ph probe op amp circuit problem

[cdr]

I have a standard glass pH probe connected to a Microchip MCP607 op amp which passes the signal to the A/D on a PIC16F877. The problem is that I have some stray voltage produced by high-output flourecent bulbs over an aquarium. When the lights are on and the pH probe is in the water the op amp output signal is 0V. When the lights are off the output signal is correct. When i place a ground probe in the tank with the pH probe and the lights are on I get a signal of 0V.

As soon as I turn the lights off everything goes back to normal. What's wrong with my circuit?

My op-amp circuit is the top circuit on the attachement.

thanks

 

[gerty]

How close is your circuit to the light? Is it ac powered (thru transformer)
from the same outlet? sounds like you need to move or shield one from the other.. Maybe a grounded metal enclosure for your circuit is all you need.. You could also try pwering either the light or the circuit from a different outlet, lamp ballasts do induce a lot of electrical 'noise' on the line....

 

[cdr]

I thought that too. So I left everything the same except I took the pH probe out of the tank and place it in a cup of tank water about 2 feet away from the lights the pH probe and circuit work fine. I only have problems when the the pH probe is in the water under the lights. I also have problems when the lights are off but the ground probe is in the water, I suspect this is noise in the line.

To answer your questions, the circuit (PIC and all) is plugged into the same outlet as the electronic light ballest. The PIC and pH circuit are powered through a 5V DC transformer. I tried moving the PIC to a seperate outlet but the results were the same.

To make this even more of a headache I borrowed a friends store bought pH probe and meter (mine is obviously homemade) and put it in the tank under the same conditions and it works fine, regardless of the lights or ground probe. That's why I though I had something wrong in my op amp circuit.

 

[gerty]

Here's a few things to check: How well is your power supply filtered?
Try adding a small value (.1uf) caps from each of your probe leads to gnd. you can do this at the ic or where the leads attach to the board, this may help attenuate the noise..Are your probe lead wires shielded?

 

[Sebi]

This thing also depend from ballast type. If You have only inductive ballast, probably better. I mean, the noise coming from gas-discharge lamp (F-tube),not via power supply. With inductive ballast radiated 50(60) Hz, but with electronic ballast radiated 30...100kHz.

 

[cdr]

I tried the .1uF caps, they did not seem to help.

It realy does seem to be noise from the lights. Also, if I put a ground probe that is plugged into the ground line of my wall socket I get erroneous ph reading, even with the lights off.

What else can I try to isolate my circuit from the noise?

 

[Roff]

Try the mods shown below. The 1uF caps need to be nonpolarized, low leakage parts. I think it will minimize the interference from the fluorescent light - at least it did in simulation. If you can't find a suitable 1uF cap, scale all the resistors up by a factor of 10, and scale the caps down by a factor of 10.

 

[cdr]

You are the man!!! That did it, it works great now. Thanks for all the help.

Go

 

[mishras1132]

Dear Sir, I am doing a similar project with pic microcontroller. So can you please share your project file with me. Thanks in advance.

 

[JimB]

Mishras
This is a very old thread, cdr (the thread starter) has not been seen here for 12 years.
JimB

 

[mishras1132]

Ok thank you sir for informing.