Sync-ing multiple 40106 LFOs

Short question: Is it possible to have a 40106 chip LFO synced by multiple LFOs at the same time?

Long question: I am planning an LFO module of a single 40106 hex inverter chip to make three square wave LFOs with different value capacitors, which gives a slow, medium, and fast rate LFO. I am looking into to placing switches between the three LFOs to engage sync between them:
[slow LFO] < switch > [medium LFO] < switch > [fast LFO]

The switches are planned as DPDT On-Off-On switches, so that: a) On 1: the slow LFO can input (sync) the medium LFO, b) Off: LFOs are not synced up, and c) On 2: the medium LFO inputs the slow LFO.

In this setup, the medium LFO can be synced by the slow LFO and the fast LFO. Is it possible to have the medium LFO synced up with both the slow and fast LFO at the same time, if I place 1N4148 diodes between the slow and medium LFOs at the one hand, and between the fast and medium LFO at the other hand? Or does this cause issues with the medium LFO, when synced by multiple LFOs?

In this simulation:

syncing a slower oscillator with a faster one, with no diode, makes its output the same as the syncing oscillator, but inverted.

Syncing with a diode pointing right, the oscillator never turns on. With a diode pointing left, it never turns off.

If syncing a faster oscillator with a slower one:

With no diode, synced oscillator output is the same as syncing one, inverted.

With diode pointing either left or right, I get synchronized behavior.

So I’m not seeing how to sync with both faster and slower oscillators separately, let alone at the same time.


It could be more interesting to have the oscillators run at different frequencies and mix them together to get a wacky waveform, as others have done before.


Just to be sure as to which oscillator is the one ‘synching’: Is the oscillator on the left in the schematics referred to as the one ‘synching’ the oscillator on the right?

Do you mean a wacky LFO waveform, or an audio waveform?

The module I am planning is intented to be a small 10x10cm clock LFO module with three LFOs with different ranges (slow, medium, high), mostly to clock a 4040 divider module I am planning at the same time.
While I was planning and breadboarding, I just though: why not introduce synching between the LFOs? :smile:

If you want something really small you can make a clock out of a 555 or an op amp. You can mix the 40106 oscillators at either audio rates to create metallic sounds, or at sub audio to get a quasi random LFO.

Have a look at CGS721 - Super Psycho LFO (though there’s no sync): CGS721 - Super Psycho LFO

I have the basics of the different 40106 LFOs now planned out, so back with another question. :smiley:

How can I hard sync 40106 oscillators with an external oscillator? I thought about putting a diode before the 40106 LFO to avoid negative input voltages, but does this work?
If I want to hard sync with an external LFO, and not have the sync influence the inputting LFO, how should I make the two connect? On pin 1 or 2 of the 40106 inverter?

Pin 2 is output. You don’t want to be connecting another signal there!


If you connect another oscillator to pin 1 it charges and discharges your oscillator’s capacitor overriding your inverter and resistor. Your oscillator just runs at the same frequency as the upstream, which isn’t what you want. So you use a diode: then when the upstream output is above the diode’s forward voltage it charges/discharges the capacitor and your oscillator follows the upstream frequency but when it’s below the forward voltage it doesn’t and your oscillator runs at its own frequency. (Or similarly but reversed if the diode points the other way.)

In a modular system to control the load on the upstream (external) oscillator) I’d probably add another inverter as a buffer, with a 100k to ground to set the input resistance, and maybe a Zener with an input resistor if the 40106 is running on +5 V to protect against overvoltage. With the inverter it makes sense to reverse the diode direction. So something along the lines of:

Alternatively you could buffer with an op amp in either a comparator or a voltage follower configuration, with the diode in the original direction. All three would behave a little differently with non pulse wave inputs.


Wow, thanks!!! I will definately try this out!

What is diode D2 used for? As it is directed away from GND, I assume voltage can only flow from GND, if the voltage of GND is higher than what is in the system. Is this to take out negative input voltages?

For diode D2, ypu used a 1N4733A. Could a 1N4148 be used? Or does that have characteristics that do not match? For instance: different voltage tresholds?

It’s a Zener diode. If the input voltage goes higher than 5.1 V it conducts to ground, preventing overvoltage on the 40106 input. As you note it also prevents negative voltages on that input. The presumption here is that the 40106 is powered with +5 V and a +12 V input would be out of spec. But if it’s powered with +12 V (and if the input can’t be anything higher than that) D2 could be a Schottky instead like a 1N5817.

1 Like

If I power the 40106 with 12V, should a 1N4733A for D2 work? I tried it with a 1N4733A, but an incoming LFO (almost 12V) does nothing to the 40106 I want to sync.

I do not have a 1N5817 laying around, so I could not test with that.

If you power the 40106 with 12 V you want to limit the voltage on the input pin to about 0 V (slightly lower) to 12 V (slightly higher). If there is nothing in your synth that produces for instance 15 V then that may not be a concern and you could just use a Schottky diode to protect against negative voltages. If you do feel the need to add 12 V protection you would want to use a 12 V Zener like a 1N4742A. The 1N4733A limits the voltage to 5.1 V, which probably isn’t high enough.

You might get away with a 1N4148 but in principle it should be something with a lower forward voltage. In fact you might get away with no diode at all since the 40106 has internal protection diodes; they’re intended for static protection, though. Not meant for handling negative inputs and cannot handle continuous current above a very small level, so the 1k input resistor probably ought to be made 10k or maybe 100k. I don’t regard that as good design though I have seen it in one commercial module.

I agree its not rigorous enough for a commercial product, for DIY should be ok 95% of the time, though definitely with a 10K or greater input resistor for current limiting into the 40106

@SynthesiS : This seems like a really cool idea for a module!

1 Like

Thanks for taking the time to ecplain things!

I have some 1N5819 Schottky diodes. Reading the datasheet, it seems as though they have the same properties as the 1N5817, only rated for higher maximum voltage. Therefore, will a 1N5819 diodd suffice for D2 in your schematic for the input signal?

1 Like

Should be fine.tiny


No matter what I try, I am not able to make this schematic work. The LFO I am trying to hard sync only turns on when its rate exceeds the rate of the input LFO.
Could you by any chance point me in the direction of an original schematic that uses the schematic you uploaded? Maybe that could help me look for differences with my setup.

See the MFOS Weird Sound Generator:

There’s other stuff going on there, the relevant part is something like this:


With the right oscillator frequency lower than the left, the outputs look like this (left oscillator on top, right oscillator on bottom):

With the right oscillator frequency higher than the left, the outputs look like this:

In both cases the right oscillator frequency is synced to that of the left oscillator; in the first case it’s just a plain pulse wave with the same frequency, in the second case the waveform has bursts with the right oscillator’s frequency embedded in it.