Op amp level shifter, 0 to 2.5 V into −3 to −1.5 V: 20 k, 7.5 k, 12 k, 10 k on E24, ±68.5 mV at 1%

A 0 to 2.5 V DAC into a −3.0 to −1.5 V GaN gate bias is Vout = 0.6·Vin − 3 V: R1 20 kΩ and R2 7.5 kΩ divide the DAC into the + input, and RF 12 kΩ with RG 10 kΩ from the 2.5 V reference set the gain and the offset, all on E24. With an ideal op-amp it reads −3.000 V and −1.500 V, 0.0 mV off at both ends.

The Rescaler wizard on 0 to 2.5 V in, −3.0 to −1.5 V out: three E24 sets and their error, then the best one placed and solved.

With 1% resistors and a 0.1% reference the worst corner is −3.064 V to −2.938 V at code 0 and −1.569 V to −1.432 V at full scale: a 137 mV spread at the gate.

Open in Op-Amp Lab

Change this: in the error budget, set the R tolerance to 0.1% and the worst case at code 0 shrinks from ±63.0 mV to ±9.0 mV.

Key numbers

QuantityCode 0, Vin = 0 VFull scale, Vin = 2.5 V
Output, ideal op-amp−3.0000 V−1.5000 V
Error from the target0.0 mV0.0 mV
Worst case, linear (1% R, 0.1% reference)±63.0 mV±68.5 mV
Worst corner, low−3.0637 V−1.5691 V
Worst corner, high−2.9377 V−1.4322 V
Worst case, linear (0.1% R)±9.0 mV±9.5 mV

Every result here is the Op-Amp Lab's own, computed when the site was built from the circuit the button opens, and checked against a hand calculation. How the engine is checked.

The circuit

Vout = Vin·R2/(R1 + R2)·(1 + RF/RG) − Vref·RF/RG

PartValueWhere
RG10 kΩReference to the − input
R120 kΩInput to the + input
R27.5 kΩ+ input to ground
RF12 kΩOutput to the − input
VREF2.5 VReference

The numbers use an ideal op-amp on ±15 V rails. Set U1's Vos and Ib on the sheet and the error budget adds their share; its Monte Carlo run gives the spread of 1000 boards.

How the four values follow from the two points, and the same circuit on a 3.3 V DAC and other gate windows: op amp level shifter design.

Other solved op-amp circuits

Question about this circuit, or a correction?