Simple octave down schematic?

[nick12391]

Hi,

I'm looking for a schematic for a simple analog octave down pedal. I've built a Green Ringer octave up circuit so I'm trying to find an octave down circuit of comparable complexity. I'm also eventually going to add a major third and a perfect fifth component (all analog), so if anyone has schematics for those that would be great!
(Here's the green ringer schematic if anyone's interested: http://www.generalguitargadgets.com/pdf/ggg_gro_sc.pdf )

Thanks!
Nick

[nocentelli]

http://www.diystompboxes.com/analogalchemy/sch/shocktave.html

[Mark Hammer]

Intervals other than octaves quickly become more complex than you are seeking.  Dividing by 2 is relatively easy.  ividing by 4 simply means dividing by 2 twice.  Other intervals requires counter chips to divide by other values.

[vigilante397]

I'm also eventually going to add a major third and a perfect fifth component (all analog), so if anyone has schematics for those that would be great!

If anyone has schematics for those I would be very impressed As Mark said, not really possible with analog.

[Marcos - Munky]

There's this one: http://hammer.ampage.org/files/EMMHarmoGen.PDF. But it isn't a simple build. I don't know how it sounds, didn't found any samples of it.

[Mark Hammer]

I built one of those, but was never able to get it working.  Or rather, never put in the troubleshooting time to get it working.  I'll work on it after I retire, and...um...after I get the HYperflange working.    I'll get back to you then.

I should point out that there are no counter chips that divide by fractions, only whole numbers.  What that sarcastic comment means is that, if one wishes to provide intervals between same-pitch and an octave down, one needs to start off with a pitch higher than the original, so that it can be divided appropriately by the counter chip.  In the project file that Marcos/Munky linked to, a PLL chip synthesizes a note higher than the original, such that if I divide a note by 3, but the note I'm dividing is twice the original frequency, then what I end up with is less than same-pitch, but not as low as an octave down.

Make sense?

And that, my friend, is why almost all the analog "other-pitch" pedals are all octave up or octave/sub-octave down.

[Marcos - Munky]

Yes, makes a lot of sense.

[slacker]

There's this one: http://hammer.ampage.org/files/EMMHarmoGen.PDF. But it isn't a simple build. I don't know how it sounds, didn't found any samples of it.

I bread boarded it years ago, it sounds surprisingly good on single notes, especially when mixed with the dry, tracks pretty well providing you play cleanly just like you have to with most analogue pitch shifters. Obviously completely useless with chords.

[Mark Hammer]

Thanks for that note of encouragement.  I'll move it a little higher up in the stack of things-to-get-to-when-I-retire.   

[PRR]

Side-topic:

It is non-trivial (impossible) to get the Frequency of a new note as quick as the ear thinks it does. So most "pitch trackers" are poor approximations on the beginning of a note.

However with sufficient brains (CPUs), we can take advantage of the fact that most musical instruments are "fixed pitch" (more or less). If the last several "A"s were say 217Hz, a new note in the "A" area is likely to be going for 217Hz.

This will be true for piano and straight (no wammy, no finger-bend) guitar, and near-so for most brass and woodwinds and good vocalists. Obviously it may go wonky with wammy, or a synth's pitch-bend. Or if a guitarist (or Buck Trent) touches-up a tuning peg mid-song. Even so, the target pitch is liable to be pretty consistent, or the audience will feel uneasy.

Many instruments vary in pitch at start-up due to pluck tension or air column build-up. Again a smart-enuff CPU could track this, and note that 219Hz start-up settles to 217Hz sustain, which helps guide pitch-picking from the start-up.

After a full set it should know exactly what the average pitch is, if instruments would hold pitch that long. More likely we should track the recent minute and discount older observations. But to be really clever, it could note that "A" has dropped 2Hz in 20 minutes, and anticipate another 1hz over the next 10 minutes, and micro-fudge its target pitch prediction to compare to observations. This might also bridge passages where no "A" is played for a while (key changes).

All of this, times 12 notes per octave (hey, you might use the odd notes), times as many octaves as you want covered.

Returning to topic: given a brain that can correctly output 217Hz when "A" is struck, it would be trivial to compute 217Hz * 4/5 to get the perfect Third down at 173.6Hz. This does mean a system to convert Hz numbers into actual sounds, preferably pleasing. It is hard to pop a VCO onto frequency in a few cycles. You can compute a waveform but the arithmetic suggests a whole second CPU (or DSP).

{edit -- bulb icon added}

[Tony Forestiere]

Wow Paul. Just Wow. As I read, I could clearly see that scenario in my mind's eye. Wonderfully written description meshed with your innate ability to explain things via clear and direct "splainin'".

Thanks. Heaps. 

PS: I miss the old light bulb emoticon. I would have used it.

[Mark Hammer]

If a person wishes to use analog means to yield other pitches from a single plucked note, I strongly recommend they use heavier gauge (.012 E-string) flatwound strings.  For several reasons: 1) stronger fundamental, relative to harmonic content (i.e., dull-sounding), 2) stiffer, so less incentive to bend, 3) stiffer, so less string deformity at initial pick.

There are so many circuits predicated on the sort of stable, uncomplicated pitch of an electronic oscillator.  We too easily forget about the physical realities of plucking skinny metal strings, with or without whammy bars.