Homemade PH probe

[ktek]

Hello everyone
I need some tips on how to DIY an ultra cheap ph sensor
My idea is to make a network of sensors to monitor the pH in a garden or a field
Each sensor will be connected to a low-cost microcontroller
I saw this:

http://www.lowes.com/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&storeId=10151&catalogId=10051&cId=SEARCH&productId=3142183&cm_mmc=SCE_nextag-_-nextag-_-11102010-_-Matthews%20Four%20Seasons%20PH%20Moisture%20Light%20Meter%20MFS42

Such an electrode is more robust than the traditional glass

I do not know what material the electrodes are made, they sound like aluminum tubes and copper-bottomed lead??
 I also saw that in the past made them pH with antimony, but is poisonous and is not found anywhere
 There someone who knows me say what metal to use? .

[pickit2]

http://www.explainthatstuff.com/how-ph-meters-work.html

[DarkClover]

Hy!
Long time ago I made a ph probe by myself but I only found this link in my folder.
http://www.sensorland.com/HowPage037.html

However if I could find the other stuff I'll upload it somewhere.

[PeterMcMonty]

...late but maybe pertinent. How to make an homemade pH electrode:

https://noisebridge.net/wiki/BioBoard/Documentation/pH

almost the same stuff, but with no pictures:

http://www.instructables.com/id/cheap-DIY-electronic-pH-meter/?ALLSTEPS

[motox]

From what I was told (by people that tried to make homemade ph sensors) the calibration is very difficult. In the end,  it is cheaper to buy the sensor than spending soo much time and effort trying do it.

[PeterMcMonty]

I agree motox!
I've not tried yet, but I don't expect outstanding results.

But may be interesting to see what happens

[dogipic]

Hi.

I've made an simple interface using a low noise opamp when I was still in college. Basically it's an amplifier for the probe connected to an ADC. It has also a possibility to calibrate the probe. I've also built and tested it.
Right now I don't have access to my computer. As soon as I come home wil look for it and post it.

Posted on: April 16, 2015, 08:48:02 08:48 - Automerged
Hi.

Here is the schematics.
BNC1 is the input, JP2 is the output. Trimmer R3 is for setting the amplitude, R10 and R8 for zero calibration. R1 can be a 100k resistor or you can replace it with a termistor. This way the circuit is temperature compensated.

[Vineyards]

It should work. However, you should take into account below points:

1-pH probe outputs a voltage of roughly 56mV per pH unit change and this goes between plus/minus 414mV, zero standing for pH7.
2-Probe output current is in the femto amper level therefore Ib and input Z of the opamp is of great importance here. I would recommend a higher Z amp with a lower Ib.
3-High Z causes two main problems: high noise susceptibility and damage risk caused by noise signals exceeding safe limits.
4-Ground loops are a big concern. Ground isolation is a must if this pH meter is meant for an industrial environment.
5-You could use a more precise voltage divider like TLE2425.
6-Cables, sockets etc might introduce additional noise by way of static discharge and leakages. Clean all inputs on the circuit with ethy alcohol to reduce it.
7-If you use more expensive high Z opamps and solder all the leads on the board you nullify all your gains by doing so. Solder the coax core on the input pin itself through a resistor.
8-You could use a two pole active filter in the output stage.
 

[PeterMcMonty]

I agreed with all your suggestions, Vineyards.
pH probes are very high impedance and very low signal devices and requires a careful design of input interface.
Moreover, pH readout should be temperature compensated.
So, it's not a simple task, but from my point of view, I think the most challenging task is to build the probe.
That's why I would like to try and see what happens, before giving it up and buy a commercial probe on e-bay

Meanwhile, I found on internet some references to a "pH masurement and control circuit" for aquariums that is said "the best all over the internet" (not my worlds ).

Unfortunately, the links to geocities are all dead, thanks to Yahoo , with a lot of interesting personal pages all gone.
But there is a chance, thanks to webarchive.org that saved millions of pages and lets everybody to view them:

https://web.archive.org/web/20010607135242/http://www.geocities.com/Petsburgh/Zoo/1705/pH.html

I've downloaded the whole page and saved in pdf format, as well as the electric schematics (as gif image) and bill of materials (in txt format).

So, I include all this stuff,  just to compare.

What I've noticed is that TL082 and TL072 (almost the same) are quite extensively used in all schematics: may be there are better op-amps, but for sure these are the cheapest with a very high Zin.

[Vineyards]

Because most of those designs are from 20-30 years ago.

Posted on: April 17, 2015, 11:44:08 23:44 - Automerged
If you can make a pH electrode, you can make a valve too. It requires a lot of manual skill which I seem to lack.

[PeterMcMonty]

I'm not able to get vacuum inside... at least at home so I'm not able to make a valve.

But I think I can make the active and reference electrodes of a pH sensors, following the instructions in the links I've posted in this comment:

http://www.sonsivri.to/forum/index.php?topic=35118.msg171885#msg171885

I will try, just for fun. The only thing I will change will be that I will not use a Christmas ball, as suggested, but a vial from a medicine: smaller and available even after Christmas and Easter

As I said, I don't expect great results, so I will end to buy some e-bay or others. But I 'm curious and I like to experiment.

My mother, my father, my grandmother... all of them had a quite good manual skill. I always loved to make things by myself... may be it's genetic? But everybody has his/her skills and lacks skill in other fields, so I'm not particulary proud of mine neither I'm not shamed if I'm not partculary skilled in other fields (example: I'm italian, national sport is soccer but I always been a disaster to play... never mind: I did other sports ).

By the way, I agreed that most of designs are dated, so for a much better performance I would use LMC6482 instead of TL082:
- Ib Input Current: 20 fA (typ) / 4 pA (max) for LM6482 vs. 30 pA (typ) /100 pA for TL082
- Rin Input resistance:  >10 Tera ohms for LM6482 vs. 1 Tera ohms for TL082
- $ cost: 0.54 to 0.80 USD for LMC6482 vs. 0.19 to 0.23 USd for TL082 (1k quantities)

Not a novelty, but an improvement. And the cost is not a big issue, for an homemade unit.
Maybe there are even better op-amps, but I think it's a good starting point.

Effectively, the main problem is the pcb resistance, that could vanish any effort, but a "dead bug" layout, at least for the first stage, may help in solving the problem.

Posted on: April 18, 2015, 01:39:10 - Automerged
Ok, There are better ones:
LMP7221: 3 fA (typ) / 20 fA (max) Input current, single, 4.70 USD|1kpcs non standard single pinout
LMC6001: 10 fA (typ) / < 25 fA (tested) Input current, single, 5.76 USD|1kpcs
and, descending to Earth:
LPV542: 100 fA (typ) / 1 pA (max) Input current, dual, 0.70 USD|1kpcs

But the old LMC6482 is still valid.

Posted on: April 18, 2015, 02:39:39 - Automerged
About LMP7221, I found a couple of TI appnotes that fits like a cherry over the cake

AN-1852 Designing With pH Electrodes (Rev. A)
http://www.ti.com/lit/an/snoa529a/snoa529a.pdf

and

AN-1798 Designing with Electro-Chemical Sensors (Rev. C)
http://www.ti.com/lit/an/snoa514c/snoa514c.pdf

[Vineyards]

What matters is curiousity and the positive energy that it generates.
Now that you have the manual skill, the next obvious step must be to either buy or make the tools that you need along the way.
I would suggest you should consider laying your hands on a more serious glass bulb membrane because much of the action happens there. For example, the high Z character of the probe can be attributed to the tiny pathway between the measured solution and the probe solution which restricts all but a few of them by comparison. The bulb has a porous surface which permeates ions. I believe this is achieved by treating the membrane surface with phosphoric acid which has the strange capability of dissolving glass. ALthough I don't know how it is done the cycle must go like this: Start with as small a bulb as possible, the glass must be as thin as possible too. Dip this into phosphoric acid (handle very carefully, never put in glass containers and preferably work outdoors. Never inhale the fumes.) and further erode it until it becomes porous. You may polish this with very fine grit or sand paper (respectively nunbers 400-800-1000). Commercial electrodes usually feature a porous ceramic membrane. and it is also another matter. Once you have produced the individual elements the next step would be to properly install them. Since a practical unit will have a built in reference junction, the next challenge is to install that in that tiny housing in proper shape.

I suggested you consider valve production because it also involves glasswork and you will probably use similar tools so one investment can be extended to another (put them on e-bay later on).

I tried many of the opamps on the list. There are specialized opamps in new generation where input pins and ground are deliberately designed wide apart from each other to avoid leakages. I tried LMC6001 and CA3140 both of which Works quite well. LMC6001 is now obsolete but still hard to beat.

[PeterMcMonty]

Neither LMC6001 nor CA3140 are obsolete:

http://www.ti.com/product/lmc6001/samplebuy

https://www.intersil.com/en/products/amplifiers-and-buffers/all-amplifiers/amplifiers/CA3140.html#order

If you register, you can even get free sampes of both.

I don't want to tamper with phosphoric acid. In fact the guess I did is that the original choice of Christmas bulbs is because their glass is very thin, so (maybe) somewhat porous.

About your advice, you confirm my fear that my idea of substitute Christmas bulb with a medicine vial probably won't work: vials glass is not thin enough.
And even the original homebrew electrode has probably a different sensitivity than the 56 mV/pH unit of commercial ones, but it has to be calibrated anyway.

I think the two key-ideas that make possible to homemade the probe, at least with didactic purposes, are:
1) select Christmas glass bulb as a suitable porous-enough glass object to use for active electrode,
2) make the reference electrode as a separate one, without any attempt to embed the second into the first.

This saves a glasswork that just Murano glassmakers could be able to do.

By the way, Murano is the island close to Venice  famous for the glass artworks and their art glassmakers.
A small Murano glass factory was committed by CERN to produce some 10 thousand lead doped glasses of extreme purity for photomultipliers in a detector of Delphi experiment (an old LEP experiment, in the very same tunnel of LHC) back in 1984... I know it because I was working there in the same Delphi experiment (but another detector) at that time so, they are for sure quite able to make great things with glass.

So, I think it's better I go to commercial probes on e-bay and concentrate on designing a good front-end amplifier for that probe, with one of the op-amps discussed above and with a careful layout that includes an input guard ring.

I have a portable pH/EC/TDS + Temp °C tester by Hanna Instruments (mod. HI98129) that works quite well, but I'm thinking to make a uC based unit to keep my aquarium tanks constantly monitored and controlled in temperature and, perhaps, in CO2 dosing.

Actually I have an homemade uC controlled fan based temperature controller, very efficent to keep low water temperature during hot summers that I described here (you know very well, since you commented it):

http://www.sonsivri.to/forum/index.php?topic=58162.msg171597#msg171597

but I would like to improve my project by including heater control (and that is easy), CO2 erogator control (and that's why I need to measure pH) and perhaps an EC monitor.

I don't know if I'll do it: I'm mumbling a lot and sometimes I don't even start, but... who knows?

[PeterMcMonty]

By the way... someone found the way to homemade tubes

https://vimeo.com/14490689

[dogipic]

"So, I think it's better I go to commercial probes on e-bay and concentrate on designing a good front-end amplifier for that probe, with one of the op-amps discussed above and with a careful layout that includes an input guard ring."

I absolutely agree with that. You get pH probes verry cheap.

The circuit which I posted was also used in an aquarium. I built a complete automation system and this was a part of it. I also bought a cheap pH probe in an electronics store called Conrad, here in Europe. The accuracy was in the range of typical 0.1pH with a verry good repeatability. Also you need good quality calibration fluids. You also have to consider the life time of such probes. The life time of pH probes varry a lot. The one which I bought had a life time of only 6 months. Also when u consider to use such a probe in an aquarium wich has constant temperature then the temperature compensation isn't required (because it is a constant).

[Vineyards]

Furthermore, if you take a look at the temperature compensation curve you will see that at pH7 the offset caused by t is zero and remains minimal a few pH levels up and down. Unless, there is a big shift in temperature, the error rate will always be higher than the effect of temperature at these moderate pH levels.

I have also built commercial grade pH, ORP, conductivity and dO meters. I believe most of the error component lies on the sensor side. Chemical contamination is the number one culprit. Electronicswise, ground loops and noise are the two most common headaches.

Posted on: April 23, 2015, 04:56:43 16:56 - Automerged

By the way... someone found the way to homemade tubes

https://vimeo.com/14490689

Yes, I watched these quite a while ago. What we need is someone with skillful hands and a desire to create... You can safely rule me out

[Balderdash]

I have recently gained an interest in a pH Sensor for monitoring plants, not aquariums, and the cost/fragility of traditional pH Sensors isn't suitable for my application (multiple units in a garden, wirelessly reporting several pieces of data several times per day).

I've seen some Patent papers about solid-state pH sensors, but haven't found anything for sale, yet, and the multi-sensor linked in the first article in this thread having pH as one of its 3 measurements and retailing for less than $8 leads me to believe I can fabricate my own pH Sensor Electrode.  Towards this end, I don't see the need for 0.02pH accuracy/repeatability; a repeatable value in the range of +0.1 pH will probably be more than adequate (most vegetables have an acceptable range of 1-2 pH units for proper growth).  Consequently, I am looking at the LMP7712 instead of the considerably more expensive opamps referenced in other posts (100 fA instead of 10, but the spec sheet still lists it as suitable for high impedance sensor buffering).

I'm curious as to the need for a thin glass insulator in my application (moist, but not wet, soil instead of a fluid).  Why not just two different metal wires, one coated with Silver-Chloride (the typical sensor material) and the other being plain copper (or maybe HgCl, although I'd prefer to avoid dealing with Mercury)?
  Marking this out as I just read the link in pickit2's post.


Any ideas on what I should consider before I commit the Alpha prototype to schematic/PCB?


Thanks in advance,

Jeff

[bobcat1]

@Balderdash

long time ago i have encountered an article who show how to use LED as a PH meter a work done by student
try search google for this article - it might help you with your project

Bobi

[crunx]

Interesting!

I have earlier seen articles about various semiconductor-based pH sensors, but they seemed to be quite difficult to fabricate without proper laboratories.

I looked around a bit more and found this:
http://www.mdpi.com/1424-8220/6/8/848

Could that one be the one you meant, bobcat1 ?

[Balderdash]

I saw this in my research.  While fascinating, it won't mesh with my application because I want to measure the pH of soil, not a fluid.  Even if commercially available, I doubt this particular implementation would meet my needs.


Balderdash