The following is a demonstration of how to make a Graphic VCO or stepped/variable waveform VCO using a Doepfer A-155 sequencer module.
Part 1: A few things to consider before we begin...
- If you have modified your A-155 as described on the Doepfer DIY with Modification #2, it is not suitable for this application. A switch could be installed to allow one to disable the mod when using it as a Graphic VCO. Also, in an email communication with Doepfer, I was told that the A-155/154 combination is also unsuitable for use as a Graphic VCO. However, this is how my system is set up and how I made these examples. The email response is shown here:
"The debouncing mod reduces the max. clock frequency and has to be removed for Graphic VCO applications. As the max. clock frequency is
also limited in combination with the A-154, the A-155 should be used
without A-154 for this application."
- Dieter Doepfer
- DC offset: a standard VCO output has a level of 10Vpp (peak to peak) or -5/+5v. However, an analog sequencer generally outputs 0V as the lowest note and a positive voltage as the high note (ie. 1,2,4V etc, depending on the sequencer model and the range settings you choose). This is a fully positive DC offset and not ideal for your audio signal. In my experience, something else in the signal path will compensate for this. This is because another module is using a capacitor at the input to correct DC offset and center the waveform on 0V. If you look at the video clips here, and/or listen to the audio, you will notice changes in the overall level because the max level is only as high as your highest set step. In order to keep a more consistent level, one might keep at least one step set at max voltage and one set an minimum voltage. But still, if your sequencer is outputting 0-4V, you will only have 4Vpp, or -2/+2V. This is not as loud as a normal VCO output. For best results, one might amplify the signal to make it -5/+5V, otherwise your filters with overdrive won't actually overdrive as they are expecting a 10Vpp signal. Another email from Doepfer confirms this:
"The DC output is indeed a problem when DC coupled modules are
used to process the graphic VCO signal..... The only solution is an
AC coupled input of the succeeding module. All A-100 filters and
some A-100 VCAs have AC coupled inputs."
- Dieter Doepfer
NOTE: Consider building a x2.5 amplifier module with an AC coupled input (both circuits are described on the Doepfer DIY page) if you plan to use this Graphic VCO application often, especially if you want to use filters with a built-in overdrive at the input. The overdrive appears at the 8-10 level range only if the input signal is 10Vpp!
Part 2: What is a "Graphic VCO" and how do we create the basic patch?
The A-155 Analog/Trigger Sequencer makes a stepped waveform whose shape is dependent on the values set by the 8-step voltage sequence. This sounds rather obvious and redundant. Normally this sequence would control the pitch of a VCO or perhaps the cutoff of a VCF. But in this instance it will actually be producing the audio signal by being clocked at a higher rate by a Doepfer A-110 VCO square wave (instead of something like a 1/16th note gate pattern at 120bpm). To "drive" the sequencer to audio rate, simply take the Square wave output of an oscillator and connect it to the Clock input on the A-155, or the Ext Clock input if using the A-154 Controller. Now when you change the pitch of the square wave VCO, the "pitch" of the sequencer waveform will change with it. Simply treat the sequencer output like you would any other VCO output (whether using the Pre or Post out) and patch it to a filter input, VCA input, or use it as an FM source (preferably AC coupled) for your other VCOs or VCFs.
"The A-111-1 is a better driver for the A-155 in the graphic VCO mode because it's frequency goes up to 40 kHz with good 1V/Oct tracking." - Dieter Doepfer
The oscillator I used to drive
my sequencer to audio range is
shown here....
Modified A-110 VCO, modifications
do not affect this application.
my sequencer to audio range is
shown here....
Modified A-110 VCO, modifications
do not affect this application.
Part 3: What does it look like?
Imagine a normal typical Sawtooth oscillator waveform. If the values of the sequencer row are set in descending increments you can create a stepped sawtooth shape as shown here. You can also imagine that setting steps 1-4 to the maximum value and steps 5-8 to the minimum, you would create a square waveform. 1-6 max, 7-8 min creates a 75% pulse. 1-2 max, 3-8 min creates a 25% pulse. More interesting results can be found by trying different values for the 8 steps to make uncommon waveshapes as shown below.
Part 4: Modulating the Waveform
A great feature of the A-155 is that the lower of the two sequence rows has separate inputs for each step (see the image at the top of this page) which allow you to interrupt the voltage set by the knob with any external source. The knob then acts as an attenuator for this external signal. There are many applications for this in a large modular system, like cycling through different LFO shapes/rates or even oscillator waveforms. Trills can also be created by including a voltage quantizer on the sequencer's output and injecting an LFO to one step of a pitch sequence. In this Graphic VCO example, I used three different LFOs (sine and triangle shapes, all different rates, produced by two Doepfer A-146 and an A-147) to replace the set voltage of 3 of the 8 steps. This provided a constantly changing, yet uncontrolled, waveform. An example is shown in the video clip below.
Doepfer A-155 Sequencer as graphic VCO from N K on Vimeo.
The clip is monotone, and it has a bit of LPF modulation provided by an A-121
filter and yet another triangle LFO. You can see the steps of the waveform
being modulated, and also the effect of the filter is both audible and visual,
seen as the smoothing or rounding of the steps' edges.
Doepfer A-155 Analog Sequencer used as graphic VCO (part 2) from N K on Vimeo.
The above video shows the same sort of setup, although a different set of three steps were modulated. In this example, the square wave clock source is swept by hand and the octave select dial is turned. The slow manual sweep seems to create some effect like phase or sync.
For a more deliberate result, a multi-output MIDI to CV converter could be used to program the waveform changes with a computer or MIDI sequencer. Any modulation source can be used though, like random generators or joysticks.
Part 5: A sequenced example
Finally, the values of the sequenced steps can of course be manipulated by hand in real time. In the following audio-only example, I used a Roland SH-101 to output a CV/Gate sequence to a Doepfer A-156 Dual Quantizer module and an A-140 ADSR. The quantized CV was then sent to the Pitch CV input of an A-110 VCO. The Square wave output of the A-110 VCO was used as the clock source for the A-155 Sequencer via the A-154 Sequencer Controller's External Clock In. Remember that the SH-101 and the A-110 are not used for audio, but only to control the A-155. The SH-101 sequencer is "playing" the A-155 "VCO". All you will hear is the A-155's Row 2 Post Output, which is after the Glide processor on-board the A-155. I recorded the audio clip into my computer using Nuendo 4.x.
During the recording, I adjusted the settings of the 8 steps of A-155, row 2 by hand. This is somewhat awkward to do, and it's hard to predict the results
while you do it. You can hear the changing tone in this clip, and while you may
not want to have this kind of morphing waveshape in your track, this clip
demonstrates a variety of the tones that are possible. Using this method it may be possible to find just the right tone to fit in your mix. After recording, I rendered the audio with a small amount of plate reverb. Other than the reverb, you are only hearing the changing tone of the A-155 as I somewhat randomly and frantically turned the knobs of Row 2.
Part 6: A little extra info about my A-155.....
It may also interest some of you to know that I modified my A-155, however the mods do not affect the Graphic VCO application.
***Warning: any modifications will void your factory warranty. Contact Doepfer to have any modifications performed if you are concerned about your warranty.
The modifications I did are the following:
- an "expansion module": a vertical row of 8 sockets and LEDs mounted on a 4HP faceplate which make gate signals for the individual steps available to use with other modules, like envelopes, VC switches, sequential switches, resets, etc. and even some features on the A-155 itself. It basically functions like the Doepfer A-161 Clock Sequencer. I made a small PCB and mounted the sockets, LEDs and resistors to it. I tapped the trigger signals from the center pin of the trigger output selector switches and grounded my expansion module back to the A-155. This add-on is shown on the left of the A-155 in the image below.
- two sockets that interrupt the Pre Out signals going to the on-board slew
limiters, so that the Pre Out signals can be run to a voltage quantizer (in my
case a Doepfer A-156 Dual Quantizer) and then back into my added sockets to receive Glide from inside the A-155. The quantized sequence with glide is then available at the Post Out sockets. The two sockets are shown in the image below. They are on the bottom row of sockets, below the Scale and Glide control knobs for Row 2.
- DPDT switch : I installed a switch above the Trig 1 label on the A-155 faceplate, which can be seen in some of the pictures above. Originally, this switch was for the Modification #2 suggested on the Doepfer DIY page. But I never did that modification since I've never had the problem it is meant to correct (which is the trigger state remaining high when the sequence is stopped during an active step). I am now considering using this DPDT switch to replace the some of the jumpers inside. One set of jumpers (J2) controls the S+H Control polarity. Another set (J1) controls the Glide Control polarity.
