One-Quadrant Multiplier

[Tim__x]

I'm in need of a simple one-quadrant multiplier for a bias generator.
I'm aware of the standard opamp with a pn junction in the feedback loop based ones, but I'd rather have a simpler discrete solution.
I don't need much linearity, I don't need thermal stability, I don't need accuracy, I don't need a divisor input (Z). All I need is something that can give me approximately X*Y with some speed. Any type of input and output is fine, voltage, current, referenced to any rail, or even to two different rails, any will work.

Any ideas?

[clem_o]

Voltage into the conduction zone of a diode... (?) Unfortunately, input voltage range would need to be small, and somehow the tempco of the diode would have to be compensated...


Cheers

[Elvee]

Hi

The simplest practicable thing I can think of is a differential amplifier whose tail is fed by a current representing one of the multiplicand. The other multiplicand is fed as a voltage to the differential pair. This voltage has to be scaled to be limited to tens of mV at most. Obviously, you'll need some ancillary conversion/level shifting/amplifying circuitry around to inject and extract your signals.
Maybe an OTA would be simpler after all? Or a modulator of the MC1496 variety. A voltage divider including a FET maybe, but you'd problems of linearity and large signals handling capability.
LV

[phase_accurate]

Maybe an OTA would be simpler after all?

This would indeed be a good solution. The only disadvantage is that it might perform better than needed !!!

Regards

Charles

[peranders]

Basicly a multiplier is rather easy to design if ytou have low demands.

http://www.national.com/ms/LB/LB-25.pdf

http://search.national.com/iphrase/q...ery=multiplier

http://www.intersil.com/cgi-bin2/Msm...ERY=multiplier

I used Google here and I got lot's of hits. You can also check ICL8043