another Octave Up

[onboard]

Looking at the transformer type octave boxes, like the Tycobrahe/Bobtavia/Neoctavia/SOU, made me wonder about trying parallel inverting & non-inverting opamp stages (hence the earlier thread about non-inverting opamp gain - thanks Andrew!)

I know there's a slew of octave builds out there, but I was curious to see if my hunch was right. The circuit ended up consisting of a single gain stage feeding inverting and non-inverting voltage followers in parallel, each driving a diode.  The output from each diode is then passively combined into a third voltage follower/buffer.

It works great! The octave is good and there's more sustain than I would expect. There's a higher parts count than the other opamp/shunt diode octave builds, and certainly more than driving a trasformer with a single gain stage. But a nice layout with a quad opamp could be very compact.  

Here's a link to the schem for the current version. I'm sure it could be cleaned up/tweaked even more - right now it's just bare bones functional. If anyone feels like trying it out or commenting, that would be cool. I'm just psyched that it works

[Doug_H]

Check out Gus's Octave Up Sick Box.

Doug

[onboard]

Right - I found that after the fact. Seems to be a trend with me at this point...

Anyhow, it works, and may be useful otherwise. What's the difference between the two appraoches? Specifically with regard to comparing a shunt diode configuration - Gus's OUSB -  to a circuit such as the transformer types circuits or my version. Guess I'm trying to wrap my mind around how this relates to diode clipping and full wave rectifying.

Working backwards is a little strange sometimes. Back to the search function.

[jmusser]

I would be interested to know the differences between the two approaches myself. Also, a request: Doug, would you resubmit the OUSB sound sample? It was awesome, and shows the different textures of that effect. I like the "clucking" pick attack on that sample. Mine isn't as apparent as yours is.

[petemoore]

Wellm, if it bears rendering, my OUSB worked quite differently when types of opamps were substituted in it.

[jmusser]

I don't know if Onboard was eluding to this or not, but I was asking, what is the difference between the two processes of inverting the wave forms? It seems like transformers do it naturally, because (if I remember right) the induced voltage from the primary into the secondary is 180 degrees out of phase with the primary. How op amps do it, I don't know. My bigger question, is if you already have the wave form rectified so that you have the negative going portion wrapped on top of the reference, how do you now, use that as a reference to make a second octave up?

[brett]

Hi.
The op-amps simply work as phase-shifters (sometimes called phase "splitters").  The diodes actually do the rectifying ("splitting") of the +ve portions of the signal, and the last op-amp recombines them.  

Could I suggest that it might work better if
1. the diodes are placed before the caps (ie run at Vref)
2. load each op-amp output with a small resistor (say 4.7k) to ground.  Place this after the diode and before the cap

This would allow continual positive conduction by the diodes, improving sustain (ie avoiding the "gating" effect).  Have a look at the schematic for the Tyco octavia.  From memory it uses a similar approach, and this is probably one of the reasons it has had so much success.  Interestingly, the Tyco circuits that I've built have never had equal rectified signals, so the amount of octave is not as strong as it could be.  

Have fun

[onboard]

...what is the difference between the two processes of inverting the wave forms?... (regarding the phase shifting) How op amps do it, I don't know

Yeah Jim, you answered your own question - as brett explained. Rather than use a transformer to create a signal that's 180 degrees out of phase (secondary) with the original (primary), my thought was to simply combine the rectified signals from non-inverting & inverting opamp stages. Same result.

Here's something cool -  the other octave up circuits really like it when your guitar's tone control is rolled all the way back. This circuit gives just as good an octave with the tone control wide open  :wink:

Brett, I'll have to try your suggestions about the diodes *before* the caps, and loading the opamp outputs. Very nice.

BTW, sorry to vanish for a while - new 40+ hr. a week job and other things at home, it's been pure insanity for the last month.

[jrc4558]

Lift Vref from the positive input of u2. I think the way it is in your first post, it will mute the u2 (Not a band thing, but Edge would mind).

[Mark Hammer]

Nice work.

My hunch is that bandwidth is often a key to a pleasing octave sound.  Consequently, I noted the rather high value feedback cap on the first stage with a certain amount of pleasure.

Here's my question.  What do you think of maybe reducing the cap on the first stage (to say, 100pf), and then sticking *another* cap on the inverting section after it (again, another 100pf to maybe 220pf)?  The idea is that the portion of the signal added in to produce the apparent doubling would be largely restricted to the fundamentals, with the harmonics greatly attenuated.

[puretube]

...  :wink:

[Quackzed]

this is also used in tim escobedo's octup blender circuit. only in a slightly different configuration..