• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 49 subscribers
  • Views 1341 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

INA116 pH meter questions

Janez Vizjak
Prodigy 30 points

Other Parts Discussed in Thread: INA116, LMP91200

Hello,

I am a chemistry student, and I'm currently trying to build a custom pH meter. I've decided to use the INA116 amp because it provides its own guards, and has an increddibly low input bias current. This is perfect for pH electrodes that have internal resistance of several hundred megaohms. However after reading the datasheet I have a few questions about this IC.


A) First of all I'd like to know a few details about the input bias current path. The datasheet suggests (page 8, figure 3) that some resistors be connected between each amplifier input pin, and the ground. According to that schematics I see two 100 MOhm resistors connected to each input, and the ground. Is this correct - are these really needed in my case? The sensor I will be using is a pH electrode, and they have internal resistances up to 1000 MOhm. So basically what we just did here is connected a 200 MOhm resistor in parallel to the INA116 inputs.

This means we just added a low-resistance bypass path for the already-weak current that is originationg from the electrode. We are basically measuring the voltage drop on the 200 MOhm resistor, not the pH electrode. The resistance/impedance of the INA is so high that it can be neglected. The pH electrode gives a voltage of 413.7 mV at best, and let's assume the internal resistance is 1000 MOhm. The total resistance of the current path is therefore 1000 + 200 = 1200 MOhm. According to ohm's law (I=U/R), the current at this voltage will be 3.4475E-10 A. That's about 0.3 nA or 344.75 pA. The voltage drop will therefore be U=I*R, and the resistor causes the voltage drop of 68.95 mV, where the electrode has a voltage drop of 344.75 mV due to its internal resistance.

We're therefore measuring 69 mV instead of 413 mV - that's about 83% rel. error we are causing by adding these two resistors here!
What os the proper solution to this?


B) The datasheet also states that guards should fully encircle both input pins on both sides of the PCB. In this case how are the former resistors (assuming we're measuring a lower-impedance source, so that the previous question doesn't apply) supposed to be connected to the input pins? Can they be connected over the guard?


C) What is the maximum output current that INA116 is able to provide? According to the datasheet on pg 2, this would be +5 / -12 mA. Is this correct? If so, what is the proper way to level-shift the voltage to adapt it for a 5V ADC like AD7791? A resistor voltage divider can cause a large error in the ADC measurement. Too low resistor values cause a large voltage drop. Too high, and the ADC doesn't get enough current. What's the proper way to voltage-shift, if the INA is being powered by a single-end power supply of 12V, and has a virtual ground of 6.0V? The ADC is on 5VDC (12V => LM7805 => 5V).


D) Can you offer any other tips&tricks, hints, comments or suggestions about pH meter design?


Thanks a lot.
Best regards!

  • 0
    Prodigy 30 points

    Update!

    A) I read the datasheets again, and I believe I somehow misunderstood the wiring concept. The document doesn't seem to mention it, but I take it the two resistors must only be used when there is some sort of capacitive device (crystal, capacitors) used in front of the sensor? So the resistors are not needed for a pH electrode I take it, as seen on page 9 (figure 7) in the datasheet.

    Anyway, I spotted another potential problem. In the figures 5 and 7 the solution seems to be grounded despite the differential inputs being used. What is the proper way to connect a pH electrode with a single coax cable to INA116? The cable has no grounding present. My conception was that the differential inputs should always use twisted pair cables, but in this case
    I cannot change the coax cable of the electrode. Does the solution absolutely *have* to be grounded?


    B) If the resistors aren't needed then this answers question B.


    C) Looking again at my ADC datasheet it seems that it already has a buffer present, and the typical analog input current is about 1 nA. This should be low enough to use on a simple two-resistor voltage divider. I'll probably want to use higher resistor values here so that the amplifier is not starved of current. The ratio should be 7 to 5, and this will also protect the ADC from overvoltage, provided that the supply voltage does not rise above 12V. Unless I get a better idea, I am going to use the voltage divider here.


    D) Still waiting for some hints!



    Best regards!

  • 0 in reply to Janez Vizjak
    TI__Genius 15410 points

    Hi Janez,

    Here are the answers to your first forum post:

    A) Yes, resistors are needed to provide a path for the input bias current. Please see this forum post regarding the use of resistors to provide a path for input bias current http://e2e.ti.com/support/amplifiers/precision_amplifiers/f/14/p/347357/1220299.aspx#1220299

    Also here is a blog about providing a path for input bias currents http://e2e.ti.com/blogs_/archives/b/thesignal/archive/2012/03/27/instrumentation-amplifiers-avoiding-a-common-pitfall.aspx

    B) You should connect the guards as shown in Figure 4 and Figure 5 in the datasheet. Be sure to fully encircle the input pins on both sides of the PCB.

    C) Yes you are correct, this is the maximum current the device can provide. However, as the device increases the current it is supplying the voltage output will decrease and approach mid supply of the device. To avoid this from happening I recommend applying a reference voltage of 2.5V (half of the 5V ADC) to the reference pin and then setting the gain of the device such that the output swings from 2.5V to 5V for a positive differential signal and then from 2.5V to 0V for a negative differential signal. Be sure to supply the INA116 such that the output can swing from 0V to 5V. Please see voltage output limitations on page 2 of the INA116 datasheet.

    D) Here is an application note about pH sensors. http://www.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=snoa529a&fileType=pdf

    I recommend searching the web for any application notes about pH sensors.


    Here are the answers to the second forum post:

    A) The INA116 is intended to be used with two separate electrodes, a pH electrode and a reference electrode. This can be seen in Figure 7 of the datasheet. More commonly used today are the combination electrodes which contain both the pH electrodes and reference electrode in a single sensor. The combination electrodes would not be able to provide a connection for the guard rings because as you found they use a single coax cable. You might want to consider the LMP91200 for your design.

    B) Answered above.

    C) Answered above.

    D) Answered above.