{"id":1658,"date":"2015-02-11T00:51:40","date_gmt":"2015-02-10T23:51:40","guid":{"rendered":"http:\/\/www.untergeek.de\/?p=1658"},"modified":"2015-02-21T11:34:58","modified_gmt":"2015-02-21T10:34:58","slug":"adding-a-slightly-odd-sub-oscillator-circuit-to-jenny","status":"publish","type":"post","link":"https:\/\/www.untergeek.de\/de\/2015\/02\/adding-a-slightly-odd-sub-oscillator-circuit-to-jenny\/","title":{"rendered":"Adding a slightly odd sub-oscillator circuit to Jenny"},"content":{"rendered":"
<\/div>

Let’s start with some good news: Jenny<\/a>’s here to stay with me, so I can start some serious modding. The single VCO and the filter design make it hard for her to growl credibly, so I tried to make her bark and her bite a bit tougher, using bits and gates from my basement supplies: adding a sub-oscillator, and pre-filter overdrive.<\/p>\n

\"Jen<\/a><\/p>\n

Subjenny<\/h2>\n

The sub-oscillator is simple and has been done by many<\/a> great<\/a> modders<\/a>: You add a divider circuit to produce a one-octave (or two-octave) square-wave sub-oscillator and feed the signal to the unused „Off“ terminal of the noise selector switch (over a 100k resistor so that the sub-osc signal is not overly loud). So now when the noise generators are switched off you can use the noise dial to add some deepness.<\/p>\n

<\/p>\n

\"IMG_2676.JPG\"<\/a>Nothing difficult in here, the only complication being that I did not have any 4013 dual D-type flip-flops that everybody except Neil<\/a> uses as frequency-divide-by-two circuitry. I came to use something a bit more esoteric: a 40192 BCD(!) programmable up-down counter<\/a>, because these were the only suitable CMOS frequency dividers I had on stock. I also did not want to solder anything to the original JEN IC’s, so I designed a small piggyback board to go between the 4013 on the oscillator board and its socket. It diverts the supply voltage and the square oscillation to the 40192. I also found that you’ll have to reduce the signal a bit if you do not want it to drown out the main osc; a 100k resistor in the path means -6dB if I’m not mistaken. <\/p>\n

\"Circuit<\/a><\/p>\n

Last modification of the modification: I had a four-position rotary switch to replace the standard 3-position switch for noise selection. Old maker’s words: If you have it, use it.<\/p>\n

So I tried out some rather unusual sub-oscillator sonics.<\/p>\n

Ask not what a frequency divider can do for you…<\/h2>\n

As mentioned before, my sub-osc is built as a piggy-back replacement for the 4013 on the VCO board.<\/p>\n

\"IMG_2683.JPG\"<\/a><\/p>\n

My first attempt hat the 4013 the wrong way round, so the add-on PCB is not remotely as pretty as it could be. Anyway, it works. Here’s a comparison of the Jen’s original sawtooth wave, and a sawtooth wave played one octave higher with some sub-osc added.<\/p>\n

\"Sawtooth<\/a><\/p>\n

I also switched some resistors on the noise board following Ken’s level adjustment advice to raise the noise signal levels a bit (see his PDF<\/a>) but found that the new sub-osc signals were too loud. So I added the 100k resistor in the path as mentioned above; it becomes of the voltage divider via the noise level pot and effectively halves the signal level (i.e. -6db). <\/p>\n

The madness of 40192<\/h2>\n

As mentioned above, I also switched the noise selector switch for one with 4 positions, so I could have another option to the -1oct square wave source. Maybe a -2oct wave? Well, not quite.<\/p>\n

The distinguishing feature of my frequency divider is that is a BCD counter. BCD meaning: Binary-coded Decimal, i.e. one full cylcle is 10 steps rather than 16, as with a hexadecimal counter decade. And while bit 0 of the counter just puts out a f\/2 square wave, bit 1 is asymmetrical:
\n\"--__--_-<\/a><\/p>\n

Okay, that’s positively weird. But it sounds interesting – here’s a simple sawtooth, then with some sub-osc added, then with the -2.5oct wave:<\/p>\n