Interesting pseudo-ringmod Idea

[Eddododo]

First off, a BIG thanks to R.G. / geofex.com, not only for inspiration for this, but for making up about 65% of my electronics knowledge at this point in my journey.   I constantly read and reread his entire site, understanding a little more each time.


ANYWAY
I was thinking about a ring-modulation-esque effect using LED/light resistors to pan between inverting  and non-inverting opamp stages.  There are obvious things I left out for simplicity, and several expansions and tweaks that are floating around my head, and lurking in the physics wilderness waiting to eat me; I hope merely to offer a concept, and hopefully my diagram makes SOME sense.



I 'know' there are LED's with about 5nS 'on' time, and if the LDR's can 'sweep' quickly as well, there could be an almost square-carrier signal. If the sweep' happens more slowly, perhaps there can be an interesting harmonic content as a result (i guess it could almost be sinous?).  
  What i am curious about is a switching method for the polarity affecting the parallel LEDs- is there a way for an AC signal via an oscillator to drive them? (I guess this would need to be able to swing to negative voltage ; requires a bipolar supply?) . I am open to more conventional switching ..


I guess this idea is a little more convoluted than elegant, but I intend on seeing it through- I will be pestering you guys thoroughly!

Thanks for your time


p.s. I am coming back with a picture, I didnt realize i could not upload directly

[psychedelicfish]

The one thing I would suggest is to use a different switch, LDRs have relatively slow response times (slow enough to be useless for ring mod frequencies). You could use a P-channel and an N-channel MOSFET , and connect the gates together and apply your carrier signal (a square wave), and use the drain and the source of each MOSFET as the terminals of your LDR

[samhay]

A ring modulator works by multiplying two signals together. Unelss your schematic is simplified more than you intended, it shows you summing two signals, which will probably generate some form of tremolo effect.

If you want to do your mixing a different way, you might be able to do the mixing using voltage divider consisting of a single LDR (or other variable resistor) and a fixed resistor. It isn't a perfect solution, but is a much smaller parts-count one - http://www.diystompboxes.com/smfforum/index.php?topic=102414.0

[Eddododo]

A ring modulator works by multiplying two signals together. Unelss your schematic is simplified more than you intended, it shows you summing two signals, which will probably generate some form of tremolo effect.

If you want to do your mixing a different way, you might be able to do the mixing using voltage divider consisting of a single LDR (or other variable resistor) and a fixed resistor. It isn't a perfect solution, but is a much smaller parts-count one - http://www.diystompboxes.com/smfforum/index.php?topic=102414.0

Well.. the idea is that the mix is swept/switched at a high enough rate that it (the rate of 'panning') would be in the audio frequency range; it would not be a ring modulator, but it would theoretically add some new and hopefully ringmod'ish harmonics to the signal.  Think of some of the diy ringmod designs- a few of them are basically tremolos for which the 'LFO' goes well into the audible range; By summing two opposite-phase signals, I feel like it would produce an interesting form of amplitude-modulation-related harmonics, even if by virtue of the imperfections of the LDR's, or perhaps eq/filtering differences [intentional or otherwise] between the two Op-amp signals.


Now i need to explore transistor switching etc... its been something I look AT alot, but I have not played with electrical switching/lfo's much AT ALL...
hint hint.. a great thread suggestion would be muchly appreciated, even if I probably find it before someone responds :p

Thanks

[PRR]

LDRs are MILLI-seconds, not nanos.

The highest useful carrier will be 20Hz, 50Hz, sumthing like that.

> it shows you summing two signals

I think one is inverted. This WILL modulate. If you beat 33hz signal against 22hz carrier there will be a 55Hz tone in the output (also 11Hz and others). So the right idea, but *probably* too slow for human music.

[samhay]

PRR - you're absolutely right

Eddododo - as a slightly faster alternative to LED/LDRs, you may want to take a look at at the LED FET optocouplers like the H11F1M (or, even better, one of the dual package ones). Ray Ring recently used one of these in a ring modulator, which seems to work quite nicely - http://circuitsaladdotcom.files.wordpress.com/2013/06/simple-ring-modulator1.gif

[midwayfair]

LDRs are MILLI-seconds, not nanos.

The highest useful carrier will be 20Hz, 50Hz, sumthing like that.

> it shows you summing two signals

I think one is inverted. This WILL modulate. If you beat 33hz signal against 22hz carrier there will be a 55Hz tone in the output (also 11Hz and others). So the right idea, but *probably* too slow for human music.

Came here to post something like this (minus numbers) based on PRR's previous posts.

OP, if you don't care about distortion (and if it's a ring mod, you probably don't!), you can use FETs in place of the LDRs.

Use a dual op amp oscillator like what's in the tremulus lune; replace the 10nF cap in the first op amp with a smaller cap. Several people have done that to make ring moddish sounds, but it's still not quite there because the LDR is just too slow. So instead of hooking up the output to an LED, decouple it and you can send that to the gate of one FET. I'm sure there's a way to invert the op amp signal. Or use RG's op amp setup that blinks LEDs out of phase.

I tried something like this for a tremolo and it sounded bad for tremolo use. I think ~arph was able to get it to work, though. I'd search through his old posts for his version of the Cardinal Tremolo.

I'd post links etc. but it's not convenient to do so at the moment.

Obviously this is all theoretical with the set-up you've posted.

[earthtonesaudio]

Search for the "switch hitter" opamp thing.

Or check out this example.

[Eddododo]

ALOT of good stuff to look through here, fellas.. after a little digging, I'll have something useful to say.

I should post here more, instead of lurking. This is a VERY helpful community- much thanks

[Eddododo]

With regards to the complimentary mosfets-

Would I basically run each signal path through the mosfets as a 'common gate' amplifier?
Would I encounter problems from the low impedance seen at the source?

I could make it much simpler and just use a single switch mosfet that inverts an op-amp, but I have made A LOT of progress and gone through a conceptual evolution, so long story short, I like the idea of 'square-panning' two inverted signals, rather than rapidly inverting a single signal..

Thanks for any help!

[Eddododo]

in other news, while trying to to goober my way through designing a circuit, i kind of reinvented the green ringer. Which led me to the green ringer, which uses a bias voltage through the rectifying diodes... interesting stuff.

So the evolution, and least conceptually, or more simply by the knobs and switches i want, goes something like this.  

A ring-moddish box, that contains three flavors of ringish harmonics
       
        1) the tremolo-ish circuit that this thread was started over
        2)my mosfet green ringer re-imagining (ill post some planning schematics soonish)- the catch? the same oscillator driving the tremolo circuit will be fed to the FRONT of this circuit, for those harsh whacky overtones that are caused by polyphony in these kinds of circuits
         3) a gain-starved rectified 'jawari' style .. um.. thing

Any good hints for a clean blend on some of this mess? the 'tremolo' ringer should be a pain in the ass, what with a rapidly inverting signal to mix with clean...

Thanks, part 1000

[Eddododo]

With regards to the complimentary mosfets-

Would I basically run each signal path through the mosfets as a 'common gate' amplifier?
Would I encounter problems from the low impedance seen at the source?

I could make it much simpler and just use a single switch mosfet that inverts an op-amp, but I have made A LOT of progress and gone through a conceptual evolution, so long story short, I like the idea of 'square-panning' two inverted signals, rather than rapidly inverting a single signal..

Thanks for any help!

Sorry to spas-post, but I just realized how dumb that was-
There is no reason to not use common source (right? )

[Digital Larry]

Semi-OT thought here... I have been doing a lot of experimenting with the Spin FV-1 DSP chip in the past 6 months or so.  Making a ring modulator with that is incredibly easy.  One thing I noticed is that if you put your audio signal through a low pass filter prior to ring modulating it, it really helps create a clearer tone with less "hash" all over it.  You still get lots of ring mod bizarreness.  So you might consider adding something like a tunable LPF for adjustable "flavor".

[Eddododo]

Semi-OT thought here... I have been doing a lot of experimenting with the Spin FV-1 DSP chip in the past 6 months or so.  Making a ring modulator with that is incredibly easy.  One thing I noticed is that if you put your audio signal through a low pass filter prior to ring modulating it, it really helps create a clearer tone with less "hash" all over it.  You still get lots of ring mod bizarreness.  So you might consider adding something like a tunable LPF for adjustable "flavor".

I will look into that chip!

I was already thinking about that problem specifically, glad to know I am on a good path.. Another thing I would like to try is asymmetrical 0^o and 180^o signals; one side clips one way, the other, or have them filtered differently. This is for the 'tremolo' type ring circuit..

[Eddododo]

OK so here is my rough schematic of what I am thinking..
I am VERY young at this, so be kind
Some components are known/obvious, some are TBD as I get further and further from knowing nothing about what I am doing.

The white mosfet stage outputs a 0^o and a 180^o output

The blue chunk is an astable square wave generator;
The Purple chunk around it is a crude attempt at an envelope-controlled gate for said square wave

The green stages are switched 'on' by the squarewave output. Note they are a positive AND a negative channel

The Yellow switches are shown in 'fringer' position.
The 'fringer' is my accidental reinvention of the famed Dan armstrong's green ringer. I was drawing mosfet stuff and thought about a geofex article showing positive and negative phase outputs-
later on i stumbled into the green ringer circuit, and was tickled... i definitely intend to borrow that diode on-bias trick..
Remember that square wave? remember that gating that was done? Well the waveform is mixed in at the front of the fringer circuit- If i am lucky, it will produce those AWESOME pseudo-ringy harmonics that so many rectified octavers get from polyphony.


The Red switch is to lower gain in the 'fringer' stage to try to get a 'Jawari' type effect



I CLEARLY owe alot to:
Jack Orman's http://www.muzique.com/schem/mosfet.htm
R.G's GEOFEX, specifically his MOSFET ARTICLES

I fully intend to figure this out and not copy and paste values and exact topologies from these guru's- consider this a conceptual diagram