I want to start building full wave precision rectifier, but the problem is I do not know which one I should implement. Can anyone recommend which Full Wave Precision rectifier I should build?

Sorry for asking this question, but I do not know the difference between the two implementations.

 

The circuits perform the same function in two different ways. I don't think there would be much difference in performance between the two. A Spice simulation of the two would point out any differences, which would likely show up at the upper frequency limits of the circuits.

One small difference is that the first requires 5 resistors and the second requires 4 (possibly 3, since R1 is not mandatory in the second for dc coupled sources with no series capacitor).

 

Both circuits are perfectly functional. Personally I would go with the bottom circuit since it simpler, has less components and the input impedance can be set very high.

 

Both circuits are perfectly functional. Personally I would go with the bottom circuit since it simpler, has less components and the input impedance can be set very high.

Yeah, the input impedance of the top scheme is 6.7k as drawn.

 

I've built a variety of absolute-value circuits from various sources, and they all seem to have the common flaw with non-linearity near zero. If you're planning to use those circuits for precision DC application...don't.

The most well-behaved I came across was from an analog devices note. It used two rail-to-rail opamps and no diodes.

The link (figure 3):
http://www.analog.com/library/analogDialogue/archives/44-04/absolute.html
good luck

 

The most well-behaved I came across was from an analog devices note. It used two rail-to-rail opamps and no diodes.

The link (figure 3):
http://www.analog.com/library/analogDialogue/archives/44-04/absolute.html
good luck

Never heard of it up to now. Good to give a try.

Gracias.

 

This also seemed interesting.

http://www.diodes.com/zetex/_pdfs/3.0/appnotes/apps/an55.pdf

 

I have used this circuit at 60hZ--does poorly at higher frequencies because the op amps have to recover from saturation each half-cycle. It must have load resistor R4 because there is a sneak path through R1 & R2 when the input is positive. R3 and D1 are necessary to keep the U1B from 'latching' when the op amp input goes negative beyond its common mode input range. It has the advantage of single supply operation.

 

I've built a variety of absolute-value circuits from various sources, and they all seem to have the common flaw with non-linearity near zero. If you're planning to use those circuits for precision DC application...don't.

The most well-behaved I came across was from an analog devices note. It used two rail-to-rail opamps and no diodes.

The link (figure 3):
http://www.analog.com/library/analogDialogue/archives/44-04/absolute.html
good luck

The AD spice model is bad for simulating this circuit. With 0V as the negative rail, the output swings to about -80mV when it should stop at or above the rail.