TKB patch tips

64-step sequence:

The TKB's horizontal sequences (no matter how long or short) will appear to hang out on whichever row is illuminated by the LED at the far left of each row (well, if the TKB has those LEDS. Judging from the catalog photo I'm using for these patch illustrations, this wasn't always the case!). In actuality, the voltage levels defined by the knobs for all four rows are always available at their respective outputs labeled A, B, C and D.

Additionally, even with nothing inserted in the vertical clock jack, the voltages defined by the knob settings for the illuminated row will appear sequentially at the output ABCD (read "row A, then B, then C, then D"). Only when a trigger is sent to VERT CLOCK will the vertical sequence advance to the next row, and then that row's voltage settings appear sequentially at the ABCD output.

Wait, you knew that the TKB is really two independent sequencers, right? Well, if you didn't, congratulations on your bonus sequencer because you'll need both for this patch!

A 64-step sequence employs the VERT CLOCK and ABCD jacks. At the same time, each row's voltage settings are available at their own outputs and can be used simultaneously to control other things.

Patch a clock source to CLOCK. Patch the stage 1 trigger output to VERT CLOCK. Patch the ABCD output to the input of an oscillator, a filter, or any VC input in the system.

When the sequence reaches stage 1, a trigger from that stage is sent to VERT CLOCK and row A advances to row B. When stage 1 is reached again, a trigger is sent and row B advances to row C. This continues until row D wraps around to row A for a total of 64 steps, and the whole thing starts again.

The change of rows happens on the stage that triggers the vertical clock. So if you want the row changes to coincide with (say) stage 5, use that stage output instead.

Creating and skipping sequence sections:

Patch a clock source to CLOCK. Patch the stage 7 and stage 12 trigger outputs to RESET. Set the KEYBOARD switch to the UP position.

When a stage is patched to RESET, a sequence is created whereby the final stage in the sequence will be the one preceding the reset stage. With the patch above, three sequence sections are created: steps 1, 2, 3, 4, 5, 6; steps 8, 9, 10, 11; and steps 13, 14, 15, 16, 1, 2, 3, 4, 5, 6. The active sequence section is determined by the last pad that was touched.

Now play with the touch pads on the keyboard, and observe the following:

When pad 1 is touched: 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6...
When pad 3 is touched: 3, 4, 5, 6, 3, 4, 5, 6...
When pad 8 is touched: 8, 9, 10, 11, 8, 9, 10, 11...
When pad 13 is touched: 13, 14, 15, 16, 1, 2, 3, 4, 5, 6, 13, 14, 15, 16, 1, 2, 3, 4, 5, 6...
When pad 15 is touched: 15, 16, 1, 2, 3, 4, 5, 6, 15, 16, 1, 2, 3, 4, 5, 6...

Here's a cool thing: Even though stages 7 and 12 are patched to RESET, and would seem to not be included in the three sequencer sections described above, they can actually still be used in two of them: the second and third sections.

When pad 7 is touched: 7, 8, 9, 10, 11, 7, 8, 9, 10, 11...
When pad 12 is touched: 12, 13, 14, 15, 16, 1, 2, 3, 4, 5, 6, 12, 13, 14, 15, 16, 1, 2, 3, 4, 5, 6...

In this patch, stages 7 and 12 have another potential use. Each time a reset occurs, those stages trigger (in fact, when a reset occurs, both stages trigger since they're effectively patched to each other via the RESET jack -- even though they're both outputs!). Even if those stages are not in the active sequence, either or both of those stages' trigger outputs can be patched somewhere else (e.g. a trigger input on a DSG) to fire off events.

Further variations can be created using individual stage triggers to advance VERT CLOCK to the next row. For example, add to the patch above a cable from stage 6's output to VERT CLOCK, and then press touch pad 4. This will invoke the first sequence section that we created, and the voltages for each row will appear in succession at the ABCD output. The resulting sequence will be:

(row A) 6, 4, 5, (row B) 6, 4, 5, (row C) 6, 4, 5, (row D) 6, 4, 5, (row A) 6, 4, 5....

Throw in another twist by patching stage 16's output to VERT CLOCK as well. When touch pad 15 is pressed, the middle sequence section will be skipped as seen earlier, and the change from row to row will happen twice per cycle:

(row A) 6, 15, (row B) 16, 1, 2, 3, 4, 5, (row C) 6, 15, (row D) 16, 1, 2, 3, 4, 5 (row A) 6, 15...

And, as pointed out above, since stages 6 and 16 are effectively patched together via VERT CLOCK, both stages fire simultaneously and can be used to trigger events elsewhere in the system.

Synchronous random:

This is called synchronous random because one clock is driving two actions with each pulse: 1) advancing the sequence, and 2) randomly selecting the active stage.

Patch a clock source to RANDOM SELECT. Patch RESET to RANDOM SELECT.

Asynchronous random:

This is called asynchronous random because two clock sources are used for separate actions: one advances the sequence normally, and one advances the sequence randomly.

Patch a clock source to CLOCK. Patch a different clock source to RANDOM SELECT. Patch RESET to RANDOM SELECT.

First try this with clock source 1 moving faster than clock source 2. With each pulse from clock 1, the sequence advances to the next stage as usual. Each time a pulse from clock 2 is received, a random stage is selected, and the sequence being driven by clock 1 continues normally from the newly selected stage.

If clock 2's rate is not a subdivision of clock 1, the random stage changes will be rhythmically asymmetric; meaning the timing of the randomly selected stage will occur with no regard to clock 1's timing, and the resulting sequence will have an irregular feel to it.

If clock 2's rate is a subdivision of clock 1, the timing of the random stage changes will line up with the pulses from clock 1; so the sequence rhythm will be regular (assuming clock 1's timing is regular, as with a cycling LFO).

Here's a variation: First, make sure clock 2's rate is not a related subdivision of clock 1, and then swap clock sources 1 and 2. With the faster clock 1 now inserted in RANDOM SELECT, each new stage will be random. But occasionally the slower clock 2 (now inserted in CLOCK) will sneak in a pulse between those of the faster clock, causing the sequence to advance normally until the next pulse from clock 1 sends it off randomly somewhere else.