An isolation issue?

[SCOTTIE]

Hi,

The main instruments in my project are: a 6020E DAQ Pad and a pH meter. They are both hooked up to a laptop by usb interface. The pH meter keeps reading pH into LabVIEW program from a biochemical sample. Depending on conditions required, the LabVIEW program asks the DAQ Pad to output a certain analog voltage. The voltage is then converted to a current via a constant current source circuit (two-OP circuit) and the current is passing through the sample.

PH meter obtains pH valus by detecting potentials of acid concentration in hundred "mV" range. However, the pH meter always gets wrong pH values (out of measurement range) after that DAQ card is connected and turned on. I'm thinking this is happening due to "isolation issue" between two usb slots because it won't happen like this if I hook up DAQ and pH meter separately to two laptops. I wonder can I eliminate this issue by connecting an opto-isolator device, such as an "usb opto-isolator", to the pH meter instead of connecting to the laptop directly? I don't have quite understanding on this isolation issue. Can anyone give me a clue? Thanks in advance.

BTW, this problem also happens if the pH meter is connected by RS-232 interface.


Scottie

[LYu]

Hi Scottie,

A few troubleshooting suggestions to determine where the issue might lie:

- Does the same thing happen regardless of which laptop the pH meter is connected to? If you try connecting the two devices to a computer instead, does the same thing happen?
- Is the serial connection made to the laptop also? If when only the pH meter is connected you're seeing this problem, is it possible that the issue lies in the hardware of the pH meter? Is this a new meter or have you been able to use it successfully via the RS-232 interface before?

Please try the above and let us know your results.

Thanks,
Lesley Y.

[SCOTTIE]

Lesley,

I forgot to mention that the RS-232 is also on the laptop.

1. Same thing happens regardless of which laptop the pH meter is connected to. However, I have not tried this experiment on a desktop. I'll do it soon.

2. Besides, this pH meter works well no matter what interface we used before. The probelm only comes up when two instruments are connected to a laptop at the same time.

This means, if "sample <- pH meter <-(usb1 or rs232) laptop (usb2)-> DAQ card -> sample" forms a loop, the pH meter doesn't work well. If we cut off the connection between laptop and DAQ card or between pH meter and laptop, pH meter runs unaffectedly.


SCOTTIE

[AnalogKid2DigitalMan]

Scottie:

Is the current source electrically isolated from the DAQ card? (it may be worth trying to isolate the two and also make the power the supply that drives the current source totally floating with no reference to earth ground) If not isolated, it may be that the current flow is 'biasing' the pH meter (current x resistance = voltage) and causing it to give erroneous readings. Just a hunch, I have no firsthand experience with such meters.

As a sanity check, if it is possible, is to put the pH meter in the sample and then run a current that is totally electrically separate from everything (skip the DAQ card and use a programmable supply?) through the sample to see if the pH meter readings change. If they do, then I don't think it is an isolation issue but that the current is actually the culprit.

A link to a schematic would be of great help here.

'Basic' electronics? (as opposed to acidic?) Is your 'neutral' grounded? SORRY, just couldn't pass up the puns!

[shoneill]

Scottie,

sounds kind of familiar. I actually developed pH electrodes for a few years, so I've run into things like this a few times (My background is in chemical sensors). A lot of the time, in order to push prices down (A quality pH-meter high-impedance input is really expensive) the ground of the instrument's communications port is not isolated from the ground of the instrument itself. I've seen many cases where simple having two sensors in one solution has lead to exactly this problem as soon as both instruments are connected to the same PC, and it was almost always solved by using an opto-isolator. So your first guess was quite good.

Another question. Does the current you're outputting have any electrical contact to the sample you are measuring the pH in? If so then I'm convinced this is the problem. The exact same problem is often seen when measuring pH and conductivity in the same sample. Try an opto-isolation and see if it fixes the problem.

Hope this helps

Shane.

PS Out of pure curiosity, which meter / electrode combination are you using in what kind of sample (Old habits die hard)?

[SCOTTIE]

AnalogKid2DigitalMan:

Actually, current biasing is not supposed to be the factor of pH shift. This was indeed an issue a couple weeks ago while two pH meter's electrodes (working and reference) were not put together. This is also called "IR drop". However, this issue was naturally solved after I tried to stick those two pH electrodes together. No matter how large currents from were delieved to the sample, the pH meter would not sense any difference due to currents. PH shift only happens when either DAQ card or pH meter was connected to laptop.

I'll try to get a schematic soon.


Scottie

[SCOTTIE]

shoneill,

Thanks for your reply. I do send a certain current into the sample when measuring real-time pH. As you mentioned, an opto-isolator would help solve the problem. What kind of the isolator do you mean? I suppose the pH shift is coming from "DAQ card(or OPAMP circuit) -> laptop -> pH meter", not from "DAQ card(or OPAMP circuit) -> sample -> pH meter". Therefore, what I'm thinking is to put an opto-isolator(rs232 interface) between pH meter and laptop. Do you think I am right? Actually, I have been reading a bunch information about opto-isolators packed in IC-style. Oh, they are tens times cheaper compared with usb or rs232 opto-isolators. Do you think those kind of IC-style opto-isolators might also work if I put them behind all voltage outputs from the DAQ card?

Beside, responding to your curiosity, the pH meter I'm using is FISHER AR15 equipped with a combination pH electrode, switching to Lazar ResLab micro mono pH electrode when dealing with micro-liter samples. What kind of sample? Actually I'm not quite familiar with that kind of biochemical stuff. Let me copy one of the names I couldn't remember at all for you: acetylcholine ester.


Scottie

[shoneill]

Hi Scottie,

To the best of my understanding (I developed electrodes, not the electronics, but I heard quite a lot from the guys who did the electronics), a shift in the two ground potentials can cause the A/D converter, because it's not isolated, to measure against the wrong ground. Ideally the reference electrode should form the "ground" for the A/D converter, but in a system where the pH input isn't isolated the ground and the reference electrode are often connected over a shunt resistor. In such a system, the real "ground" for the A/D converter is a wierd mix out of the reference electrode potential and the ground from the main supply (or battery if it's a bettery-driven meter). What happens then is that the "external" ground influences the "internal" ground over the shunt resistor and the signals shift. This is quite easy to imagine when you consider that the actual measuring circuit is a low-voltage high impedance circuit which is just begging to be corrupted.

The problem is basically a ground-loop where you end up shifting small currents through the reference electrode, thus changing the reference electrode potential. Because of the very small currents involved in pH measurements, it doesn't take much to have a big effect on the pH reading. Also, even passing small currents through the reference electrode will greatly shorten the lifetime of the reference electrode (The silver chloride will end up all being oxidised or reduced depending on which direction the current is flowing - a properly functioning reference electrode needs both metallic and reduced silver in immediate vicinity of the lead-off wire). Both problems will be solved by isolating the meter with an optically isolated RS232 connection.

It's truly an electronics problem. A properly isolated pH meter will not suffer from this problem (But they're expensinve of course). Simply sticking an RS232 optocoupler between the instrument and the computer usually does the job.

Shane.

PS

Do I understand that you are performing pH measurements in non-aqueous solutions? If so, make sure you're using the correct electrode for the job. Non-aqueous measurements are extremely difficult, and only the correct combination of electrode and meter will give correct results. This has a lot to do with the very high impedance of the solution which in turn greatly increases the impedance of the liquid junction of the reference electrode (Which a non-isolated, standard pH meter simply cannot handle). You then end up with a double high-impedance circuit. To get anything close to good readings (which is hard at the best of times in non-aqueous), you need a good meter AND a good electrode. I've published on this matter before. A sleeve-type reference electrode is really the only choice for non-aqueous measurements. Do a search for "Solvotrode" to see what an optimised non-aqueous pH-electrode looks like.
<\rant>

Message Edited by shoneill on 03-01-2005 09:22 AM

[AnalogKid2DigitalMan]

Scottie:

You can build up your own RS-232 isolator circuits out of discrete isolators, but do your design homework to prevent communication issues. Personally, I'd spend the couple hundred bucks for an off the shelf RS-232 unit. Black-Box

Optocouplers will not work for the DAQ outputs unless you design a linear circuit. Opto's are typically used as on/off devices, but can be coaxed to work in the linear realm. Again, there are off the shelf modules to buffer and isolate voltage outputs for a DAQ card. Dataforth, Analog Devices

[SCOTTIE]

I'm supposed to receive the opto-isolator today shipped by UPS. Hope it works. Wait for my good news. 😄