9172 chassis or 9203 module coupled w. pH meters gives strange output

[ShaunDustin]

I am using a bank of 17 Omega DP24-pH meters in conjunction with several cDAQ 9172 chassis and 9203 current modules.  The problem we are running into: If I put two or more pH probes into the same solution then hook the current output from the meters to the same 9203 block, the readings go nutty.  One will drop to a decimal value around zero, and the other will go up.  If I short across the current output terminals to eliminate the influence of the NI equipment, readings return to zero.  The meters are fine if I don't connect them to the 9203--I have put up to 6 in the same solution and they read just fine, so I don't believe it is probe interference, it has to be output signal interference, right?  The output signals are wired into the module in blocks 0-7, with the common grouped together and wired into 9. 

Any ideas on how to resolve this would be appreciated.  I must be missing somthing in the wiring?

[ShaunDustin]

Whoops--I said "readings return to zero"--should have been "readings return to expected values"

[Intaris]

You have a current loop.

Ideally a pH meter would have a differential input where both the pH and Reference inputs are isolated from ground.  This is not the case with cheaper models.  They often have a mere 10k resistance to ground or somethimes even less.

Introducing a current into the measuring loop will totally destroy the ability of the pH meter to make a useful reading.

Sometimes it's not even the reference electrode but the communications port (RS-232) which shares a ground with the reference electrode channel.  I've had applications where everything worked fine until the RS-232 cable was connected!  Luckily enough, there are a few RS-232 isolators on the market.....

Try finding an opto-coupler to place between the electrode and pH meter to counter-act this effect.  Actually, you'll need 17 opto-couplers.  I remember that German Metrohm had such a device on the market which worked well (at least 4 years ago when I still had contact with them).  I just did a Google search and came up empty-handed.  I know they're pretty rare and they do cost a few $$.

This is the problem with pH meters of less-than-optimal quality.

Also, grouping the reference channel of 6 different electrodes together is a bad idea.  Each reference electrode will have a different potential (can be up to 0.1V) and given the low resistance between them, this will eat up your electrodes.  It's just a matter of time.  Why not just connect up ONE of the reference electrodes.  What type of sample do you have, how far apart are the electrodes?

In case you're wondering, I developed pH electrodes for a while, so I'm pretty much up to speed on the pitfalls of different pH measurements.....

Shane.

[ShaunDustin]

Thx for the response.

The opto-coupler looks like it would involve building a set of BNC-optocoupler-BNC cables by soldering bnc connections via jumper wires to the optocoupler; is that what you are saying?

Since this is a research project, not a long-term monitoring application, I might be able to get away with this.  I don't understand how the optocouplers will capture the nuances of an analog signal, though.  Is there a specific type I should look for to accomplish this?

I also can't seem to get it through my thick head how all the meters can run along just fine in close proximity to each other, then the hookup to the NI hardware screws them all up. Can I isolate the grounds some other way to eliminate the feedback in the current loop?  The electronic equivalent of what us Civil engineers would call a check valve?

The other alternative, tying all the measurements to a single reference electrode might be tough--I am trying to monitor pH at 2 different locations and ORP at three different locations in a set of 3 different 60 L reactors, total of 15 channels, plus 2 more in a seperate location.  The probes range from 6 to 36 inches apart. I don't anticipate significant differences in pH, but  I do expect ORP to change considerably over that range.  Wouldn't the ORP reference vary as well?

[Intaris]

Well clearly if you have electrodes in different reactor, then they can't share the same reference....

The ORP measurement is just as dependent on the reference electrode as the pH measurement.  Principle is the same for the reference electrode.

I know something like this HAS been done in the past, but I am no longer in touch with the people who implemented it.  As to which parts.... I'm not really strong in electronics or circuits.  Sorry.

I think if they all work in combination, then the easiest answer would be to isolate the output from the meter instead of the pH electrode input.  This will be a much easier task as the input impedence on the electrode side is about 10^12 Ohms (or it should be).  On the analog out it'll be about a lot less, making circuit design a lot easier.

I'm not familiar with either the meter nor the NI hardware you're using, so I can't give any really specific tips.  It sounds to me like the current output of the meter is not sufficiently (or at all) isolated from the electrode measuring circuit.  If you can find an optocoupled voltage follower, try it out on one of the meter outputs.  I know an optically isolated voltage follower works for voltages, but I'm not sure how to implement one for current.....

Shane.

[AnalogKid2DigitalMan]

Been there, done that for various meters/controllers with 4-20mA outputs. Went with an off-the-shelf solution, not cheap around $400 for a 3 channel loop-powered isolator unit.

http://www.dataforth.com/catalog/doc_generator.asp?doc_id=3&attr_index_id=1277