Super Klee Sequencer

fonik · Posted: Sun Oct 15, 2006 1:14 pm Post subject:

as i understand it, it would make a melodic programming much easier and the "free" programming is possible still, so i would prefer this new approach.
on the other hand the melodic part can be done by an external quantizer? would be more intuitive, perhaps.
i wouldn't complain about one more control though...

Coriolis · Joined: Apr 11, 2005 Posts: 616 Location: Stilling, Denmark

Well, I think you're on to something (listening to the sample confirms this suspicion Cool

). I don't think skimping on panel controls is really in the spirit of this beast - no reason to get ridiculous of course, but so far the controls you are proposing seem to add to the fun of the Klee. I really like the idea of having a "user preset" voltage.

I say give it a 10-turn pot annacoupla rotaries!

It's not like it has to fit into a certain panelspace - it'll get the space it needs (at least mine will - when I build one)!

Thank you for this project Scott Very Happy

C

Scott Stites · Posted: Sun Oct 15, 2006 1:41 pm Post subject:

vtl5c3 · Posted: Sun Oct 15, 2006 1:59 pm Post subject:

I think you kinda lost me. Are you talking about having a 9 position rotary switch, where the 1st or the 9th position is where you use a multi turn pot to dial in any interval you desire? If so, I like this idea. I'm not concerned with western tonality... but that's just me and the rotten fruit the audience throws at me

blue hell · Posted: Sun Oct 15, 2006 2:39 pm Post subject:

I can't really oversee how the two rotary solution would interfere with less often used tuning systems - but that's really the only possible issue I see with it.

The "problem" is that the interaction between the step pots defines a sort of a scale from which patterns select notes, but there are so many possible combinations that it's almost impossible to see if they would be meaning full for all possible patterns.

Intuitivly it seems as if enough freedom for exploration would remain for the somewhat more limited possible set of reference voltages (when rotary switches are used).

It might be more work to get it setup for some tuning system. Most likely though most people will not use it that way, but rather pick a musical reference interval and have the pots at rational fractions, like closed, half or fullty opened, maybe quarters but that seems complex already in how the resulting scale would turn out to be structured.

Scott Stites · Posted: Sun Oct 15, 2006 4:17 pm Post subject:

Romeo - yes, the idea is that one position of the rotary would enable a pot to set the voltage range.

janvanvolt · Joined: Nov 24, 2005 Posts: 285 Location: Mainz, Germany

Out of curiosity: Are they some EAGLE files (or maybe a finished) for the Super Klee PCB ?

While reading and watching the thread it seems to be a perfect addition to my Q119 and MFOS sequencer ...

Scott Stites · Posted: Sun Oct 15, 2006 9:08 pm Post subject:

Nothing as yet, though Uncle K is tracking a moving target for a stripboard version.

Cheers,
Scott

Scott Stites · Posted: Wed Oct 18, 2006 11:42 am Post subject:

Well, I'm stuck in an endless loop whether to go ahead with the more versatile max range, either with a variable pot on the 2V range or two rotary switches, one of which selects half step increments and one of which selects octave increments.

The additional pot would fit right in with this schematic. For the rotaries, I think I would re-do the circuit a bit - put in a constant current source and a resistor string for the half step increments, and another for the octave increments. CD4051 would have to go, which wouldn't be a problem.

Here is the schematic using the 78L10 as the 10V reference with the major 3rd, perfect 4th, perfect 5th and minor 6th range settings, and then the octaves. I think I'll do one with the pot replacing the 2V range, then see if I can't do a version with a decent current source and two rotaries. It's something I should breadboard and try out to see if it really is all that useful. Sorry for the end-stage waffledom here, guys.

Cheers,
Scott

blue hell · Posted: Wed Oct 18, 2006 12:35 pm Post subject:

vtl5c3 · Posted: Wed Oct 18, 2006 1:22 pm Post subject:

Seriously, if you come up with several variations that are all good - publish them. It may be easy enough to get multiple variants on the same pcb layout.

fonik · Posted: Wed Oct 18, 2006 2:31 pm Post subject:

Scott Stites · Posted: Wed Oct 18, 2006 2:33 pm Post subject:

Ah, thanks guys, I'll just do the different variants. As for our stripboarder, it's up to him which he prefers to do.

Thanks all!

Cheers,
Scott

Uncle Krunkus · Posted: Wed Oct 18, 2006 3:03 pm Post subject:

Can you make music with this stuff? Wowww!!! Laughing

I'm sure I'll get around to that some day! Embarassed

I haven't actually done enough of the scaler/decoder to be committed either way at the moment. (as a male of the species I try to keep my options open for as long as possible! Laughing

)
Even when hearing the Sorcerer is only an hour's work away, I can't say if that will happen today, tomorrow, or next week. That's parenthood I s'pose. Cool

It's definately worth taking a couple of photos though. I'll do that now before I forget. Rolling Eyes

Scott Stites · Posted: Wed Oct 18, 2006 6:29 pm Post subject:

Uncle Krunkus · Posted: Thu Oct 19, 2006 2:35 am Post subject:

Yeah,
the aboriginal name for Ayers Rock is Uluru (pronounced OOLaRoo) There was an official re-naming a number of years back as part of the reconciliation processes which various governments have supported to greater or lesser degrees.

Scott Stites · Posted: Thu Oct 19, 2006 6:17 am Post subject:

Ah, thanks! My memory was already turning it into Uhuru rock, which perplexed me - why name it after the communications officer on the Enterprise? Why not Kirk Rock, or Sulu Rock, or Bones Rock, or (my choice) Spock Rock? Very Happy

Hmmm... Oh, yes, the Klee - I'll get the version with the external pot up. Twin rotaries will take some breadboarding (which means I have to go shopping for some acreage - the Klee expanded beyond its borders again with the last round of breadboarding and the CD4050 coup). Probably won't happen this weekend - I've got a camping trip scheduled for the wilds of Northeast Kansas.

Cheerios,
Scott

Uncle Krunkus · Posted: Thu Oct 19, 2006 6:52 am Post subject:

"Spock Rock"! Very Happy

That would make a great name for a minimalist electro retro glam band!

Scott Stites · Posted: Thu Oct 19, 2006 10:51 am Post subject:

I heard that! Glam means green make-up, long ears would be allowed!

Here's the schemo for the panel mount 2V (2 octave) variable range version of the decoder. As you can see, it's not much of a departure from the non-variable pot version.

The twin rotary, however, will be much more of a departure. I envision the first rotary to step from 0V to one half step below a full octave - twelve positions, and the second rotary to vary from 0 to 8V (9 steps). Using these two rotaries, any step in a 9 Octave range could be set as max voltage for the pots, though in truth, the first two octaves would be the most crucial for Klee type stuff. Nice benefit of having 0V for both rotaries is one could set the ranges to 0V and set the base frequency of VCO's being modulated without stopping the clock on an unoccupied stage (which is something of a trick once you have multiple bits in the pattern) or setting the pattern to all off.

The departure concerns how the voltages will be derived - as I mentioned before, I plan to get rid of the CD4051 and use constant current sources and resistor strings for each rotary pot.

Having never actually played with constant current sources, I've been nosing about a bit. I've come across two candidate circuits to base this one off of - the constant current source used in the ARP Odyssey keyboard or Ray's ingenious use of the LM3900 in his keyboard circuit.

The ARP Oddysey would entail possibly the least parts/complexity. Ray's may or may not be more temperature stable, though I have a feeling that the ARP circuit may do just fine as well - it uses a diode to swamp out any temp sensitivity of the transistor. Before I'd use Ray's, I'd certainly ask if it was OK with him - not that the Klee is selling for money or anything, but I hope it isn't construed as something that would bite into his board sales for his sequencer. That and I have the greatest respect for him, like everyone else here.

Anyhoo, I've also attached a circuit fragment of the ARP transistor current source. Values would probably have to be altered somewhat. Any other suggestions welcomed as well.

Cheers,
Scott

ARP_constant_current.JPG

Description:

Constant Current Source Transistor from ARP Odyessy Keyboard

Filesize:

46.57 KB

Viewed:

10226 Time(s)

Scott Stites · Posted: Sat Oct 21, 2006 9:11 pm Post subject:

Working out the constant current source for the 1 Octave rotary was a piece of cake.

I found that finagling the values for a constant current source across 8 octaves, using the ARP circuit as a basis, doesn't work out so well. However, I think using a constant current source for the critical small voltages of the first rotary, and a buffered divider from the 78L10 for the larger voltage steps of the second rotary works fine.

I've also refined the range of the larger rotary - 0 to 7 volts (seven octaves). A combo of the two rotaries can get as high as a half step below 8 octaves. 8 positions on the second rotary just seem a bit more natural.

Schematics to follow.

Cheers,
Scott

Scott Stites · Posted: Sun Oct 22, 2006 12:15 pm Post subject:

Here's a schematic of the discussed third voltage range option - the two rotary switches.

Q1, D1, and associated resistors form a constant current source that is fed to R1 through R12. R30 adjusts the source to 833 microvolts. This creates a voltage drop of 1V between the top of R12 and ground - each resistor provides a .083V drop from that point, which corresponds to half step increments in a V/Oct system. SW1 selects the voltage tap, from 0V to 0.916 (which is one half step under one full octave) and sends this to U2A, which is configured as a voltage buffer.

The octave increments use the regulated voltage supplied by U1 to create the 1V octave increments. R31 and R25 are in series with R13 through R19. R31 is calibrated so that there is a 1V drop from the top of R13 to ground - this will ensure maximum resolution where needed most - at the lower octaves. Because R13 through R19 are equal in value, each tap will provide a drop of exactly 1V, which corresponds to 1 octave in a V/Octave system. Switch 2 selects from 0V at the bottom of R13 to 7V at the top of R19, and feeds this voltage to U2B, which is configured as a voltage buffer.

Buffers U2A and U2B provide the voltages from the two switches for summation by U2C and U2D. From there, the voltage is applied to the CD4066 switches on the decoder board. R32/R29 provide a means to zero any offset present at the output of U2D - it is adjusted for 0V at the output of U2D when each switch is set for its 0V tap.

I'm going to reconfigure the Decoder schematic - the CD4066 sections, etc. will make up that schematic. Then there will be three attenuator/scaler circuits that can alternately 'plug in' to the decoder - that way, anyone can select which of the three options one wants to put into one's Klee. Once that's complete, I'll upload the complete zip of the schematics, and put a fork in it.

Option 1 - a single rotary switch to select specific ranges.

Option 2 - A rotary switch to select specific ranges, with one range being a continuously variable range of 2V set by an additional pot.

Option 3 - Two rotary switches - one switch selects half steps up to one half step below one octave (12 position) and one switch that selects octaves from 0 through 7 (8 positions).

Cheers,
Scott

vtl5c3 · Posted: Tue Oct 24, 2006 8:04 am Post subject:

Hahaha... "put a fork in it"!!!

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Coriolis · Joined: Apr 11, 2005 Posts: 616 Location: Stilling, Denmark

Chief Wiggum?

Scott Stites · Posted: Sat Oct 28, 2006 8:48 am Post subject:

It's been quite a week - I've accepted a new position at work, which leaves me a bit wrung out in the evenings to do a whole lot. Today I'm aiming at incorporating the three scaling options into the zip file and getting that uploaded.

In the meantime, I did spend a little time playing with the Klee enough to record a couple of samples, which I've merged into one sample here. It's a fork dance for our hapless frog skin attired triclopian exotic dancer. Very Happy

Both sections of this sample incorporate three VCOs - the first VCO is controlled by the A output and passed through one section of the dual MS-20 clone filter, the second VCO is controlled by the B output and passed through the 2040 clone filter, and the third LFO is controlled by the A+B output, it's triangle wave is passed through the wave multiplier then through the second half of the dual MS-20 clone filter. Each VCO/filter/VCA voice is controlled by a separate EG, each of which in turn are controlled by I forget which outputs of the gate bus - I don't bother tracing that spaghetti-substituted-for-switch section of the breadboard very often.

Anyway, these samples represent a mode I don't believe I've posted samples of - 8X2 mode with section A in random mode and section B looping. The random source is the noise output of the DSC2000. This is an interesting mode, because the voice controlled by the B output constantly cycles an 8 step pattern, the voice controlled by the A output is a constantly changing random pattern, and the voice controlled by A+B is a mixture of the two.

The first section of the sample has no intervention from me - that thing played all night and came up with the coolest 'lead' lines from the random output - enough to exhaust Iris, that's for sure. The second part has a bit of intervention from me - I adjust the EG decay times here and there, and the folds setting on the wave multiplier - that's it (2040 filter is also controlled by a slow triangle LFO for sweeping it's cutoff along with the EG and the Klee).

Cheers,
Scott

forkdance3.mp3

Description:

Amphibial Fork Dance Medley

Download (listen)

Filename:

forkdance3.mp3

Filesize:

1.39 MB

Downloaded:

1179 Time(s)

Scott Stites · Posted: Sat Oct 28, 2006 1:29 pm Post subject:

@#$%!!! son of a #$%! !!!!! The entire schematic source files for the entire last rev are gone - fried in a shuffle of computers and a @#$!! Sony POS CDR drive.

I'll have to print out the JPGs and do a re-draw of the entire last revision.

@#$!!!! Evil or Very Mad