analog multiplier

....... Phil

View\\fixed font .---------+------o | | V+ .-. .-. | |R1 | |R1 | | | | ___ '-' '-' .--|___|-. | | | R2 | | | | |\\ | | +----+---|-\\ | | | | >-+--o +---------)----+---|+/ Vout | | | |/ | | | ___ | | '--|___|-GND |/ \\| R2 o-| LM394 |-GND Vx |>

The only method I know of uses the fact that log(f*g) = log(f) + log(g)

hence you convert your signals to there logs then add and take the exponential.

i.e.

f*g = exp(log(f) + log(g))

You can make logarithmic and exponential output configurations (employing a diode in the op-amp circuit), and a summing amplifier. So you could put together multiplying and dividing circuits by combining them appropriately. Have a look here:

Also google "Gilbert cell multiplier."

John

Hi, Bob. You can, but I wouldn't.

Multipliers made from op amps and discrete components are subject to major drift and non-linearity issues. Also, less complex designs have the limitation of only working in one or two quadrants (meaning one or both of the multiplier inputs can only be positive).

You can get a good four-quadrant IC analog multiplier (Analog Devices AD633) for several dollars in single quantity. It has differential inputs, better than 2% error across the input voltage range, works from DC to well above audio frequencies, and doesn't require *any* external components. Perfect for newbies.

If you have requirements for higher frequency, more accuracy, or special functions (division, square root), other products which can help you are available from Analog Devices and Burr-Brown (now TI).

If you're thinking about actually making one instead of just answering a question, the headache of trying to cobble together a working multiplier from discrete components just isn't worth it.

Good luck Chris

Others have given you good info regarding multipliers that use device nonlinearities. These are the methods of choice for high-speed applications (Gilbert cell especially). But for low-speed situations it's possible to build a multiplier based on pulse-width modulation. The basic idea is that you feed a fast triangle wave into one input of a comparator, and one of the values to be multiplied into the other. The output is a rectangular wave whose duty cycle is proportional to the input value. You then use this to modulate (chop) the second input, and average the results (low-pass filter) to get the product output. This works well at low frequencies and can be made very linear. As you try to move to higher frequencies things get trickier.

Best regards,

Bob Masta dqatechATdaqartaDOTcom D A Q A R T A Data AcQuisition And Real-Time Analysis