Super Klee Sequencer

[Scott Stites]

Thanks, Uncle K. You're really right about that hole - this one is like a crater for me. I was honored to be invited to speak about Jeff's synthesizer and music yesterday at his memorial service. It was an amazing event - that's the way I'd like things to go when I kick off. It was truly a celebration of his life - I never saw a tear or a hanky. It was held in the chapel at KSU, where Jeff was an associate professor of statistics. The service consisted of his colleagues, family and friends speaking of the various aspects of his life they'd come to know and love. Truly uplifting, and true to form of the kind of guy Jeff was.

Take care,
Scott

[Scott Stites]

A few belated samples of the basic Model 2 Klee functionality.

A brief overview - this is a 1X16 shift register, with 16 switches to parallel load bit patterns. Right now there is no random input. Patterns are loaded either by sending a pulse to the 'load' input, hitting the load button, or by connecting a bit output to the stage load input (same sort of thing as a sequencer that can reset at a certain step for different lengths of sequences). Right now there are only four pots controlling the output voltage - one pot each on the 4th, 8th, 12th and 16th bit. In the final version, there will be 16 pots so this section can also act as a 'normal' sequencer.

This method of forming sequences I've come to refer to as 'found sequencing'. The process is that one sets the four pots for intervals, and then tries different bit patterns to explore the relationship that has been set up with the pots. The number and location of the selected bits can change the sequence *radically* or slightly. Using the stage load function creates even greater variation and makes it easier to synchronize pattern changes.

On sample 6, I set the pots up specifically for standard 'western' intervals by tuning each pot separately to a specific note, using my DW6000 as a reference. I let the sequence run, hit stage load to shorten the length, and while it's in stage load, I vary one switch position up and down before releasing stage load. This creates a nice little pattern within a pattern. 2.05 MB.

http://mypeoplepc.com/members/scottnoanh/sitebuildercontent/sitebuilderfiles/kleemod2_sample6.mp3

Sample 2 is just me running the circuit through its paces so that one can kind of get an idea what switching a few of the switches around does. It can change much more radically than what is heard, but I just stuck to a few switch changes and used the stage load to more or less sync the changes. I vary the clock and the tuning of the voices themselves (though not of the pots on the circuit) as the sample progresses. It gives a glimpse into 'finding' a sequence. 2.9 MB

http://mypeoplepc.com/members/scottnoanh/sitebuildercontent/sitebuilderfiles/kleemod2_sample2.mp3

Sample 3 is just a sequence with no pattern changes - just a little demo of how 'normal' the function can sound using four pots and multiple bits rather than a single bit like a 'normal' sequencer does. 1.7 MB

http://mypeoplepc.com/members/scottnoanh/sitebuildercontent/sitebuilderfiles/kleemod2_sample3.mp3

The circuit itself is quite simple, considering what it can do. I'll post it when I get a decent copy drawn up. Of course, it will expand, too.

Cheers,
Scott

Edit: Added file size info

[Scott Stites]

Some pics and a sample.

---

klee_patch.JPG
Description:	Pic of the Klee breadboard driving a KS-01 patch.
Filesize:	125.89 KB
Viewed:	783 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

klee_bb.JPG
Description:	The Model 2 Klee circuit on breadboard (with alligator clip position switches).
Filesize:	136.17 KB
Viewed:	791 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[Uncle Krunkus]

Cool!

[v-un-v]

Scott, perhaps you should change 'klee' to jackson pollock??? hehe

[Scott Stites]

Thanks Unca K and V-un-v. The breadboard does sort of look like something Pollock would have put to canvas.

In this sample the Klee is acting more like a 'normal' sequencer - I didn't change any of its pattern switches throughout. Instead, I had it driving the four singal sources I used in the sequence, but also had the keyboard patched to the extra expo input of one of my Schmitz VCO3's - this is what provided the extra little quite notes heard throughout. The sequence was 'found' with a particular switch setup and the four Klee pots.

There was only one VCO (the VCO3) used for the voices. I used one half of my dual late MS-20 filter in self resonance as a signal source for the main 'low' part heard in the sample. It's patched through the wave multiplier. Before it was sent to wave multiplier it was multed to an input on my ring modulater. The triangle wave of the VCO was patched to the other input, and the ring modulated product of the two was used as the third signal source. The VCO, wave multiplied resonant MS-20 filter, and the ring modulated product of the two signals were patched into my 2040 clone filter, then through a VCO into my mixer module.

The fourth signal source was the Dark Star Chaos 2000 patched into the other half of the MS-20 filter in high pass mode (for a bit of wang), passed through a second VCA and into the second input of the mixer module. This is the atonal sequence that is faded out about halfway through.

The Klee was providing voltage control to the two halves of the MS-20 filter, the 2040 filter, the VCO and the DSC2000. It, in turn, was being clocked by a pulse wave LFO. The Klee was providing gate and trigger output to the EG controlling the 2040 filter and its VCA. The same LFO used to clock the Klee was providing a gate to the second EG, which was controlling the HP filter the DSC2000 was passing through and its VCA.

As mentioned before, the keyboard was also controlling the intial pitch of the VCO - this is what provided the variation in pitch of the rapid notes heard intermittently through the sample. Because the VCO was also controlled by the Klee, pitch changes would also occur when the Klee changed its control voltage. Towards the end of this sample, the Klee control was disconnected from the VCO, and the notes heard do not follow the Klee, though they are filtered and ring modulated with a filter and signal source that were both controlled by the Klee. In fact, the VCO is only heard mixed in at the very beginning of the sample - I then mix it out so that only its product is heard in the ring modulated voice. In the longer version, it starts out that way, then I mix in the straight VCO signal (sawtooth) to the mixer.

In the longer version, I start out with the direct VCO mixed out, then I connect the square output of the wave multiplier to the sync input of the VCO, mix in the VCO, disconnect the VCO from the Klee, disconnect the sync, then reconnect the Klee for variation in timbre and towards the end of that section is where I started this shorter sample. It was 3 MB, so I opted just to post the tail end of it.

Cheerio,
Scott

[Uncle Krunkus]

My stripboard finger's gettin' itchy!

[seraph]

[v-un-v]

[vtl5c3]

So this is where you hide out! Your writeups on the super Klee are very interesting reading. I'm looking forward to the schematics.

[Scott Stites]

Hey Romeo! Nice to see you here!



Guess I better get to work on those schematics

Cheerio,
Scott

[Scott Stites]

I added a gate/trigger bus yesterday to the Klee. As mentioned before, there will be two busses. Each stage will have a On-Off-On SPDT select either bus or neither. The stages are ANDed with the clock, from which the gates and triggers are derived. Further, there is a control that will allow adjacent selected stage gates to merge, creating one gate and trigger instead of two or more with 'adjacent' selected stages.

I found that simply ANDing the stages with the clock didn't quite cut it, at least for my EGs (my EG's have comparators on the gate and trigger inputs, which makes them very easily triggered by the slimmest of voltage spikes). The problem was that the clock would advance the stage and be ANDed with a slim part of the the current stage before it advanced a few microseconds later. I had to create a slightly delayed (around 20 uS) version of the clock to do my ANDing - otherwise, that slim spike would trigger my EG. That was a simple enough matter to fix.

This addition really explodes the capability of the Klee. The hard part about working with it is that I always want to stop and record what it's doing at the moment. This has got to be one of the funnest things I've ever worked with. It really is a blast.

Here are a couple of samples I recorded. The first one, sample 8, actually uses a function I'm considering - inversion of the selected stages for a bus to create the gates and triggers - in other words, every stage will fire a gate and trigger *except* the stages selected for that bus.

Sample 11 is a bombastic little ditty that I recorded while manipulating between 'merged' gates and triggers and then either one or two selected stages. Manipulation of these switches can create some great diversity in a sequence.

Everything in these samples was driven by the Klee, and they're all just one take (no overdubbing). On each one, I additionally controlled a VCO with keyboard while at the same time it was controlled by the Klee. Each sample is from the same patch, with different settings.

There's plenty to scratch your stripboard finger with, Uncle K


http://mypeoplepc.com/members/scottnoanh/sitebuildercontent/sitebuilderfiles/kleemod2_sample8.mp3

http://mypeoplepc.com/members/scottnoanh/sitebuildercontent/sitebuilderfiles/kleemod2_sample11.mp3

Cheerioio,
Scott

---

kleemod2_sample11.mp3
Description:	Super Klee Sequence, Alternately Using Merged and Non-Merged Adjacent Select Stages For Gate/Trigger Sources	Download (listen)
Filename:	kleemod2_sample11.mp3
Filesize:	2.24 MB
Downloaded:	1404 Time(s)

---

[v-un-v]

[piedwagtail]

Love it,love it!

The first one had me back in the eighties,and my first hearing of Sun Ra's Disco 3000.Something about that sclicking drum sound.

I've been rewiring my tube preamp this weekend,i'm pleased that you don't do tube electronics,that ratsnest of yours would be scary.

Robert

[Scott Stites]

[piedwagtail]

Actually i was slightly scared when the neon voltage stabiliser started flicking off.
But my new redesign had turned the power supply into a 150V lfo complete with spark off at every cycle.Nevermind....

I'm finding that 'when to stop' is critical!

Robert

[v-un-v]

[v-un-v]

[Uncle Krunkus]

Great samples Scott!
You could make a great track just out of the first bar of No. 11!

When you do the schematics, make sure there's a good emphasis on what front panel controls it needs and your recommendations as to how they should be laid out. That's the only thing which, if not done properly from the start, could make designing the stripboards a bit ambiguous or frustrating. Does that make sense?
+/-15kV is fine with me! You might just have to find some big stripboard! But then you'd need some big components I s'pose!
I think Tom missed a "k" in there!

[piedwagtail]

V-un-V

http://www.juno.co.uk/products/200934-01.htm

Check the mp3 of Third Planet which has the drum machine.
The title track actually has the drum sound i previously mentioned but its not in the sample,or at least is very submerged in the mix 3/4 in.
Nice to hear Space is the Place again though.
They used to hop just before yeah.

Perhaps they'll need a 15kV Superklee to distribute power if there's any energy crisis up ahead and folks need to be rationed .

Robert

[vtl5c3]

Nice samples! #11 has quite the bombastic bassline. What did you use for that?

So, Friday night I was working an old NY Times crossword puzzle. 37 Across was "Twittering Machine artist".... and I thought "Holy DejaVu!!!"... check the link to see what I mean: http://www.moma.org/collection/browse_results.php?object_id=37347



I think the cosmos is telling me to build a SuperKlee.

[piedwagtail]

[germaniac]

Sample #11: I could listen to that riff all day--Klee Kosmic Kalypso! Great sounds as usual Scott!

Back to the peanut gallery for me. . . .

Joe

[Scott Stites]

Hey Romeo,

Wow - Twittering Machine - it does seem to fit. It would be nice if I could put that on the web page I will eventually do for this project. I'm sure there are massive copyright laws to put my butt in a ringer if I did, though.

I've uploaded an illustration of the bass voice - it's actually the same patch I've used in all of these samples, more or less.

The really low, steady part of the bass is a self-resonating Schmitz late MS-20 filter in high pass mode with no input. It's run through the wave multiplier, which has the folds and offset inputs modulated by two separate tri-LFOs. It's controlled by the CV out of the Klee. It goes from the wave multiplier triangle out to the 2040 filter. I used the MS-20 filter as a simple VCO, mainly because I was too lazy to hook up a second one.

The part that plays the high and low parts along with it is a more standard thing - a VCO controlled by both the Klee and my keyboard. The pulse wave goes straight to the 2040 filter.

The VCO triangle is ring modulated with the straight signal out of the MS-20 filter. This is fed to the third input of the 2040 filter.

When the sample first starts, I've got the ring modulated signal attenuated all the way down. Halfway through, I turn it up (the 2040 filter has input attenuators on each input). Then I hook up the pulse output of the wave multiplier to the sync input of the VCO, so the VCO gets synced to the self-resonating MS-20 filter. This is where more bom is put into the bombast for the second half of the sample. You can hear the higher notes get more gurgly/gratey from the ring modulation and sync.

The EG modulating the 2040 filter and the VCA is fired by two selected gates/triggers from the Klee gate bus. I flip on the 'merge' function for the breaks you hear in the sample.

The drummy/percussiony sound is controlled by the master gate and trig output (just follows the clock, so it's constant). That was the DSC2000 through low pass late MS-20 through wave multiplier grinder section through a VCA. The signal from the 2040 filter and the grinder output both went to my mixer module, where I control the relative level between the two signals before going out to the amplifier. The grinder, BTW, is what put that clicky, pitched sound on that voice, the one Robert mentioned.

Klee Kosmic Kalypso - I love it!

Cheerio,
Scott

---

kleepatch1.JPG
Description:
Filesize:	83.85 KB
Viewed:	673 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[v-un-v]