problem is he has problems with orchestration and composition, but he's getting better. Here's his latest track: https://soundcloud.com/#bolshevikofficial/execution

Hey, you know I love supporting other artists. In other news, I totally neglected to say what I got for Christmas! It's mostly electronics-related, but proves that quantity can easily be pwned by quality:
1: Korg MS-20 T-shirt with Japanese text (which I totally want to frame)
3: Elenco Hobbyist's kit, which has so much awesome stuff I can't list it all. Let's just say I could give my dad his tools back and got a lot more :)
4: (I actually got this later, but it's ok) small electronics utility table (handmade)
5: Elenco Adjustable Power Supply Kit, which I am currently building because I actually just got it today.

What can I say, Christmas can last a few days around here! Big thanks to my dad for, well, supporting my hobby :)

Oh, and I'm getting a TON of awesome components soon. My dad went on Mouser (my preferred dealer) and got me 5 4017's and 10 red LED's, so I can make (and hopefully not screw up) my sequencer, and adjust it as need be. Then today, I joined Texas Instrument's my.TI, which allows you to get free samples of their ICs (not even a shipping fee!). I got 3 quad opamps, 3 quad comparators, and 3 5-transistor arrays (two of those and you can build a Moog Ladder Filter, y'know). I could have gotten more but didn't want to intrude on my luck, and also didn't know what else to get. TI has a HUGE selection.

I also updated the Xanu website a bit: https://sites.google.com/site/xanuguitars/

Finally, I've uploaded a few new things to Soundcloud if you haven't been watching. Nothing incredible, but neat stuff: https://soundcloud.com/#yoa01.