Sciencemadness Discussion Board

Presenting: The super ghetto DIY pH meter.

Chordate - 4-3-2011 at 12:45

A while ago I went and built a DIY pH meter from the ground up: op-amp, glass electrodes, the works. Presenting: the super ghetto DIY pH meter.


In the early 1900s it was demonstrated that a glass electrode dipped in an acidic medium gave rise to a voltage potential directly related to the concentration of the acid.

Later it was realized that this was a simple form of ion-selective electrode, the acidic hydrogens exchange metal ions on the surface of the glass electrode forming a micrometers thick film, with the surface concentration being an equilibrium dependent on the acidic concentration.

This voltage potential difference is linear for any given ion species and governed by the nernst equation, which I am shamelessly stealing from wiki-images because they have prettytext.

Glass, however, has a very, very high impedance. In order to detect this voltage potential difference one must amplify gain somehow. For this we use a simple op amp circuit, with reference voltages provided by 9V batteries, though a high quality power supply voltage could work fine.

The second (non glass) reference electrode is a similar electrode, though it is placed in contact with the test solution through a salt bridge to minimize impedance while avoiding contamination of the filling solution. This doesn't need to be glass.

Materials and methods

In order to replicate, you will need:

-HCl (I used regular old pool-grade OTC as I had limited access at the time)

-Sodium Hypochlorite (I used ordinary household)

-1% AgCl in saturated KCL solution (sold as pH electrode fill solution, I got mine from cole-parmer)

-Pyrex tubing, I used 6mm outer diameter .5mm wall thickness tubing. This is the most basic substance which will work and has significant disadvanatages: more on that in a minute.

-1 tl082 op-amp, or other high impedance op-amp (maybe get three, they are cheap and can be used to improve the device)

-Glass frits, vycor. I got these from Princeton Applied Research, and they were far and away the most expensive part of the setup.

-Heat shrink tubing (will come with the frits, most likely)

-Some kind of hollow plastic tube which you can use to build a reference electrode. I used the back of a pen, and regrettably I didn't get any good photos.

-A means of providing a reference voltage: in this case, 2 cheap 9 volt batteries.

-Some means of blowing glass. A bunsen burner with an o2/propane mix with some surgical tubing and a pipet bulb should work for this small scale.

-Some solvents. Denature ethanol and DI water.

The first step is to the formation of the glass electrode body. Commercial electrode suppliers use carefully process controlled lithium doped glass. As was stated above, voltage dependence is linear for a given ion species. If your glass has multiple competing ions your voltage dependence will be linear over a narrower range, and less reproducable. I used pyrex, and as such my electrode was shit. I also made the walls far too thick, and because of this my electrodes took roughly 4-8 hours to equilibrate. Learn from my mistakes. Here is the before and after look for my glass bulb, taken with a cell-phone camera.

Don't let that image fool you, I blew about seven before I got a few that looked usable, and these were still not thin enough.

The reference electrodes are silver chlorde/silver reference electrode. These can be made quite easily by soaking fine silver wire (I aquired mine as jewlers scrap, good silver speaker cable should also work) in sodium hypochlorite solution in the dark overnight, then rinsing thoroughly with solvent of choice. I used ethanol. The wire will tarnish and should provide a stable reference voltage.

The filling solution for the reference electrodes will be a mixture of KCl/AgCl, I bought mine pre-made. Remember that AgCl is light sensitive and will decompose.

The non-glass reference electrode was made by taking the back of a plastic pen (selected because the clip made for easy mounting in a test solution), sealing the end with the tip of little chromatography column (selected because the heat shrink tubing fit easily over the end), and sealing the fritted glass salt bridge over the end. here is a picture of the disassembled reference electrode:

The red material is a bit of wax used to provide a good seal. The idea here is all you need is a watertight container with a salt bridge, filled with KCl and with a silver chloride plated silver wire in it. Once assembled allow the thing to soak in a dilute HCL/KCL solution for 2-3 days to allow water to saturate the frit, so that it is conductive.

The glass electrode is constructed similarly, a silver chloride plated wire sitting in a KCl/AgCl solution filling the glass bulb.

Here is a pic of the filled bulb.

I will return later tonight to finish this here post and show the op amp circuit and the results. I have a place to be.

bfesser - 4-3-2011 at 18:23

Can't wait for the sequel! :D

Chordate - 4-3-2011 at 19:59

Part 2:

So the basic layout of this cell is:

AgCl Reference electrode - Filling solution | doped glass | *test solution* | saturated salt bridge | filling solution - AgCl reference electrode.

When these two electrodes are placed in solution, completing the circuit, a very small voltage is produced. If you check the nernst equation above the Z value used will be one, and ph is already based on log 10 concentration, therefore for every pH unit the cell -should- produce about 0.05912 volts of current. This can be measured with a multimeter, but you have to amplify this signal to overcome the high resistance of the glass.

So you use a cheap op-amp. There are some great references out there for building great circuits which convert the output voltage to 1 volt per pH unit, cancel noise, and are useful across a broad range of pH values.

This is not that guide. Behold my sloppy circuits:

This shows the two batteries used as reference voltages, and the bridge at the bottom which serves as a ground.

And this is the actual OP-Amp wiring.

This layout might be a little confusing so the following is a diagram of the whole cell.

You can extrapolate a wiring diagram from this by checking the pin layouts for the TL082 here.

So, after about two weeks worth, teaching myself to glass blow, and about 80 dollars worth of materials, what were the results?

Educational, but not really worth it. The voltage depenedence was only linear from ph's of about 4-9. The damn thing took 4-8 hours to equilibrate, and the values it gave drifted every time you attempted to use it. BUT! Voltage was pH dependent, and the op-amp circuit worked just fine with a commercial pH electrode. (I am currently using a slightly more complex version of this circuit which I will write up at some point in the near future)

Why? Likely two reasons. The first being the thickness of the glass. I had no real way of measuring the thickness of the glass while blowing. The bulbs should really be very delicate, almost gossamer. If you decide to attempt to improve on this design you should try and blow a number of really delicate bulbs, of ascending thickness, and test each until you find a compromise between durability and sensitivity.

I think more importantly, pyrex is sodium borosillicate glass, with a fair amount of potassium and lithium present. Each of these will contribute to pH readings in a different way making the bulb's sensitivity less linear and less sensitive over a broad range. I couldn't find a supplier for lithium glass at the time I built it, and I haven't tried since.

In the future I think I'd like to try and play with some lithium doped polymer resins to see if they could be used in place of glass. At present, this isn't a very economical way to go about pH measurement. It was fairly educational I suppose.

Any questions?

Magpie - 5-3-2011 at 09:18

This is very interesting and I learned a lot from your posts. Thanks.

m1tanker78 - 7-3-2011 at 17:55

Looks like you did your homework. Reminds me of something out of a Popular Mechanics magazine. Nothing like hands-on work, wouldn't you say?

Have you ever messed with microcontrollers such as the PIC or AVR? It's a little more involved but it's damned ADDICTIVE!


Chordate - 8-3-2011 at 12:29

I hadn't played with any programmable microcontrollers until about a month ago, when my friend showed me some stuff involving custom built flashy-light hats. All my original electronics experience came from being a projection mechanic at a theater and working on small vehicles. Lots of mechanical relays and diodes and capacitors, no microcode.

That said, I am starting to open up to their neatness, but currently all my hobby money is going into trying to make some homebrew graphene composite materials. If that goes well the next step is to build airplanes out of them, at which point he'll be teaching me about DIY avionics and fly by wire systems. It is hard to understate the joys of having a collaborator.

Edit: Oh and he got me one of those "electronics projects for the evil genius" books, which have silly titles but are actually very educational.

[Edited on 8-3-2011 by Chordate]

madcedar - 8-3-2011 at 20:11

Am I the only one who can't see the pictures in this post? All I get are pictures saying "Domain Unregistered. To view, register at:"

I'm assuming the domain is in which case it should already be registered because others can obviously see the pictures.

I'm looking at this thread via https but I can't see the pictures with the nornal http access either.

Chordate - 8-3-2011 at 22:07

Hmm. Very strange. I just tried it on two computers, and used a VPN to see if maybe it was IP related issue. Here is a list of the direct URLs. Let me know if they work.

madcedar - 9-3-2011 at 01:11


Thanks for your post. I'm currently in the Philippines and all the images are blocked. I have a VPS located in the United States and I can see all the images without a problem. So I guess your assumption that it's blocked by IP is correct.


m1tanker78 - 18-3-2011 at 21:50

Chordate, if you ever decide to get into PICs, I highly recommend the 18Fxxxx series (18F1320 is my personal favorite for general use). They're cheap as hell and have MANY features to play around with. The 18F1320 is an 18 pin PDIP package; it fits the breadboards perfectly. Although for a project like this one, all you really need is an 8 pin like the 12Fxxx (from memory).

Judging by the nature of your pH meter project, I think you'd have a blast with the analog to digital converter(s). You can even drive a character LCD and incorporate a couple of buttons for calibration or whatever. Let's see, a buzzer to indicate high/low pH in the case of continuous monitoring....

Got a little ahead of myself there. :D