Hello, When I put PH probe inside 7.0 PH solution, values goes from 6.xx to 7.xx and value never stabilize. I used Leophi circuit schematic but moved stuff around, link below. I'm thinking something wrong with circuit or wrong components hopefully someone can help me out.
JayMan:
You have to calibrate the sensor within your code.
Not going to work for OP before the stability is solved.
So you built this board yourself?
Hard to say what could possibly be wrong, but I don't think the TL072 is suitable at all for this purpose - the data sheet doesn't mention it being a precision OpAmp or so, it's so old (data sheet mentions 1978!) it's not even a CMOS part. It has an input bias current of 65 pA which translates to an input impedance of just 15 GΩ, not bad for such an old design but far from good enough for a pH probe.
Try the LMP2232, a precision OpAmp with an input bias of 20 fA instead (that's over three orders of magnitude lower, making this suitable for pH probes, and there are OpAmps out there that do another 2-3 orders of magnitude better), and one that I have seen being used in pH sensor boards before.
I realize you are following project instructions. However you should consider using one of the pre-made Delta-Sigma A/D boards one can get on ebay. You will get stable gain and much much better noise performance.
Input bias current is still far too high for that one. Did your read the data sheet of that one?
The output impedance of a typical pH probe is 10-100 GΩ. Make sure your OpAmp is at least 1, preferably 2 orders of magnitude better, so 10 TΩ for your OpAmp, or <0.1 pA bias current. This TLE2072 has 15 pA input bias. Better than the TL072 but not good enough.
You're also not looking for a drop-in replacement. You're looking for an alternative. Most OpAmps are fully pin compatible so easy to swap out one for another, except the highly specialised ones such as the LMP7721 which has a pinout designed to be used with the BNC connectors of pH probes. See also this AN. This OpAmp features a bias current of just 3 fA, but it's a pretty expensive part.
Do read that AN, and study a bit more on how to properly read a pH probe. It's not trivial, component choice and PCB layout (keep signal leads super short; add shielding on the PCB) are both very important for the success of such a project.
wvmarle:
Input bias current is still far too high for that one. Did your read the data sheet of that one?
The output impedance of a typical pH probe is 10-100 GΩ. Make sure your OpAmp is at least 1, preferably 2 orders of magnitude better, so 10 TΩ for your OpAmp, or <0.1 pA bias current. This TLE2072 has 15 pA input bias. Better than the TL072 but not good enough.
At these impedances guard traces around all signal traces are essential, and PTFE pcb is perhaps needed -FR4 is not fit for 10^13 ohms
So far I've only seen regular FR4 board used for pH sensors, and also never seen the recommendation to use different material in ANs. The signal trace is to be kept short - and by placing the OpAmp next to the BNC connector you can keep this one trace down to about 5 mm in length.
original design is the one in red from sparkyswidgets.com, modified is a bit messy because I will be adding OLED, battery etc.. to it. he uses STMicroelectronics TL072. I couldn't figure out how to remove the solder mask with easyeda, will have to ask in the forums later on.
Hi,
The main problem will be, because you are trying to amplify a mV DC signal, it is very susceptible to noise.
This noise can come from the power supply, components, how the circuit is shielded from outside EM fields and other DIGITAL circuitry on the PCB.
The component layout can cause problems.
I note that you have moved the controller IC closer to the mV amplifier circuitry.
If you remove the probe from the circuit and replace it with a short circuit, what reading do you get.
The branding of the TL072 should not be a problem, as long as it is a TL072.
TomGeorge:
The main problem will be, because you are trying to amplify a mV DC signal, it is very susceptible to noise.
I would point the finger more at the ridiculously high output impedance of a pH probe... it really has no way to deal with any electrical interference for that reason. The voltage level is much less important, and in many cases is not even amplified (the OpAmp is in most cases wired as unity gain).
From that PCB view no idea what you're connecting where and how.
The bias goes to the shield of the pH wire. It is kind of a virtual ground (and one of the reasons why ground loops are a common problem with pH probes).
