Harmonic Multiplier

[soggybag]

Has anyone built the Harmonic Multiplier from JC Maillet's site? This look like a very interesting and unique approach to octave effects. With a minimum of parts.

[/url]

[jmusser]

This is the first I've seen of it, and I'm not exactly sure what a second harmonic is, unless he means a one octave down, with everything else removed, but he also says something about ring modulation in there too. The description is over my head I'm afraid, but just building the circuit, looks fairly straight forward. The schematic is not very readable unless it's magnified, but it prints out well and to scale. Hopefully someone will enlighten me on this effect.

[Eb7+9]

... a very clean analogue octaver - based on the idea of tying both inputs of a ring mod to a single source ... in this EMS derived circuit the fundamental and higher order terms can be easily nulled to better than 80db ... my circuit is used to simulate tube "warmth" when mixed with dry signal in low amounts (an idea borrowed from Eric Pritchard's research), and pretty insane octavia effect at full mix ...

... the waveshaping is based on linear waveform multiplication, and not by full-wave rectification like pretty much every other analogue octave-up circuit ever made - some of which are excellent of course, but different in terms of transient response and coloration ...

... interestingly there's at least one other analogue based method for achieving parabolic transfer that never seems to have been used for instrument processing ... a naive Laplacianist can easily overlook the potential of such circuit techniques since their operation isn't at all described in terms of small-signal circuit behavior - that's not where their mojo hangs anyway ...

[Paul Perry (Frostwave)]

Eb7+9,   seeing you are as one with the maths, what about a log/antilog device so you have a knob giving arbitrary powers of whatever the input is? (which would include the 'squared' or ocatave doubler as a special case).
I hope this makes sense, i just had a couple of teeth out & the various painkillers are messing with my mind... yeah there's going to be a problem with the negative half cycle :oops:

[Ge_Whiz]

I'm not exactly sure what a second harmonic is, unless he means a one octave down,

Second harmonic = first overtone = one octave up. If you feed the same sine wave into BOTH inputs of a ring modulator, you get a pure second harmonic (in theory). In practice, you get 'ring mod tones' for anything that is less than pure.

[jmusser]

Thanks for the detailed explaination. This looks to be a very cool circuit! It looks like a moderate level build.

[Paul Perry (Frostwave)]

Something people don't seem to mention, dealing with the analog multiplier style of freqyuency doubler, is that the dynamics of the input signal is increased. (think about it.. if you multiply X by X you get X squared, but if you multiply 2X by 2X you get 4 times Xsquared.
Of course, when you have a ring modulator with a constant amplitude local oscillator, the dynamics are unchanged.

[puretube]

err, this has been mentioned in an earlier thread...  :wink:

[Mark Hammer]

In one of their units, Aphex used something like this to generate added harmonic content.  Many "aural exciters" use distortion to acquire the extra top end to mix in with the original.  Aphex ran the original through a highpass filter, then into a VCA modulated by the highpassed signal, yielding doubled harmonics, which were then mixed back in with the original.

Nifty.  Not sure how it sounded though.  Steve Giles sent me the schem.  He may know something about the audio behaviour of this unit.

[Pedal love]

I had always assumed that the balanced modulator/demodulator, like the MC1496, produced too much distortion to be used as an octaver. Doesn't it produce on the order of 1% distortion.pl

[puretube]

that 1% won`t hurt a guitar-sig...  :wink:

[loscha]

Paul, maybe you should do a mod to the Blue Ringer that routes signal to both sides of the Ring for octave up?

Bill and I have routed a booster --> both sides, and it sounds quite good on guitar.

[Paul Perry (Frostwave)]

Paul, maybe you should do a mod to the Blue Ringer that routes signal to both sides of the Ring for octave up?

As you have found, it isn't a difficult mod, you just plug a Y phono splitter in to the audio in, and run a cable across to the alternative external carrier input.
So long as you have the input sig boosted up, you can have fun running it into the voltage control input for the local oscillator, also.
Hell, you can run the output into the external carrier if it comes to that. Anyway, I enjoyed it

[Eb7+9]

To be clear the EMS circuit it's based on is a synth processing circuit that can handle something like 3 to 5 volts of input before the modulator saturates ... this is contrary to most full-wave octavers which produces their cleanest (fundamental most suppressed) octave with least input signal ... I remember Roger Mayer's Octavia manual mentioning something to that effect ...

Now, it's true you get side-terms when the input waveform isn't purely sinusoidal - producing the characteristic ringing - but still you'll hear the difference between one type of octaver and another with a guitar signal ... this is because on a clean input tone the upper partials (4th, 6th, 8th ... which are mixed in with the multiplied terms in the non-sinusoidal case) are nulled out in this circuit ... this doesn't happen with rectitifer type octavers since the two paths can never be that well matched ... so the octave along with side-terms are cleaner in this circuit ... it's not a huge deal, it's just different ... the nice thing is what happens when you mix in 5-10% of this stuff to dry signal - you get nice tube fakeness, which is what Eric Pritchard was after ...

Paul ... that's a good point - it explains the "jumpy" nature of octavers - not sure if you would want to get rid of it - interesting idea though ... for some reason I never got into log/antilog amps that much, something I never thought of trying with octavers  ... these days I'm trying to wrap my head around the Art Levelar control circuit which linearizes the 1/x current-to-resistance relationship of diode optocouplers to the LA2A style voltage divider - I'm just getting into that stuff ...

Mark ... interesting ideas - must be a challenge too keep the noise down on all that stuff ... for a while I was toying with the idea of using the control voltage of a PLL to drive a linear bandpass VCF as a means of "capturing" harmonics thinking it might produce a tracking wah slash feedback inducer ... it's probably useless as such but likely weird - especially as the loop is locking onto the signal ... I wouldn't mind checking out that Aphex schematic if you please ...

If anybody's serious about building this thing pm me and I'll mail you back a copy of the EMS nulling instructions ...

... jcm

[puretube]

"Elektor" had a doubler, where one channel`s amplitude was held constant...

[soggybag]

A few years ago I was into building modular synth stuff. I seem to remember seeing a ring modulator that used a 1495 chip. I think it was described as a four quadrant multiplier.

[Paul Perry (Frostwave)]

I seem to remember seeing a ring modulator that used a 1495 chip. I think it was described as a four quadrant multiplier.

You certainly did. A very low point of my life, when I was just starting in FX, was designing a ring modulator using MC1495s and then realising that they had gone out of production, just as I was trying to make a commercial quantity of ring modulators. So I got the AD633.......
If anyone ever needs a MC1495, I have a few, and can swap them for Duplo animals. (always need Duplo animals for a toddler I know).