Fullscreen HTML5Med Res (??MB)HTML4

Using Electronic Modules with Jim Rowe Analog Liquid pH Meter This module is designed to form the basis of a liquid pH meter, for testing the acidity or alkalinity of things like the water in fish tanks or swimming pools, or the liquid in a vat when you’re making beer or wine. It comes complete with two pH sensor probes, and can be easily hooked up to an Arduino or other microcontroller to form a pH meter. T he ‘pH’ of a liquid indicates how strong of an acid or alkali it is; or perhaps it’s midway between the two and thus ‘neutral’, like distilled water. In my school days (long ago!), we used strips of ‘Litmus paper’ to test this – the paper changed colour when it was dipped into a liquid, with the colour providing a guide to whether the liquid was an acid or an alkali. Nowadays, though, this kind of testing is done using a more precise device called a pH Meter. The concept of ‘pH’ was first proposed in 1909 by Soren Sorenson, a Danish chemist working at the Carlsberg Laboratory. It is generally regarded as indicating the inverse concentration of hydrogen (H+) ions in an aqueous liquid, or the ratio between H+ ions and OH− (hydroxide) ions in the liquid. As shown in Fig.1, the pH scale runs from 0 to 14, with 0 representing an extremely strong acid, like battery acid, and 14 representing an extremely strong alkali (or base), like liquid drain cleaner. In the middle of the scale (pH = 7) is the neutral point. The first electronic method for measuring pH was developed in 1934 by Arnold Beckman, a professor at the California Institute of Technology, to help local citrus growers test the pH of lemons they were picking. He formed a company to manufacture and market pH meters, and since then, pH meters have been used in a wide range of industries. They include testing water quality, swimming pool maintenance and wine or beer brewing. They are also widely used in healthcare and food processing. The pH probe The key component of a pH meter is the pH probe. This contains two electrodes, designed so that when they are both in contact with the liquid to be tested, a small voltage difference is developed between them. The polarity and amplitude of this voltage difference is proportional to the pH of the liquid. Originally, pH meters used two separate electrode probes: a hydrogen ion sensing probe and a reference probe. But nowadays, most pH meters use what is called a ‘combination’ probe, which includes both electrodes in a single probe body, shown in Fig.2. The main H+ sensing electrode is Fig.1: the table on the left shows the pH scale from zero to 14 with hydrogen and hydroxide concentrations (pH values normally lie in this range). The right-hand table shows example liquids with their typical pH values. 36 Silicon Chip Australia's electronics magazine siliconchip.com.au inside a small central glass tube which usually ends in a small spherical bulb of very thin, porous glass. This sensing electrode is generally made of silver, with a very thin silver wire used to make the electrical connection to it. The interior of this H+ sensing electrode tube is filled with a solution of silver chloride (AgCl), its electrolyte. The reference electrode is similar in construction, but housed in the outer part of the probe body and surrounded by a different electrolyte; usually, a solution of potassium chloride (KCl). This area of the probe ends in a porous ‘reference junction’ around the central glass tube, just above the glass sphere housing the main H+ sensing electrode. As a result, when the bottom of the probe is submerged in a liquid, a voltage difference is generated between the two electrodes. The small hole shown in Fig.2 near the top of the inner glass tube is provided because some of these probes are designed to allow the H+ electrolyte solution to be ‘topped up’ from time to time, if it has seeped away through the porous sensing membrane at the bottom. Many pH probes do not offer this feature, though. The electrical output of an ‘ideal’ probe is shown in Fig.3, which plots the voltage difference between the H+ electrode and the reference electrode for liquids with a pH varying from 0 to 14. The voltage rises from 0mV at pH = 7 to over +400mV for pH = 0 (red line), while it falls to beyond -400mV for pH = 14 (blue line). Both the red (acid region) and the blue (alkali region) lines have a slope of -59.16mV per pH unit, assuming the liquid being tested is at 25°C. So an ideal composite pH probe has a linear output voltage swing of from +414.12mV to -414.12mV for the pH range of 0-14, swinging positive for acids and negative for bases from 0mV at the pH = 7 neutral point. The output from a pH probe has a very high source impedance, typically between 10MW and 100MW. So it needs to be connected to a very high impedance load to avoid attenuation. (analog-to-digital converter) inside a microcontroller unit (MCU) like an Arduino. The module shown in the photos is a low-cost unit we obtained from an AliExpress vendor in China, “Mi Yu Koung”. It comes complete with two pH sensor probes (one ‘refillable’ and the other not), each with a 1m-long cable fitted with a BNC plug. They also came with a small container of electrolyte for topping up the refillable probe, four 10mm-long M3 screws and four matching nuts, for mounting the module. There was also a mounting nut and spring washer for the module’s BNC socket, providing the alternative option of mounting it behind a panel. This module ‘kit’ cost us $11.52 plus $9.75 for shipping, for a total of $21.27. We found an identical kit is available from an eBay supplier called Garmenthouse No.1, for just under $20 with free delivery. We found that another module called the DFRobot Gravity pH Meter V2.0 is available in Australia, from suppliers such as Core Electronics and element14. This one comes with only one pH probe, for about $82.00 plus $10 for express delivery. Module circuit details Returning to the module shown in the pictures, it is on a 42×32mm PCB with the input BNC socket protruding from one end, and a 6-pin SIL output header at the other. The full circuit is Fig.2: an example of a combination probe, which has both electrodes in a single probe body. The main electrode is located inside a very thin, porous glass membrane. Fig.3: the electrical output of an ‘ideal’ probe should be a linear change in voltage relative to pH as shown in this graph. The sensor module The job of the pH meter module is essentially to amplify this low output voltage swing from the probe, boosting it to a level where it can be measured accurately by the ADC siliconchip.com.au Australia's electronics magazine September 2023  37 Fig.4: the circuit diagram for a cheap pH module which was purchased from AliExpress. The top half of the circuitry involves processing the signal from the pH probe, while the lower half provides an analog signal indicating the module’s temperature. shown in Fig.4, but don’t be fooled by its apparent complexity. The only section involved in processing the signal from the pH probe connected to CON1 (the BNC socket) is the top half, involving shunt regulator VREF1, op amps IC1a and IC1b and, to a lesser extent, IC2a. The lower half of the circuit, involving IC2b, TH1, IC3a and IC3b, is purely to provide an analog signal indicating the temperature of the module, via pin 6 (TO) of CON2. That could be useful as a way to adjust for the temperature’s effect on the pH readings, although the module’s temperature won’t necessarily be the same as the temperature of the liquid being tested. The pH+ electrode signal from the probe via CON1 goes directly to input pin 3 of op amp IC1a. IC1 is a TLC4502, a dual self-calibrating precision CMOS op amp with an input bias current of only 1pA (0.001nA). It therefore provides very little loading to the signal from the pH+ electrode. Since IC1a has negative feedback applied via the 20kW and 10kW resistive divider, it amplifies the pH+ signal by three times, sending the amplified signal to pin 4 (PO) of output connector CON2. The purpose of the circuitry at upper left, involving VREF1 and IC1b, is 38 Silicon Chip to generate a ‘bias offset’ voltage to the pH− reference electrode of the probe. Since the output voltage from the probe can swing either positive or negative with respect to zero, that could be a problem for IC1a since its output can only swing between +5V and ground (0V). By feeding a bias voltage to the probe’s P− reference electrode, the These two buffer solutions were purchased from an Australian supplier and came in 125mL containers. Most buffer solutions will have tolerance of ±0.01pH, which explains the labelling of 7.01 for a 7pH buffer. Australia's electronics magazine pH=7 ‘zero’ voltage of the P+ electrode is shifted upwards so that the output voltage of IC1a at pH=7 moves up to +2.5V, allowing it to swing up or down without problems. This also means the ADC monitoring the output signal doesn’t need to be able to deal with negative voltages. Trimpot VR1 and the 5.1kW resistor reduce the 2.5V output of VREF1 to around 0.83V, which when amplified by three times by IC1a, gives the correct 2.5V offset. The offset voltage is buffered by voltage follower IC1b before being fed to the pH− probe connection of CON1. If the pH=7 output of the probe is exactly zero (as with an ideal probe), and the gain of IC1a is exactly three times, the bias voltage applied to pin 5 of IC1b would need to be exactly 2.5V ÷ 3 = 833mV. However, with a real probe and real resistors that differ from their nominal values, that might vary. VR1 allows the bias voltage to be adjusted until the output of IC1a is close to +2.5V when pH = 7. The circuitry at centre right in Fig.4, around VR2, IC2a and LED1 detects when the output voltage from IC1a rises above a certain threshold. IC2a is connected as a simple comparator, comparing the output of IC1a siliconchip.com.au Fig.5: a plot of the nominal output voltage over the full pH range for the module, taken at pin 4 of CON2 (PO). with a reference voltage set by trimpot VR2. So when the output voltage of IC1a rises just above that level, the output of the comparator will drop to near-zero and LED1 will light. The voltage level at pin 5 (DO) of CON2 will also drop to near zero, allowing the situation to be detected by the MCU if required. At the same time, LED2 simply acts as a power-on indicator. Fig.5 is a plot of the nominal output voltage of the module at CON2 pin 4 (PO) for the full pH range from pH=0 to pH=14. It should provide an output voltage of 2.50V for a pH of 7.0, rising to 3.74236V for a pH of 0 and falling to 1.25464V for a pH of 14. DFRobot Meter differences Before moving on, I should mention that the DFRobot Gravity pH Meter V2.0 module mentioned earlier only provides an amplified analog version of the pH probe’s output, with no added ‘frills’. It also allows the pH− output of the probe to be connected directly to ground. This is done by using a DC-DC converter to provide the main op amp with a -5V supply as well as the +5V supply. It is also provided with a mini polarised 3-pin output connector (instead of the 6-pin SIL header), siliconchip.com.au plus an output cable with a matching 3-pin plug. In addition, it comes with four small containers of pH standard buffer solution, two with pH = 7.0 and two with pH = 4.0. Connecting to an MCU Since the module has an analog voltage output within the 0-5V range and is designed to operate from a DC supply voltage of 5V, it is quite easy to connect it to an MCU such as an Arduino Uno or Nano. You just need to connect its + and - power pins to the +5V and GND pins on the MCU board, and its PO output pin to one of the MCU’s analog input pins, such as A0, as shown in Fig.6. Fig.6 also shows the Arduino connected to a 16×2 character alphanumeric LCD with an I2C serial interface, so it can display the pH reading. More about this shortly. Now we just need firmware to sense the module’s output voltage and convert it into the equivalent pH value. After a bit of internet searching, I found the website www.circuitdigest.com that has an article by Debasis Parida describing a pH Meter using the module we are focusing on here, together with an Arduino Uno and a 16×2 LCD display. Australia's electronics magazine A close-up of the tip of the probe that came with the pH meter module. You should just be able to see the two electrodes, The main electrode is a very thin winding wire in the centre. September 2023  39 Fig.6: a wiring diagram showing how to connect the pH meter module to an Arduino Uno or similar. We have also incorporated a 16x2 LCD module with I2C serial interface so that it can display the pH reading. He also provided an Arduino sketch, although there were a few drawbacks: he had a parallel interfaced LCD, rather than one with an I2C serial interface, and his code for converting the module’s analog voltage readings into equivalent pH values was a bit convoluted and difficult to follow. So I decided to write a sketch of my own. It is named “Arduino_pH_ meter_sketch.ino” and is available to download from the Silicon Chip website. When you upload the sketch to the Arduino and it begins running, it gives you this opening display: Silicon Chip Liquid pH Meter After pausing for two seconds, it starts measuring the output voltage from the pH amplifier module, converts it into the equivalent pH value and then displays both the pH value and the amplifier module’s output voltage, like this: pH = 7.0 Vaverage = 2.50V It continues doing this every two seconds. If you’re wondering why the second line displays “Vaverage”, that is because the sketch calculates the average of 10 measurements to compensate for minor fluctuations in probe output. The sketch also sends the pH value and the average module output voltage back to your PC or laptop via the Arduino’s serial port if you have it connected. So if you start up the Arduino IDE’s Serial Monitor, you’ll see the 40 Silicon Chip same information appearing every two seconds. Once you have the pH module and probe connected to an Arduino as in Fig.6 and have uploaded the sketch to the Arduino and seen that it works, there is still one further step before your pH Meter is ‘ready to go’. This the important step of calibration. Probe and module calibration This step is particularly important because every pH probe is slightly different in terms of its pH to voltage conversion characteristic. Before you can start using the probe seriously, you have to test its response with liquids at a minimum of two known pH levels. This calibration needs to be done not only before you start using the pH Meter, but every time you change probes or clean/refurbish your probe. Calibration is a two-step operation. First, you place the probe into a ‘neutral’ liquid like distilled water, with a known pH of 7.0. Then you can adjust trimpot VR1 on the module (the one nearer CON1, the BNC input connector) until the LCD readout gets as close as possible to show pH = 7.0 and Vaverage = 2.50V. The second calibration step is to place the probe into a different liquid, with a known pH that is well away from 7.0; say, 4.0 or 10.0. This will allow you to work out the effective slope of the probe’s transfer characteristic. If you get a pH reading that differs significantly from the correct figure, you can make a change in the Meter’s sketch to correct for this error. Australia's electronics magazine And while you can use distilled water for the pH 7.0 reference buffer, it is not so easy to find another liquid with a known pH of 4.0 or 10.0. You really need to get a reference solution from a reputable supplier. While you can find many suppliers of reference buffer solutions on the internet, many are overseas and can only supply them in large containers that cost a lot to ship. Luckily, I found a local Australian supplier offering two 125mL bottles, one of pH7 buffer and the other of pH4 buffer, for the modest cost of $15.50 plus $8.95 for shipping. This supplier is My Slice of Life Pty Ltd, located at Shop 2, 159 Vincent Road, Wangaratta Victoria 3677. Phone: (03) 5798 3489 Web: https://mysliceoflife.com.au I ordered one of these packs of buffer solution, and they can be seen in the photo. When they arrived, I was therefore able to have a go at calibrating the pH module and one of its probes. Running into difficulties Unfortunately, I soon struck a puzzling problem: when the hardware was hooked up as in Fig.6 and either of the probes connected to CON1 of the module with its tip end submerged in the pH = 7 buffer solution, no adjustment of trimpot VR1 would allow the pH value to be displayed at anywhere even close to 7.0. The maximum pH displayed remained no higher than 2.60, with Vaverage no lower than 3.28V – much higher than the correct figure of 2.50V. At first, I suspected that trimpot siliconchip.com.au VR1 was faulty, but when I replaced it, there was no change. Then I wondered if there might be a dry joint on the module’s PCB, in the vicinity of IC1. But resoldering any joints that looked dubious still didn’t cure the problem. So it wasn’t possible to calibrate the pH module with either of the two probes supplied with it. I suspected that either the probes themselves had ‘dried out’, or that IC1a has been damaged due to static charge on one of the probe cables. One further thing I should mention: I could not find any way to gain access to the ‘refill’ opening near the top of the refillable probe. The cover ring seemed to be stuck in position, so there was no way to top up its inner electrolyte. In the hope of providing some answers to these problems, we ordered another module and an accompanying non-refillable probe. When these arrived we tried seeing if the new module and/or the new probe would give more sensible results. Cutting a long story short, the replacement module and probe didn’t perform any better than the first ones. With the probe in a pH = 7.0 solution, trimpot VR1 still would not allow the value of Vaverage to be taken below 2.93V, giving a pH reading of 4.58. Way off! I tried various things to see if I could track down the cause of this problem, including re-checking my sketch to see if I had made any programming errors, measuring the actual gain of op amp IC1a (it turned out to be 2.997 – very close to 3.0) and trying to run the module from 3.3V instead of 5V. But none of these provided any clues as to the real cause of the problem. Then I decided to see if I could make trimpot VR1 able to bring the Vaverage level down to 2.50V when the probe was in a pH = 7 buffer solution. After a bit of experimenting, I found this could be done by connecting a 5.6kW resistor in parallel with the 5.1kW resistor connecting pin 5 of IC1b to ground, bringing its effective value down to 2.67kW. This allowed the module and probe to give correct readings of pH = 7 when the probe was in either distilled water or the pH7.00 buffer solution. Astute readers may have spotted the design flaw in the circuit earlier – the reference attenuator, including siliconchip.com.au trimpt VR1, does not have enough range to reduce the 2.5V reference to the 833mV needed for calibration. By shunting the 5.1kW resistor, we are fixing that flaw and providing enough range for correct calibration. Why it was designed this way is a mystery. However, when I tried swapping the probe over to the pH4.00 buffer solution, there was still a problem: the module was now giving a pH reading of around 5.14, rather than the correct 4.00. So I had to change the value of the variable “Senslope” in my Arduino sketch, from the ‘ideal probe’ figure of 0.05916 volts per pH unit to 0.0234. So finally, after fiddling with both the hardware and software, I was able to get the probe and module combination calibrated – at least, at the two pH levels of 7.0 and 4.0. Mind you, there was still no real explanation as to why these hardware and software changes were necessary. Nor was there any way to be sure that the output characteristic of the module was still linear, so the twopoint calibration would ensure correct pH measurements at levels well away from pH = 4.0 or pH = 7.0. After further testing and analysis, I determined that the high impedance of the probe and the module’s input circuitry means they pick up a fair bit of noise and 50Hz hum, causing the readings to vary up and down. This means that the module needs to be housed in an Earthed metal case, to provide shielding. That would at least give you a chance of being able to calibrate them out-of-the-box. Summarising I can’t give these particular modules and their probes a glowing report, given that I wasn’t able to achieve calibration using the normal procedure, and it’s unclear whether the readings could be relied upon over the full pH range! The circuit design may seem to make sense at a theoretical level, and the probes and modules seem to be made correctly. The problem is that they don’t provide sufficient instructions on how to assemble the device to avoid RF and mains hum pick-up from interfering with the results. We think the DFRobot Gravity pH Meter V2.0 is more likely to work without modification, given its higher price and availability from more repSC utable sources. The AliExpress module also includes two separate pH probes (one ‘refillable’ and the other not), a small bottle of electrolyte and some mounting hardware. Australia's electronics magazine September 2023  41