7

The design for the rectifier is rather clever, and comes from TI's TIDU030: Precision Full-Wave Rectifier, Dual-Supply.

It looks as follows:

schematic

simulate this circuit – Schematic created using CircuitLab

R2=R3=R4, and this rectifier has gain of +/-1. The architecture is nice as it keeps both op-amps operating closed-loop, without saturation. D1 and D2 can be Schottky microwave RF diodes to lower crossover distortion.

Could it be made to have inherent gain greater than 1? It's not as simple as changing the R3/R2 ratio, since the gain for the positive input would still remain +1.

2 Answers2

11

It turns out it's possible by adding R0 and R1 to control the non-inverting gain.

schematic

simulate this circuit – Schematic created using CircuitLab

When the input is positive, D2 conducts, D1 is reverse-biased, and the equivalent DC circuit becomes:

schematic

simulate this circuit

When the input is negative, D1 conducts, D2 is reverse-biased, and the equivalent DC circuit is:

schematic

simulate this circuit

We want the gains in each case to have the same magnitude:

$$ A_V = 1+\frac{R_1+R_2+R_3}{R_0} = \left( 1 + {R_1 \over R_0} \right) \frac{R_3}{R_2} $$

We can substitute \$r_i=R_i/R_0\$, and

$$\begin{aligned} A_V &= 1+ r_1 + r_2 + r_3 \\ A_V &= (1+r_1)\frac{r_3}{r_2} \\ \\ A_V \frac{r_2}{r_3} &= 1 + r_1 \\ \\ A_V &= A_V \frac{r_2}{r_3} + r_2 + r_3 \\ A_V - r_3 &= \frac{r_2}{r_3} \left( A_V + r_3 \right) \\ \text{Let }k &= \boxed{ \frac{A_V-r_3}{A_V+r_3} } \\ r_2 &= \boxed{ r_3 \cdot k } \\ \\ r_1 &= A_V \frac{r_2}{r_3} - 1 \\ r_1 &= \boxed{ A_V \cdot k - 1 }. \\ \end{aligned}$$

Since \$r_1\$ must be non-negative, we have the constraint \$A_V \cdot k \ge 1\$. \$R_0\$ and \$r_3\$ are otherwise arbitrary.

Example: $$\begin{aligned} &\text{Let } A_V = \boxed{10}, \ r_3 = \boxed{5}, \text{ then} \\ \\ k &= \frac{A_V-r_3}{A_V+r_3} = \frac{10-5}{10+5} = \boxed{ \frac{1}{3} }\\ r_2 &= r_3 \cdot k = 5 \frac{1}{3} = \boxed{ \frac{5}{3} } \\ r_1 &= A_V \cdot k - 1 = 10 \frac{1}{3} - 1 = \frac{10}{3} - 1 = \boxed{ \frac{7}{3} }. \\ \end{aligned}$$

Letting \$R_0 = 1{\,\rm k\Omega}\$, we get

$$\begin{aligned} R_3 &= r_3 R_0 = 5 {\,\rm k\Omega} \\ R_2 &= r_2 R_0 = \frac{5}{3} {\,\rm k\Omega} \\ R_1 &= r_1 R_0 = \frac{7}{3} {\,\rm k\Omega} \\ \end{aligned}$$

A DC simulation confirms the derivation for this case:

schematic

simulate this circuit

2

I have often used a similar, but simpler, precision rectifier circuit, which has the advantage of using a single supply and only 3 resistors, but it has a gain of 0.5. It also requires a low impedance source and a high impedance load. A second op-amp can provide as much gain as desired as well as driving a lower impedance load.

Precision Rectifier

PStechPaul
  • 7,195
  • 1
  • 7
  • 23
  • I’ll have to put this together and compare. I’m interested in getting best performance using general purpose op-amps, and create a sort of standard circuit that people who use eg. LM324 can use and fully leverage the performance of the op-amp (for how little there is of it, it matters to get the most). – Kuba hasn't forgotten Monica Aug 13 '22 at 16:38