> Summary: How to avoid grounding problems on a breadboard while
> prototyping microcontroller circuits which use ADC?
>
> Hello,
>
> I'm developing a pH meter using a PIC 16F870 (10bit ADC). The pH probe
> is a coaxial cable with the outer wire (shield?) which is to be
> grounded. I'm using a LF356 (+5V 0 -5V) and a LM324 (+5V 0V) to amplify
> the pH probe signal which is then fed to the PIC (PORTA,0; AN0).
>
> The PIC is on a PCB development board with crystal, LCD et al on it. The
> LF356 and LM324 are on a breadboard, with a short wire from the output
> of LM324 to the PIC.
>
> As mentioned in a lot of subject matter that I read through to avoid
> grounding problems, the grounds of the LF356, LM324 and pH probe shield
> are connected together; while the PIC ground is separate from this.
> These two separate grounds (wires actually) go the another PCB board on
> which I've soldered a full-wave rectifier using 7805 and 7905 and a
> center-tapped Xmer.
>
> While running the PIC, the ADC reading varies almost cyclically about
> 150 ADC units (for eg. pH readings are 7.10 .. 7.30 .. 7.65 .. 7.93 ..
> 8.34 .. 8.67 .. 8.50 .. 8.14 .. 7.87 .. 7.56 .. 7.34 .. 7.05 and
> repeat). No software filteration is used, although a 470pf capacitor is
> soldered between AN0 and the PIC ground. The 150 ADC range is too much
> to bear.
>
> If I touch the outer shield of the pH probe, the reading is pretty
> stable.
>
> So to avoid grounding problems while prototyping such a system, what
> should I do about the grounds (and power???) ?
>
> Thanks in advance,
> Mohit.
>
> --
> http://www.piclist.com hint: The list server can filter out subtopics
> (like ads or off topics) for you. See http://www.piclist.com/#topics

> > measure the voltage potential between the
> > two grounds (while powered).
> Same, 0V on multimeter.
>
> > Possibly defective software...
> > even a problem in the display routine.
> Using 16x1 LCD, the first 8 characters show pH and the next 8
> temperature (of the form "14.00pH, 100.0^C"). While running the LCD
> shows temperature properly - using a LM335, connected to PIC PORTA,1;
> AN1. Since the ADC routine is shared by both channels (pH and
> temperature), I believe it is alright and so is the display routine.
>
> > Then disconnect the pH circuitry altogether.
> > Replace it with a simple trimpot.
> > Does the problem go away?
> Yes, the problem goes away. The PIC shows a very stable reading.
>
> > disconnect the controller from the pH circuitry.
> > Measure it's output.
> Using a multimeter, the output of the pH circuitry is stable.
>
> The temperature measurement and display is alright.The problem happens
> only when the pH circuit is connected to PIC. I've even tried connecting
> the pH probe outer to a large metal case but the fluctuations still
> persist.
>
> The point is that I get a very stable pH reading when I touch my finger
> to the pH probe outer wire. Can anybody suggest why this might be the
> case?
>

> Regards,
> Mohit.
>
>
>
>
> pic microcontroller discussion list wrote:
> > ----- Original Message -----
> > From: "Mohit Mahajan" <TakeThisOuTbiozenEraseMEspam_OUTSANCHARNET.IN>
> > To: <RemoveMEPICLISTTakeThisOuTMITVMA.MIT.EDU>
> > Sent: Wednesday, July 21, 2004 10:47 AM
> > Subject: [EE:] Grounding problems on breadboard
> >
> >
> >> Summary: How to avoid grounding problems on a breadboard while
> >> prototyping microcontroller circuits which use ADC?
> >>
> >
> > Your measurements are taken with respect to ground. I'm guessing that
> > the PIC ADC is being used, which means they are being interpreted wrt
> > ground on the PIC/display board... Note that ground leads have
> > resistance, so variations in power requirements (LCDs, PIC, etc) will
> > cause the grounds to fluctuate. I'm guessing this is your problem.
>
> that may or may not be the problem. Grounding is certainly one of the
> first things to suspect. The OP said he was having grounding problems,
> but with no details of what he ckecked. 150 bits tells me there is a
> fundamental layout problem somewhere.
>
> > If you really need to separate the grounds, I'd suggest using some
> > kind of VCO/VFO pair, coupled capacitively.
>
> IMO, that's premature at this point. Possibly defective software. Or
> maybe even a problem in the display routine.
>
> First verify what & where the problem is using basic techniques.
>
> For example, measure the resistance between the two grounds (unpowered
> of course).
>
> measure the voltage potential between the two grounds (while powered).
>
> Then disconnect the pH circuitry altogether. Replace it with a simple
> trimpot. Does the problem go away?
>
> disconnect the controller from the pH circuitry. Measure it's output. Is
> it fluctuating the same way the ADC readings are? In other words,
> perhaps your amplifier is really an oscillator...
>
> Thats all the time I have for now, but if you do some simple &
> fundamental diagnostics, i'll bet you find the problem pretty quickly.
>
> --
> http://www.piclist.com hint: The list server can filter out subtopics
> (like ads or off topics) for you. See http://www.piclist.com/#topics
>
> --
> http://www.piclist.com hint: The PICList is archived three different
> ways.  See http://www.piclist.com/#archives for details.

> -----  From: "Mohit Mahajan" <biozen@S
>
> > > measure the voltage potential between the
> > > two grounds (while powered).
> > Same, 0V on multimeter.
> >
> > > Possibly defective software...
> > > even a problem in the display routine.
> > Using 16x1 LCD, the first 8 characters show pH and the next 8
> > temperature (of the form "14.00pH, 100.0^C"). While running the LCD
> > shows temperature properly - using a LM335, connected to PIC PORTA,1;
> > AN1. Since the ADC routine is shared by both channels (pH and
> > temperature), I believe it is alright and so is the display routine.
> >
> > > Then disconnect the pH circuitry altogether.
> > > Replace it with a simple trimpot.
> > > Does the problem go away?
> > Yes, the problem goes away. The PIC shows a very stable reading.
> >
> > > disconnect the controller from the pH circuitry.
> > > Measure it's output.
> > Using a multimeter, the output of the pH circuitry is stable.
> >
> > The temperature measurement and display is alright.The problem happens
> > only when the pH circuit is connected to PIC. I've even tried connecting
> > the pH probe outer to a large metal case but the fluctuations still
> > persist.
> >
> > The point is that I get a very stable pH reading when I touch my finger
> > to the pH probe outer wire. Can anybody suggest why this might be the
> > case?
> >
>
> What is the impedance of the probe? I understand that ph probes are very
> high. Since you are reading it with a bipolar opamp, that may be an issue.
>
> http://pdfserv.maxim-ic.com/en/an/A0512.pdf
>
> > Regards,
> > Mohit.
> >
> >
> >
> >
> > pic microcontroller discussion list wrote:

> The temperature measurement and display is alright.The problem happens
> only when the pH circuit is connected to PIC. I've even tried connecting
> the pH probe outer to a large metal case but the fluctuations still
> persist.

> The point is that I get a very stable pH reading when I touch my finger
> to the pH probe outer wire. Can anybody suggest why this might be the
> case?

> Regards,
> Mohit.