CBS/arbitor doubler schematic, why impossible to DIY clean octave up?

[buildafriend]

I can't seem to find one 

also, does anyone have the knowledge who would not mind explaining why there is no clean octave up option in DIY land? and how can we actually get with it? whats holding us back?

[buildafriend]

Bump

[R.G.]

I can't seem to find one 

also, does anyone have the knowledge who would not mind explaining why there is no clean octave up option in DIY land? and how can we actually get with it? whats holding us back?

This has been asked here a number of times. The short answer is that it's really really hard to do with analog circuits and the nature of analog signals.

With digital/dsp techniques it can be done fairly readily, but then you have to first get over the hump of doing clean A->D-> before you can mess with the signal in sampled/digital.

It's not that hard to do an octave up for a sine wave, which is the simplest possible signal. An analog squaring circuit built out of what amounts to a multiplier fed the same signal twice or a device that has fairly clean square-law distortion does it. But guitar signals in particular and audio circuits in general are not sine waves. They're complex, time-varying mixtures of many sine waves at different frequencies.

Mother Nature (that is, the math that underpins the universe) said that any time you do a nonlinear operation on two signals, you get out both the operation on the two signals, plus the same operation on the sum and difference of the two signals. This is known in audio as intermodulation distortion. It's somewhere between difficult and impossible to keep the intermodulation products down low enough for rock-n-roll.

The only easy way to get octave/frequency-doubling is half or full wave rectification of the signal. This is a distortion process that produces a prominent octave up, but it also produces a hash of intermodulation products. All the other ways to get an octave up are much more complicated to do with analog circuits and don't get you to a really clean octave up anyway, only a clean-er result.

As with so many things, getting a partial if ugly result is cheap and easy. Getting a really good resultis more complicated and expensive, and in the extremely money and complexity limited world of DIY effects, as in the universe generally, the almost-good-enough has crowded out the really, really good.

[buildafriend]

Thanks, I'll chew on that for a little bit.

Has anyone ever seen the schematic I'm looking for?

[Mark Hammer]

Brother Keen is too humble to mention it, but consider exploring this circuit, whose goal was to come as close to a clean octave up as possible in a simple circuit: http://www.geofex.com/Article_Folders/jfetdoub/mudoubler.htm

I'll attempt a slightly simpler explanation that might be something you can swallow without having to chew too much. 

Virtually all analog frequency doublers operate by splitting the signal into its upper and lower halves (if one were looking at it on a scope), and recombining them in a way that mimics what it would be like if there were now two peaks for every one there was in the original signal.  It splits the signal into those two halves using diodes, or using transistors like diodes.  As you might recall, diodes normally only conduct in one direction.  The caveat for accomplishing that is that the signal has to be above a certain voltage to pass.  So, if the diode used to separate the signal into halves needs to see something at least 600mv in amplitude to pass, then anything that is LESS than 600mv in the signal simply isn't going to be part of the final output.  That, in itself, "distorts" the signal, and creates a certain glitchiness.

Additionally the ascending and descending parts of each half cycle are going to be "squared at their sides"; somethng referred to as crossover distortion.

Finally, for reasons I have never fully understood (a LOT of math involved), analog frequency doublers have a tendency to produce sideband products that make them sound a lot like ring modulators, no matter how clean.  That is, indeed, why the Dan Armstrong Green Ringer - an octave-up fuzz - is called the Green Ringer.

No matter how you slice it, because of what one needs to do to make the two-for-one peaks, there is no way an analog frequency doubler is going to sound like a digital or taped sample played back at twice the speed (or sample rate).

[buildafriend]

thanks mark! the circuit you showed me here seems to partially resemble a differential circuit. I remember seeing some crossover error in push pull style circuits. I have a little bit of knowledge of what common mode rejection ratios are from back in school and I've browsed over some op amp circuits from certain mic amps. what seems strange to me is that the first side band of a sine wave is an exact representation of its fundamental freq that can be boosted but I guess its not as ideal when you have a band of uneven frequencies. do I sound like I am grasping the issue properly? am I making sense? thanks regardless.

[analogguru]

Thanks, I'll chew on that for a little bit.

Has anyone ever seen the schematic I'm looking for?

Yes, I have seen it, but with a different name on it.  Since Gary Hurst never ever invented anything on his own  - "solder jockey of the 70´s" - you can´t be wrong when you use this schematic (tone control uses slightly different values):

http://www.generalguitargadgets.com/diagrams/ftmsc.gif?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

[buildafriend]

NOW WE'RE TALKING.

thanks A.G. 

I'm really hoping to find the original schematic. it sounds super awesome. By chance, do you have any idea how the original tone stack was configured?

Here are some pictures of the one I saw.