Ringdroid - a sample and hold ring modulator

[~arph]

Happy New Year!

Here's something I came up with lately.
It's a sample and hold ring modulator, using the input signal as noise source. The nice thing about this is that the oscillator frequency range lowers as the note decays. It also includes a noise gate based on the envelope section of the nurse quacky. The gate shuts down both the sample and hold oscillator as the VCO for the ringmod. No clock noise when you're not playing.

Here's the schematic and a little youtube fun.

Schematic: https://dl.dropbox.com/u/1849818/published_images/ringdroid_public.png

I don't have a layout for it. I'm in the process of making one, but it'll be double layered and I will order it online.

[Mark Hammer]

Lovely!  Can't see the schematic from work but I snuck a listen to the video, and I like what I hear.

From your description, it sounds like there is plenty of room for interesting mods.  For example, playing with the gate is promising, as is high/lowpass filtering of the "noise source".

[~arph]

Oh yes, plenty of mods imaginable.. there is an envelope signal available in there too. You could crossfade the clean and modulated signal based on the envelope. Using the 4066 gates, the spare schmitt trigger and an extra comparator.  Or one could use the saw wave outputs (pin 6, 7) on the PLL to create a PWM signal so we can have a variable duty cycle for the ring mod carrier signal. I wanted to keep it relatively simple.. 

[jrod]

Very cool!

[wavley]

Awesome!  Great job, I need to add this to my build list!

[slacker]

I'm the opposite of Mark can't watch the vid at work but had a peak at the schematic. Looks very interesting, nice work.

[~arph]

Thanks guys, your positive response means a lot to me.

I was thinking about adding pre/de-emphasis like in the rebote 2.5 to suppress carrier noise. Think it'll be worth it?

[Mark Hammer]

alparent was kind enough to send me the schematic.  I don't know how pre/de-emphasis would work in this context.  Normally, such "poorman's Dolby" adopts a very shotgu approach, by emphasizing and de-emphasizing ALL top end above some corner frequency, under the assumption that there will be multiple cumulative sources of hiss and other HF noise.  Carrier noise in RMs tends to be in a rather confined part of the spectrum.  I suppose one could use a bandpass and a complementary notch, but that would make the circuit noticeably more complicated.  It is also the case that often when you see pre-emphasis/de-emphasis networks, at least one of the stages is an inverting op-amp, since that makes implementation very easy (an RC pair in parallel with the input resistor).  Your use of non-inverting unity-gain stages here makes doing that require a more complex circuit.

Is there any way of using an opposite-phase version of the carrier to cancel out at a mixing node?

[reverberation66]

wow, that's the coolest thing I've heard in a long time, excellent work!!!

[~arph]

alparent was kind enough to send me the schematic.  I don't know how pre/de-emphasis would work in this context.  Normally, such "poorman's Dolby" adopts a very shotgu approach, by emphasizing and de-emphasizing ALL top end above some corner frequency, under the assumption that there will be multiple cumulative sources of hiss and other HF noise.  Carrier noise in RMs tends to be in a rather confined part of the spectrum.  I suppose one could use a bandpass and a complementary notch, but that would make the circuit noticeably more complicated.  It is also the case that often when you see pre-emphasis/de-emphasis networks, at least one of the stages is an inverting op-amp, since that makes implementation very easy (an RC pair in parallel with the input resistor).  Your use of non-inverting unity-gain stages here makes doing that require a more complex circuit.

Is there any way of using an opposite-phase version of the carrier to cancel out at a mixing node?

Thanks, seems we think alike. The carrier frequency here is smack down the middle of the audio spectrum, so that's why I didn't bother emphasis, but it did cross my mind and I wanted to hear your opinion on it. I do recall having seen a modulator pedal that used a 4066 and some complementary circuitry and another 4066 gate to cancel out the modulation noise, but I can't find the schematic no do I know what type of circuit it was part of.. I believe it was a ringmod too..  I do doubt it'll work here.

[slacker]

Are you thinking of this? http://www.aronnelson.com/gallery/main.php/v/BinOfBrett/Modulatron+sch.jpg.html

Sounds great now I've had a listen. Have you tried listening to the signal at IC2B pin 7, that should sound pretty cool assuming the sampling is at or near audio frequencies.

[~arph]

That was the circuit I was talking about!

Haven't listened to the pin yet.. Will do. Thanks for the suggestion.

[Keppy]

Wow, this is cool!

I built a ring modulator for the 1000 contest that had sample & hold & envelope follower, but it sounded nothing like this. Awesome!

[~arph]

Are you thinking of this? http://www.aronnelson.com/gallery/main.php/v/BinOfBrett/Modulatron+sch.jpg.html

Don't think this will do anything for the noise. It's just switching between inverted and non-inverted. There are still sudden changes, but then again..that is basically the whole point of a square wave modulated signal.

One could redo this with a sine VCO (simple OTA based) and a AD633. It will bump up the price and power requirements (need a voltage doubler ) should sound very good.

For the record, I also did a small test to replace the VCO and sample circuit with an attiny85 but my code probably sucked so bad it was terribly noisy. (i tried to do a PWM sine to modulate the signal)

[moosapotamus]

Sounds awesome! Nice work!

~ Charlie

[soggybag]

Awesome! great idea.

[~arph]

Thanks guys,

Here's the progress on the layout..



Fits a hammond B. I think I'll order some boards from itead studio, so if they turn out right I'll have about six or seven surplus boards. I'd be happy to send them on at cost.

[merlinb]

I like it a lot!

Can you explain what R21 is for? It looks to me like it just attenuates the bypass signal?

Also, should there be a pull-up resistor after C11?

[~arph]

Good questions. R21 indeed attenuated the bypass signal a bit. This is because the affected sound is perceived at a lower than unity volume. With R21 in place there is no volume jump when the gates jump in. I actually still need to do some testing and might give the first amplifier stage a little gain.

C11 should have a pull up, at least in theory. I either missed it migrating the breadboarded circuit to eagle or it's intentionally not there. Will check.

[John Lyons]

I'll take a board please, when you have them.