What should I know about the AD633 mutliplier before trying a new design?

[varialbender]

I'm hoping to design a trem/pan/ring mod using the AD633.
I've gone over the data sheet a few times, and I'd like to start trying to make something work.
Are there some basic things to do to make this thing work?
Paul Perry advised that I figure out the power supply first, so now I have a +/-15V power supply for it.
I think I'm also going to need to give it stronger audio input signal eh?
How should I go about doing that? Something basic? How much gain?
Best to ground the inputs that aren't in use?
Any recommended method of starting out?
Any working schematics using the AD633 to experiment with?

Any help is much appreciated.
Can't wait to get started.
Thanks a lot.

[goodrevdoc]

Ray Wilson over at musicfromouterspace.com has a ring mod circuit up, but its more tailored for synth apps. I guess it could be a start though. Here's the link:
http://www.musicfromouterspace.com/analogsynth/RingModulator/RingModulator.html
-justin

[varialbender]

Cool, thanks.
I found a few at eaced.com too.
Anyone have any other advice?

[TELEFUNKON]

datasheet & app-notes

[Transmogrifox]

With the AD633 you probably don't need to boost the input signal.  The input impedance is very high, and the gain can be controlled by your carrier amplitude.

Otherwise the function is just like the datasheet says:  (X1-X2)*(Y1-Y2)/10

For unity gain, make Y a 10V or greater amplitude, or you may put a gain of 10 on the guitar signal input.

For a simple multiplier, ground X2 and Y2, and use X1 for guitar, and Y1 for the carrier.  Ground Z.

Using the Z input, you could make a ring phaser weird thing by putting in some phaser stages and the output into the Z input.

Anyway, the chip is pretty straightforward if you understand multiplication and addition.

If you're real clever, you may be able to make the thing self oscillate so you don't even need an external oscillator for a carrier.

[varialbender]

I think I remmember reading a post, most likely by Paul Perry, about getting a much better S/N ratio when boosting the input signal. Can anyone confirm? Or am I imagining things? If it's doable without boosting, I'll try to at least get my circuit working first, and then deal with getting better sounds out of it.

Thanks transmogrifox, always great help. This circuit is going to be in with a phaser, and there'll be some strange mixing going on, but I don't think my configuration will make use of the Z input. Also, I don't know enough about oscillators to build one using this chip, but I'm hoping a different oscillator will give me what I'm looking for anyway.

[Transmogrifox]

Also, I don't know enough about oscillators to build one using this chip, but I'm hoping a different oscillator will give me what I'm looking for anyway.

Yeah, realistically a separate oscillator is the way to go.  The idea of using the AD633 as an oscillator was just for the cleverness of not having to add any additional active components.  More a novelty than anything.

You will certainly improve signal/noise ratio if you boost the input signal with a lower noise amplifier than the AD633.  I forgot to look at the noise specs, but I'll bet you would probably be better off using an op amp for gain.

It will still work without boosting the input. 

[Processaurus]

If your signal path is simple enough, go stereo!   I was messing with my alesis bitrman yesterday, and the ring mod is in stereo (panning L<>R at audio frequencies)  and sounded great, that + an expression pedal for the carrier rate (or a CV in) is a must, once you've tried it.

For an oscilator, you might look at the XR2206 (small bear has them) its a signal generator on a chip, it has triangle, square and sine waves available (I could se the sine paying off for the ring mod because of the lack of harmonics).  Best of all the amplitude and frequency can be voltage controlled.

[varialbender]

I'm going for a modified "Super Simple Dual LFO" from eaced.com.
It'll have pots for triangle/square, tilt, rate, and a switch for high/low rate.
I'm planning on making an OTA phaser, followed by this AD633, and making it super stereo.

Anyway, thanks a lot.
I just breadboarded, and it works!
Check my new thread for soundclips and schem.

[Paul Perry (Frostwave)]

I definitely recommend having the inputs (both signal and oscillator) as high as possible (before distorting) to maximise sig/noise. Plus, the X and Y inputs are a little different as regards bleedthru (you want to minimise the oscillator bleedthru)... can't remember which, see the data sheet. Also, if you are trimming, use multiturn trimmers, it's pretty pointless otherwise! (OK, you COULD use a smaller single turn trim & pad it out with resistors.. if you have the time & the patience!).