# Building a pH meter

Discussion in 'General Electronics Discussion' started by thunderdood, May 3, 2011.

1. ### thunderdood

10
0
May 3, 2011
I'm trying to build a pH meter and have done some research. I am using a 741 op-amp and wired the circuit as a voltage follower. I had saltwater in a plastic cup and I was getting a reading at around 350mV. I did some calculations from what I researched, and the equation is Vout/54.55mV= pH units. Then you subtract the neutral pH of 7.01 by the pH readings so in my case it would be 350mV/54.55mV=6.416 pH units. Then 7.01-6.416= .594 pH. I know something isn't right since the pH of saltwater would be from around 8-8.4. Any advice or help would be appreciated. Here is a schematic of my circuit. Keep in mind I have the op-amp powered by +9V and -9V which is not shown in the schematic.

2. ### Laplace

1,252
184
Apr 4, 2010
In the article "Building the Simplest Possible pH Meter" http://www.66pacific.com/ph/simplest_ph.aspx it stresses the importance of using a high impedance input amplifier such as the TL082. Could your problem stem from your choice of the 741 op-amp?

3. ### thunderdood

10
0
May 3, 2011
Yeah, after what I've read, I think I'll have to get one of those op-amps that are 20MOhm and above since the 741 is 6MOhm I believe.

4. ### duke37

5,364
769
Jan 9, 2011
You do not have a voltage folower, you have connected the probe to the - pin. To get a follower, connect the probe to the + pin and add a resistor from the output to the - pin and another from there to ground. A better op-amp will not help in this configuration which effectively shorts the probe to ground.

5. ### Resqueline

2,848
2
Jul 31, 2009
What duke said: a voltage follower is connected like this:

File size:
12.8 KB
Views:
486
6. ### thunderdood

10
0
May 3, 2011
Yeah my bad. That is how I orginally had the circuit and I was getting random voltage readings. I then grounded the positive input and connected the pH probe to the negative input. From there I was able to get a stable reading. I know the 350mV is not what I should be reading but its a step forward since it was stable.

7. ### duke37

5,364
769
Jan 9, 2011
With a very high input impedance, any external voltage variation can be coupled to the input with capacitance giving variable voltage readings. You will need to use screened input cable and keep away from devices which produce interference such as fluorescent lights.

8. ### thunderdood

10
0
May 3, 2011
By screened input cable, do you mean coax cable? The probe has a coax cable but I am using a BNC connector to be able to change from coax to bare wire for the circuit.

9. ### duke37

5,364
769
Jan 9, 2011
Yes, screened cable will be made coaxially. Coax for radio frequency purposes is often stiff and a large diameter, this you do not want. Your probe cable should be connected to your circuit with a BNC connector and then to the op-amp with screened cable leaving no more than a millimeter or two "free to air". It would be best to screen the amp in a metal screen (do you eat beans?) with power supply and output bypassed with capacitors where they pass through the screen.

10. ### thunderdood

10
0
May 3, 2011
I was able to use the lab this week and test out the op-amp and the op-amp works 4.0. I tested out the op-amp today and I was not getting any readings from it what so ever. I wired the op-amp to be a voltage follower and was getting a -6.9V at the output and input with a +9/-9V supply voltage. I also checked the voltage directly at the pin of the coax connector and I was not reading anything. Could I of burnt out pH probe?

11. ### duke37

5,364
769
Jan 9, 2011
You need to locate the problem.
1. Short the input and check the output is within a few mV of ground.
2. Replace the short with a high value resistance (10M), has output changed?
3. Input a voltage from a single cell, does the output match?
4. Reverse input polarity, still OK?
5. Try probe in a working pH meter.

The fet input op-amp is more delicate than a 741. Are you a bright spark?

12. ### thunderdood

10
0
May 3, 2011
I did short the input and the voltage was a few mV of 0. Using a powersupply for input voltage does work as a voltage follower or inverting op amp as I have tested in the lab. I will try adding a high resistance input.
I'm not a bright spark since I don't know what that is.

13. ### duke37

5,364
769
Jan 9, 2011
I used the english idiom since fet input op-amps can be sensitive to static electricity., I will try to use plain english in future.
You say that it works as a follower or invertor, it should not invert but I presume that you mean that it works with positive or negative input and gives a positive or negative output. If you test with a 10M resistor in series with the voltage input and it works, it shows that the input impedance is very high and the offset current very low.
So far, it looks as if the amp is in order and the problem is with the probe. Have you managed to test it on another instrument?

14. ### thunderdood

10
0
May 3, 2011
I have not tested the probe yet. To be honest, I'm not sure how to test the probe. Would testing the resistance of the probe when in water work?

15. ### duke37

5,364
769
Jan 9, 2011
I know nothing about pH probes but I looked up "glass electrode" on Wikipedia. The probes have a resistance between 50M and 500M so would be very difficult to measure. The probes are sensitive to storage and should never be allowed to dry out. I would think that the only way to check is to test the probe with another pH meter or to use a known working probe.

16. ### nbw

48
1
May 8, 2011
You get something like, if my aging memory serves me correctly- about 50mV per unit on the pH scale with pH probes. And they don't unfortunately last forever - 18 mths in my fish tank and it was starting to lose the plot. Don't forget you need to frequently recalibrate them! Useful to get some guaranteed 4.00 or 10.00 pH test solutions for your project also.

17. ### thunderdood

10
0
May 3, 2011
I opened up a digital pH meter circuit and it has the pH probe going to a BNC connector where the power wire goes to an ADC chip. I was trying to measure the voltage at the wire where its soldered onto the BNC connector, and I wasn't measuring anything. It was something like 6mV, but when I disconnected the multimeter it still read the same value. I noticed that when I tried to measure the voltage on the BNC connector, the pH readings would start to decrease so trying to measure the voltage had an effect on the digital output as it slowly decreased from a pH of 8.
I also tried wiring the pH probe to the op-amp set up as a voltage follower and got nothing. I was reading -6.77 volts at the output. The input measured something around 590mV but kept decreasing little by little.
There has to be a way of reading the output voltage without effecting the pH readings right?

18. ### thunderdood

10
0
May 3, 2011
Tomorrow I will be going to the lab to work on this. Would hooking the pH probe to an oscilliscope work on getting a reading?

19. ### duke37

5,364
769
Jan 9, 2011
You are not measuring pH, you are measuring voltage. If the probe has a source resistance of 100M it is necessary to have an amplifier input resistance much above this to get a reasonable reading. Any standard voltage measuring device such as DMM or scope will have an input impedance of about 1M so will effectively short the input and you will measure nothing. The op-amp is there to sense the voltage and to provide an output with a lower output resistance so that a DMM can measure it. An oscilloscope is used to detect varying voltages, you should not have any such.
Any shunt resistance across the amp input will have a disastrous effect on the reading, in one example I saw (did you give the web site?) the amp was soldered directly to the BNC terminal so there was no printed circuit board or other components which could provide a track other than the package of the amp. If you have handled this with sweaty hands or tried to clean it, there may be sufficient leakage to affect the result.
Check your amp with a voltage source in series with the highest resistance you can find. The voltages should match, both positive and negative. Connect the probe and try in an acidic and alkaline solution, any difference? If so, do they agree with the data given by nbw?
Do to the very high source resistance and the cable capacitance it will take time for the readings to settle.
You say that the input measured 590mV but how could you do this without a very high input impedance device which is what you are tryiing to construct?

20. ### thunderdood

10
0
May 3, 2011
I tested my op-amp using a power supply when I was in the lab last week and the readings were the same from the input to output. The 741 op-amp is rated at 2M ohm resistance. I would need to get a different op-amp, probably one that has at least 20M ohm input resistance, right?