X² analog multiplier/squarer jam (clips)

[g3rmanium]

Hey,

been playing with my X² analog multiplier/squarer again. I added a COT50 and



Sound clips!

Enjoy.

[nordine]

Crunchy as fresh cereals! 

sound cool, but that AD chip, so expensive 

[alanlan]

Nice sounds - I'm after a multiplier to do some heterodyning and the device you used may do the trick.

[g3rmanium]

Crunchy as fresh cereals! 

sound cool, but that AD chip, so expensive 

Come on, I could have used the AD835 or so 

[soulsonic]

I like it - I love the metallic tones. Metal springy droning things are some of my favorite sounds to hear. The textures are very synth-like as well. This is a toy I will have to play with.

[g3rmanium]

This is a toy I will have to play with.

Schematics:

[tommy.genes]

I like the sound and I will definitely have to play with it.

But where did you get those cool, little, yellow resistance boxes?

-- T. G. --

[Steben]

The response function is XY. very interesting.
Let's go into math. If one would split X in two, like you do, yet let one go into the chip (=X) and another one into an inverting attenuator "f(X)" (amplifier gain <<1).
This amplifier "aX" is biased to unity voltage for the multiplier chip. The result is X + aX. If "X" goes up "aX" goes down.
Now you have a very complex way to make a simple amplifier.

But what if you use X² as the input of "f(X)"?
Than the function is X(1+aX²)= X+aX³. The gain would be decreasing as the voltage goes up!
You should have a wave compressor, aka overdrive simulator!

no?

Or another option is to simulate the quadrant aspect in triodes. With a little attenuation you get X*aX= aX² instead of X². Which would give assymetric gain without clipping more easily.

[soulsonic]

Just to clarify something in the schematic; R1 is connected between pins 6 & 7 and R5 is connected between pin 6 and DC Ground? This is a feedback network between the output and the Z Sum input? For the purpose of gain?
Thanks!

[g3rmanium]

Just to clarify something in the schematic; R1 is connected between pins 6 & 7 and R5 is connected between pin 6 and DC Ground? This is a feedback network between the output and the Z Sum input? For the purpose of gain?

Exactly. It's explained in the data sheet of the AD633. If you vary the resistance of R5, you can increase or decrease the output level. I think I am using 1.5 K now which is close to unity gain or so.

[g3rmanium]

But where did you get those cool, little, yellow resistance boxes?

From here. They used to have them for 5 EUR or so.

[tommy.genes]

From here. They used to have them for 5 EUR or so.

Thanks. They seem to be 10,49 € now, and I bet shipping from Europe would not be cheap. Now that I've seen them, I'll be more able to spot something similar in the States.

-- T. G. --

[Austin73]

just had a quick look at this and yes it sounds rather cool (i presume the other fx really help the sound) but what is IC2? And is there any controls for this other than volume?

Cheers
Aus

[g3rmanium]

just had a quick look at this and yes it sounds rather cool (i presume the other fx really help the sound) but what is IC2? And is there any controls for this other than volume?

Yeah sorry IC2 isn't mentioned in the diagram. It's an XP Power IA0515S DC/DC converter. Other types should also work (I think I have some at home that I could try out).

No other controls except volume.

[Paul Perry (Frostwave)]

I like the AD633 - I use it in my commercial Frostwave Blue Ringer - but, if anyone wants to get into squaring on the cheap, you can do it with a CA3080, or a half a LM13700.
Look at the LM13600 or LM13700 application notes to see how.
It's a bit tricky, but the approach is this:
The output of an operational transconductance arranged as an amplifier, is proportional to the product of the input signal and the bias current.
Now if you let the bias current be proportional to a second signal voltage - by feeding the signal through a resistor - and in addition you add some offset current so that the bias current is always positive - then the output is now proportional to the first input signal multiplied by (the second input signal plus a constant offset).
So your output is the product you want , plus a constant times the first input signal, which you can subtract out.

It's tricky to get this working right, a lot of tweaking, plus the gain depends on the power supply voltage, but it CAN be done.

[g3rmanium]

I like the AD633 - I use it in my commercial Frostwave Blue Ringer [...]

Btw: I think there is some internal limiting/shutdown in this chip. If you drive it too hard, it will stop working for a while.

Can you confirm this?

[Paul Perry (Frostwave)]

I don't believe thermal shutdown is mentioned as a 'feature' for the AD633.
But a precision analog chip like this is often more complex internally that one would expect.. I can well imagine some internal biasing arrangement falling over when hot.
But, how would one cook an AD633??? It isn't a chip that runs 'warm' in normal use.