A Very Simple Circuit for Measuring pH

The circuit needs a standard 5 V power supply (without a negative rail) and outputs a signal of roughly 2.5 V ± 0.5 V or 0 V ± 0.5 V depending on the wiring and a suitable pH probe.

Signal out is common for both methods of wiring. The ground differs on type of application. For microcontrollers is better to use the out gnd-µC as that gives an always positive signal. The other method, suitable more for displaying results with the aid of a voltmeter, is to use out gnd-voltmeter which will output the signal in a few 100 mV above or bellow 0 V.

pH probes can be bought at various laboratory suppliers, Internet sites or eBay. I've bought a GX-E1 probe on eBay for approx. 12 €, which comes with two different calibration buffers. The rest of the circuit costs around 2 €.

Output of the circuit ideally follows the Nernst equation which gives a value of 0.059 V per pH unit. The lower the pH the higher the voltage and the opposite with basic pHs. Reason of the opposite sign is in how the buffer amplifier U1 works (inversion of signal).

Because the probe doesn't adhere totally to the ideal equation (usually the change in voltage is slightly less) we must calibrate the instrument with standardized buffers. Usually we use two at different pH values, but you can use more. Most common is the use of two buffers to get two voltages at known pHs which gives us then the factors k (roughly -0.059, depends on the probe and its age) and c (usually 2.5 or 0, depends on the wiring and resistors) of the equation pH = k V + c. Sometimes the equation is written as pH = β 0.059 V + c (where β is usually between 0.98 and 1). If using more buffers we get more points which can be translated to a more correct linear equation with use of the least squares method or to a polynomial equations of higher order.

If you lack a suitable buffer for testing you can prepare it yourself by mixing 2.5 g of citric acid with 0.8 g of NaOH in 100 mL of distilled water. The pH of such buffer will be roughly 5.01.

The operational amplifier U1 needs to be high-impedance as the pH probe has an internal impedance of approx. 100 MΩ. Suitable technologies are for example CMOS, JFET, FET. The higher the input impedance of the OP-AMP the less error there will be due to loading. In my test circuit I've used a Microchip MCP606 OP-AMP.

The resistors don't need to be exactly 1.5 kΩ. Any value around that would do. They are recommended to be the same, but there's no need to. The resistors form a voltage divider to bring the pH probe's voltage enough above ground, that it never goes negative. That means that as long as the equation 5 R2 / (R1 + R2) > 0.5 holds it should be fine.

For probe attachment is recommended to use a BNC female connector as that's the industry standard for pH probes. The impedance of the connector is not important as the signal is DC .

A few notes at the end. The probe ages, so with time it will need more time for the probe to settle for a number. The typical probe has to be kept in a solution of 3-4 M KCl. It cannot be fully submerged in any liquid. The glass bulb at the tip of the probe is sensitive to mechanical stress. All probes suffer from the so called sodium error which gives erroneously lower pHs at higher pHs if in a solution with high sodium ions.

[1] I've tested with 2.52 g citric acid and 0.80 g NaOH which gave me a pH of 5.01.