Super Full Wave Octaver

[Eb7+9]

I don't know how effective a filter's gonna be in removing that stuff - those upper 4th, 6th, 8th, etc all track the fundamental - you'd need a filter with cutoff tracking ... this reminds me though, I had an idea for another type of octaver circuit : track the fundamental with a PLL and play with the clocking to produce a square-wave of twice frequency, or track the octave from a dirty analogue octaver (first way probably best) - either way producing a square-wave tracking the fundamental at twice the frequency // then use some sort of square/tri/sine conversion and a mixer ... I'm sure this must have been tried somewhere but I never heard of it ...

~jc

[gez]

this reminds me though, I had an idea for another type of octaver circuit : track the fundamental with a PLL and play with the clocking to produce a square-wave of twice frequency, or track the octave from a dirty analogue octaver (first way probably best) - either way producing a square-wave tracking the fundamental at twice the frequency // then use some sort of square/tri/sine conversion and a mixer ... I'm sure this must have been tried somewhere but I never heard of it ...

The first method has been tried by a few here.  RG's mentioned it a lot if I recall.  I've done it sort of (divided a signal up using a PLL and used logic to switch in each 'segment' to get stepped PWM...well, it involves an output from the PLL many times higher in pitch  ).

[R.G.]

either way producing a square-wave tracking the fundamental at twice the frequency // then use some sort of square/tri/sine conversion and a mixer ... I'm sure this must have been tried somewhere but I never heard of it ...

You should read my posts. 

This is an old technique, and one I most recently proposed to one of Mark Hammer's questions about bass synthesis. You do whatever you have to to get a square wave that you can get a good phase lock on, then use a phase lock multiplier to get as many harmonics up and down as you need, then form the signal you want from the sub- and regular synthesized harmonics. Even as square waves, you form them down to sine and filter. The filter only has to remove the highest clock frequency you use in mixing to harmonics. This is essentially the way divider based organs make their sounds, except  you're generating the upper  partials by phase locking upwards before dividing down. 

The same trick works with a tracking filter. You can make any of an assortment of switched capacitor filters that give quick lowpass cutoffs, then phase lock multiply the fundamental up to generate the clock you need to run the filter. It tracks automatically because it's clock is generated from the signal it's tracking. It's quite difficult to get the filter tracking right any other way.

Another of my recent posts proposed another way. You can cover the full range of guitar fundamentals with a dozen 1/3 octave bandpass filters. Since the first harmonic of any sound is 2x the frequency, one octave up, then you can pick off the fundamental from such a filter array by generating a logic signal from the envelope of each signal indicating signal presence, then use any filter's "present" signal to inhibit all filters above it from having any output. Only the lowest frequency filter passes signal, and that's the one with the fundamental in it. Now you can use the fundamental for phase locking up and down.

A good phase lock octave generator can generate several octaves up and down, as well as adding intervals.

What it amounts to is that if you want a quick and necessarily dirty octave up, use one of the full wave rectifier techniques, but don't spend too much effort cleaning it up. For the same effort in cleaning up a FWR you could do a good, more flexible fundamental and multiplier.

[gez]

If I recall, I once buffered the cap used in the loop of a 4046 with an op-amp follower.  You end up with a crude frequency to voltage converter.  Hook it up to an OTA integrator, feed this with the output of the 4046 (square) and you can get a triangle with reasonably stable amplitude.  Use a LM13700 and you can run this into a second integrator to get sine shaping.  It worked well but there's a 'slight' taper in volume from low frequencies to high...but that sounded more natural to my ears.

[puretube]

Another of my recent posts proposed another way

=http://www.diystompboxes.com/smfforum/index.php?topic=39390.0

btw., Gez: interesting study cc. saw & phase & 2F:

http://www.pat2pdf.org/patents/pat2830481.pdf

[Paul Perry (Frostwave)]

Another way to kill spikes, is with a slew rate limiter. hey, we're stompbox builders, we can do whatever we want!

[jmusser]

I talked to my buddy Gez on getting this circuit in shape. When I first reviewed this circuit in another post, it ended up being basically a boost, with some up octave warmness to it, and as Marty puts it "I'm not getting much joy". I asked Gez what it would take to make a distortion out of it. I had already changed the 100K pot to 1meg, which helped some. He said, to leave that in, and put a diode across pins 2 & 6 of the second op amp, and if I didn't like that, I could put in a 15K from pin 2 of the second op amp, with a 1uf cap to ground. I did several tests with various combinations. I used a germanium, and an LED for the diode, but settled on another 4148, because it brings out a squarewave type buzz superimposed on top of the fundamental, along with some up octave product in there. The resistor and cap, really ramped up the gain, by at least times 4, which really helped the sustain. The 1uf was way too much bass for my taste, so I traded it for a .47uf, and it turned out to be just right. If you leave the diode completely out of the equation, you loose all of the squarewave type buzz, and basically go into boost town again, only at higher volume. As I told Gez, probably because of the frequencies that are enhanced, that the A & E strings have a tubey sounding quality to them, and the lower 4 strings with the squarewave buzz, as talked about earlier. The square wave (if that's actually what it is) is not as in your face, as say Tim's DOF, but it's there, and basically gives you major grind, and a lightly buzzy sustain on notes. Thes mods ended up making this a very nice OD/distortion type pedal, depending on how you're guitar volume is adjusted. I like it played on the neck pick up. On the bridge, it's way to bright and harsh to my ears, because of the up octave products in there. I also liked in 10 time better through the SS amp, instead of the tube amp. In my opinion, Gez helped me turn this circuit from enemic, to a fire breather, with very few changes.

[Doug_H]

Jeff, you are a fuzz-octave maniac...

Doug

[gez]

Thanks for the feedback Jeff (pun intended)!