LFOs - part II - sinewave LFOs

In part I, I discussed triangle wave LFOs. This time around we'll talk about sinewave LFOs. Sinewave LFOs show up mainly in Uni-Vibe type pedals, a few tremolo pedals and amplifier tremolo circuits. All of the ones I've seen are based on the classic phase-shift oscillator. To get a sinewave, we need exactly 360° phase shift at one frequency and exactly unity gain thru the amplifier and feedback loop at that frequency. If the gain is lower than unity, the oscillations will decay to zero. If the gain is higher, the oscillations will get larger and larger until clipping occurs. It's a delicate balance and it explains why sometimes a sinewave oscillator won't start up.

Below is the Uni-Vibe LFO. Q11 & Q12 are configured as a Darlington pair and provide the gain. C19-C21, R40-R44 and the dual C100K pot provide the phase shift. At one freq, the LFO's operating freq, the phase shift from the three R-C pairs adds up to 360°. Varying the resistance of the two pots changes the frequency where 360° phase shift occurs. The diodes provide soft clipping to limit the gain. The output is taken from Q12's emitter.

Sinewave LFOs gets kick-started when the power comes on and the sinewaves build gradually over several cycles and eventually stabilize. The designer of the original Uni-Vibe, Fumio Mieda, was a clever guy and put the diodes in a place where their clipping threshold varied with frequency. That way, the sinewave amplitude is small at low frequencies and gets larger at high frequencies. This helps compensate for the slow response of the light bulb used to illuminate the LDRs. Making C20 larger lessens the variable amplitude effect.
sinewave LFO.png
 
Awesome! What do you mean by "Making C20 larger lessens the variable amplitude effect." Will increasing it make the amplitude less strong, less obvious, slower? Asking for a friend... ;)
 
Allow me to clarify. When we turn the SPEED knob down the amplitude also drops. The amount of amplitude drop is less when C20 is larger. Increasing C20 only affects the amplitude at the bottom end of the SPEED knob.
 
Since the univibe C20 paired well with the bulb due to the slower response of the lighting, does this suggest that C20 should be adjusted if an LED is used in the circuit instead of a bulb?
 
No matter how big you make C20, the amplitude will vary with RATE. If you want to drive an LED with a sinewave, then the Abyss LFO is a better choice. It's basically the same phase-shift oscillator circuit, except the amplitude is controlled by Q1 going into cutoff and saturation instead of diode limiting. With this circuit, the amplitude stays constant when RATE is varied.

1614023251410.png

There are also opamp-based sinewave LFOs, such this one from the Hollis EasyVibe. It's basically a triangle wave LFO followed by a diode clipper that turns the triangle wave into something close to a sinewave. It works pretty well and is capable of a huge frequency range. Moen copied the EasyVibe circuit for the Shaky Jimi.

1614023833477.png

Here's an opamp sinewave LFO I designed for a Bass Vibe I'm building for my brother (don't tell him, it's a surprise):

1614024098592.png
 
Last edited:
Thanks Chuck. How apparent would the change in amplitude be in a univibe circuit if it were driven by an LED? I gather it would be like having the depth control turned up as the rate increased, and turned down as the rate slowed. Perhaps not musically useful.
 
Sorry for derailing the discussion about the LFO design to talk about whether using an LED to get an effect that tied the rate to the depth would be interesting. I guess it is just that the depth is an odd pairing with the rate for the two to be linked if you used the LED. Sort of reminds me of the control on the VFE Old School where the tremolo speed can vary based on how hard your are picking the strings. Neither one of those pairings is particularly intuitive. Alternatively, it could be interesting to have the depth on a modulation pedal tied to the intensity of the picking (either way).
 
Control interaction can be a blessing or a curse, sometimes both at once. For me, it comes down to how hard is it to dial-in the desired effect.
 
I found a great very simple LFO on schematic from a phaser of 1981, I tried it on 2 designs with great success, here is a link
https://www.freestompboxes.org/download/file.php?id=37873&mode=view
I added a depth pot adding a B1K before the led with
1 transitor emitter
2 led
3 ground

I, as well, put the resistor after the led but coupled with a trimpot (500R resistor + 1K trimpot) to adjust full LED brightness.

Super simple and great

Benoit
 
Not the best place to post that since that is not a sinewave LFO.

Have you noticed that at higher speeds the sweep up and sweep down are different rates?

The Phase 90 LFO has about the same parts count when you add on the emitter follower.
 
I experimented with The abyss LFO and noticed a fatigue in response of slow set up, the wave gets smaller and smaller to fade away completely.
 
Back
Top