Pitch vibrato and LFOs

[Herr Masel]

I was thinking about what composes a LFO, or how oscillators are created at all, I just know the basic theory of variable amount of voltage at low frequencies to modulate the tone, but I don't know much more than that, certainly not how it is implemented in practice. If it is relatively simple, wouldn't it be also simple to use an LFO to create a vibrato effect, or is this already common knowledge and the traditional way it is done? Are vibrato stompboxes made from pitch oscillator sections of a synthesizer?

[The Tone God]

...variable amount of voltage at low frequencies to modulate the tone, but I don't know much more than that, certainly not how it is implemented in practice.

That is not what the LFO does in a circuit like a phaser or tremolo. The LFO provides the sweeping voltage that controls something else like some type of varible resistive element (LED/LDR, FET, etc.). The LFO does not interact with the audio directly in most cases. The LFO could be removed and replaced with a pot that you swept manually and the circuit would still function. It is the way the rest of of the circuit, or atleast the audio portion, that does the actually work of modifiing the audio signal.

If it is relatively simple, wouldn't it be also simple to use an LFO to create a vibrato effect, or is this already common knowledge and the traditional way it is done?

Now you are talking about how the audio portion works which as stated above is independant of the LFO so modifying the LFO will not produce any change in the effect. You have to change the audio circuit to change the effect which may or not be "simple".

Also unfortunatly the term "vibrato" has been mis-used in many circumstances including the tremolo, phaser, and DSP worlds. True vibrato is the changing of the signal's frequency. This is not a simple task thus there is not really a "traditional" way to accomplish this unless you want to count the good ole' fingers on the instrument method.

Andrew

[markphaser]

Is there LFO circuits that are positive feedback?

Whats the difference between a positive feedback LFO VS a negative feedback LFO circuit?

[The Tone God]

Is there LFO circuits that are positive feedback?

Whats the difference between a positive feedback LFO VS a negative feedback LFO circuit?

All LFOs, and oscillators for that matter, must be positive feedback otherwise the gain generated would be nullified by the negative feedback and thus not enough return signal to cause oscillation.

Time to do some research on oscillator theory there instead of hijacking someone's thread.

Andrew

[robbiemcm]

Is there a schematic out there that will show a sort of vibrato like the one present in ZVex's Lo-Fi Looper thing (check the video on his site to see it in action - www.zvex.com )

[Herr Masel]

Thanks for the answers tone god, I wasn't thinking clearly enough, I had maths in my head. I will do some more reading on LFOs and oscillators. Could you tell me though (as briefly as you like) what the chief concept behind pitch vibrato is? With tremolo for instance I believe the signal oscillates between the output and ground, but I don't see how you can change the pitch electronically.

Hey Robbie I can actually answer that. The vibrato on the low-fi loop junky is caused by modulating the recorded singal (probably the speed) which causes the vibrato, and for some reason it is impssible to do on the fly. I guess there would be a latency between what you play and what you hear, how big of a latency though, I wonder? and how soon before chips are made with fast enough processing speed to overcome this?

[SeanCostello]

Is there LFO circuits that are positive feedback?

Whats the difference between a positive feedback LFO VS a negative feedback LFO circuit?

All LFOs, and oscillators for that matter, must be positive feedback otherwise the gain generated would be nullified by the negative feedback and thus not enough return signal to cause oscillation.

Time to do some research on oscillator theory there instead of hijacking someone's thread.

Andrew

What about 4 -pole resonant filters, a la Moog? These use negative feedback. A 4-pole filter made from OTAs can be used as a quadrature LFO. Each of the filter stages contributes 45 degrees of phase shift, and the negative feedback adds 180 degrees of phase shift, resulting in a resonant frequency where the phase shift totals 360 degrees.

And what about the oscillator that Hewlett and Packard created for Disney in the 1930's? 3 R/C stages (I don't remember the exact configuration), followed by NEGATIVE feedback. The 3 R/C stages have 60 degrees of phase shift at the resonant frequency, and the negative feedback creates an additional 180 degrees of phase shift, resulting in a resonance at the frequency where the total phase shift is 360 degrees. Worked good enough to provide the money to start a multi-billion dollar company.

Plus, there are plenty of digital systems that will oscillate with either positive or negative feedback, providing that the gain is greater than or equal to 1.0. If this wasn't the case, making digital reverbs would be a LOT easier, as you could just use negative feedback to avoid oscillations in non-unitary systems. Limit cycles could be eliminated, all sorts of good stuff - but this is not the case. Negative feedback will usually result in oscillation at different frequencies than positive feedback, but things WILL oscillate.

Speaking of it being time to do some reseach on oscillator theory...

Sean

[A.S.P.]

Barkhausen...

[SeanCostello]

Barkhausen...

Googling Barkhausen + oscillator came up with some VERY interesting links! At least interesting to me (disclaimer: I am not a EE, so I have picked up what little I know on the fly).

I like this quote:

All feedback amplifiers try to oscillate and all oscillators don't.

Sean Costello

[A.S.P.]

an early friend of Murphy...

(don`t start googling that one, though!  )

[The Tone God]

Could you tell me though (as briefly as you like) what the chief concept behind pitch vibrato is? With tremolo for instance I believe the signal oscillates between the output and ground, but I don't see how you can change the pitch electronically.

True pitch vibrato is difficult to do. There are effects to psudeo-pitch shift up octaves (as in the Octavia and the like) and down octaves (Blue Box) but those semi-tones are much tougher to do. Usually true/intelligent pitch shifting is done via DSP. The processor has software that quickly analyzes the current frequency and outputs a signal of the appropriate frequency change. That is a gross over-simplification too as there are great many detail I am glancing over, well more like all of the details , which if you are really interested you can do a search for as I am almost sure it has been discussed here and of course there would information on the "interwebs".

You may have seen "vibrato" on effects like the univibe. It really is just the phase sections output without the dry signal mixed in at the end. During sweep is sorta kinda like vibrato but If you were to hold the sweep still and listen you know that the pitch is not really changing.

Hey Robbie I can actually answer that. The vibrato on the low-fi loop junky is caused by modulating the recorded singal (probably the speed) which causes the vibrato, and for some reason it is impssible to do on the fly.

Actually not quite. Mr. Vex has posted previously that he archives the vibrato by modulating the power supply to the recorder IC ever so slightly. I have my theories as to what this does to the internals of the IC. Since this only technique only applies to the IC in operation it will not work during normal playing.

I guess there would be a latency between what you play and what you hear, how big of a latency though, I wonder? and how soon before chips are made with fast enough processing speed to overcome this?

Latency is one of the many issues. The fact is the IC in there is not intended for that function. There are ICs that have been around for a long time which can do this on the fly called DSP.

Speaking of it being time to do some reseach on oscillator theory...

Okay I lied...well more specificly I forgot to add the typical qualifiers to my statements but thats what I get for posting in a hurry. There is reason for lying which is to keep things simple for newbies atleast for the moment. Shhh. Don't tell them.

Andrew

[A.S.P.]

anyone noticed, that when stopping the clock-frequency-alteration of a BBD-based thingy at a certain point, shows a non-altered pitch/frequency very few milliseconds after having stopped the LFO?
Yet everybody will agree there has been a shift (of pitch for a single frequency & a shift of frequency for a complex signal... how ever you wanna define this in this circumstance) during the LFO action, similar to the action of a "...vibe"...

time to...   don`t ask me.

[SeanCostello]

True pitch vibrato is difficult to do. There are effects to psudeo-pitch shift up octaves (as in the Octavia and the like) and down octaves (Blue Box) but those semi-tones are much tougher to do. Usually true/intelligent pitch shifting is done via DSP. The processor has software that quickly analyzes the current frequency and outputs a signal of the appropriate frequency change.

Pitch VIBRATO is pretty easy - just use a modulated delay line. Changing a delay length will result in a signal where the output signal has the same ratio between harmonics as the input signal, and therefore is a pitch vibrato. BBD chips will work, as will interpolated delay lines in DSP. A sine wave will usually produce a sound recognizable as vibrato, while a triangle wave might produce up/down pitch shifts.

Pitch SHIFTING is, of course, really hard. Slight detuning is easy, but the more you have to shift, the more the glitches show up. So you have to deal with them somehow. Personally, I like to make the glitches sound good, instead of try and hide them, but that is another story...

Back to vibrato. Delay line techniques are really the best solution. Allpass based phasers (like the Univibe) can do an OK job, within certain parameters. The pitch DOES change, but the harmonic relationship also changes, so that the sound at the maximum modulation extremeties can sound very inharmonic. In other words, at the low part of the sweep, the pitch is perceived as lower, but the overall timbre is more like that of a bell. Frequency shifters can produce a nice vibrato, but the problem is that the same frequency offset is added to ALL frequencies, while in a true vibrato the amount of frequency offset is proportional to the frequencies in the input.

BTW, you can use delay lines to get FM of live signals, just by turning up the modulation speed to audio rates. This can produce some nice sounds.

Sean Costello

[no one ever]

about the lo-fi loop junky, the vibrato is caused by voltage swings applied to the recorder chip.

i have no idea how good ol mr. vex fit that in there.