transimpedance amp for electrochemical cell

Discussion in 'The Projects Forum' started by mdm1955, Nov 19, 2009.

  1. mdm1955

    Thread Starter New Member

    Nov 19, 2009
    4
    0
    Hi All,

    I am planning on using the circuit shown in Fig. 4-52 of the attached .pdf (Op Amp Applications Handbook, Walt Jung) as the basis for current measurement in an electrochemical cell. Similar to a photodiode, an electrochemical cell is generally modeled as a resistance in parallel with a capacitance, but depending upon the geometry and electrolyte, the capacitance and resistance can vary greatly and be highly nonlinear. The current range will probably be from about 10nA to 1mA. The current is essentially DC (apart from noise) with possible negative or positive step transients. The low pass filter I intend to use will probably cutoff at 10kHz or lower, in part due to HF/microwave excitation that may be applied to the cell cathode.

    My questions are:

    1. Since the currents I will be measuring are at least three orders of magnitude greater than those associated with the circuit shown, could a 1Mohm digital potentiometer (AD5222) be used in place of the 1000 Mohm resistor?

    2. Would a different op amp (e.g., OP27 or OP07) be more suitable than the AD795 at higher currents?

    3. With respect to the input offset null portion of the circuit, could the 100k resistor with +/- 15V supply be replaced by a software controlled analog voltage output from a data acquisition board?

    TIA
     
  2. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    More data would be nice. Stick a copper wire and a galvanized nail into a potato, and you have an electrochemical cell. Yours is perhaps different.

    You speak of a
    - is that current in the internal or an external circuit?

    To disturb the cell as little as possible, an electrometer grade op amp might be the thing, rather than a common FET input type. Look at an OPA129 for specs (and availability).
     
  3. mdm1955

    Thread Starter New Member

    Nov 19, 2009
    4
    0
    The attached schematic shows the driven electrochemical cell (e_cell.jpg). The electrode solution potential is the potential developed at electrolyte/electrode interface. There is a faradaic component of resistance which is a function of the applied voltage and the nature of the chemical species involved. The ohmic (bulk) resistance is a function of the electrolyte composition, electrode areas and separation. The capacitance is attributed largely to the double layer, and can be as high as tens of microfarads per square centimeter. My cathode area will be about .05 sq. cm or less. There is no convenient way to numerically model the cell/electrolyte combinations since basic property data is not readily available for a lot of the materials I will be using.

    I will be using water as an electrolyte solvent and I also expect to be using liquid sulfur dioxide (room temperature, 100psi). The salts I am using will have a range of solubilities. My questions concern the need for adaptability and flexibility of operation while maintaining low noise and a degree of transient response. I will be using a cio-ssh16 in conjunction with a CIO-DAS1602/16 for measurement and control (see attached manuals). This is why I would like to modify the input offset null (question #3). I wouldn't expect the input offset null to depend on the nature of the electrochemical cell.

    The purpose of the setup is to investigate spin modification of electrolytic reactions. A swept DC magnetic field will be used to vary the Zeeman splitting in reactive species. Electron spin resonance (if RF excitation is applied) or other phenomena will be detected by a change in the electrolytic current.

    As you can see from page 12 of the attached AD797 datasheet, the AD549 (electrometer grade) is one of the op amps recommended for high source resistances, whereas other op amps are recommended for lower source resistances. Since I may be working with lower source resistances, I am interested in what modifications would be necessary to Jung's circuit to optimize it for use at higher currents and lower source resistances. The electrometer grade AD549 is used with lower currents than the AD795 that Jung is using, and I expect to be using currents higher than those measured with Jung's circuit. I have already looked at the OPA129, and it is typically used with much smaller electrochemical cells and much lower currents than I anticipate. See U.S. Pat. No. 7,444,856.

    I appreciate your response to my post, but I think I would be better off using Jung's circuit as is, rather than design something else with the OPA129 or AD549. It wouldn't matter if my cell was a potato with copper and zinc. I still have to measure current over a broad range, and when it comes to precision low noise current measurements, there is no one-size-fits-all device that I have seen. That's why I intend to start with Jung's circuit and modify it.
     
    Last edited: Nov 20, 2009
  4. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    No matter the setup, it will be interesting to assign any change in the op amp output to change in internal resistance or actual change in voltage.

    From experience, I might suggest using the op amp as a buffer with small gain, as in the attached schematic. I like the Burr-Brown op amps, but there is no reason not to sub in an Analog Devices unit. Just select such that the op amp input impedance is 3 orders of magnitude above the input resistor, R1. I chose one gigaohm, as Ohmite makes values in that range in 1% tolerance. 1% in a gig ain't too breath-taking, but it's better than paying for .001%.

    Trying to adjust for current is less important than insuring that the signal be noise free. Placing the buffer in a metal enclosure in close proximity to the cell and using coaxial cabling to send the signal to another amp for further amplification works very well.
     
  5. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    BeenThere, the circuit you posted is for measuring voltage. Mdm1955 wants to measure resistance. Did I miss something?
     
    Last edited: Nov 20, 2009
  6. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    Why, I'm so tuned into electrochemical cells as voltage generators that the -
    - clue blew right past.

    Even with the external voltage source applied, my methodology is not incorrect. The fixed resistance is not good, though. Use a good quality rotary switch to switch in a number of fixed resistors and the voltage variance will reflect the cell's internal resistance. I've been doing something like this for years.

    If the inverting amp is preferred, then consider switching feedback resistors to alter the response to the high/low cell resistances.
     
    Last edited: Nov 20, 2009
  7. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Forgive me but looking at your PDFs, won't the bus and rest of the system be defining things for you?

    As Been There said this looks like an interesting project. Is there any chance of an overview diagram so we can get a feel for what you need?
     
  8. mdm1955

    Thread Starter New Member

    Nov 19, 2009
    4
    0
    I appreciate the interest and suggestions. This is a complex project and I am working at my intellectual limits. I posted some material on sciencemadness.com earlier this year. There are a number of schematics and reference articles in the following two threads.

    I first posted on this thread (as merrlin):
    http://www.sciencemadness.org/talk/viewthread.php?tid=12024#pid150436

    I then started this thread:
    http://www.sciencemadness.org/talk/viewthread.php?tid=12028&page=2#pid150769

    As you can see, there were some people that were interested and helpful, but there was also one in particular that took issue with my project. I have since talked with Buchachenko and corresponded by email with him and other experts on spin chemistry. The consensus is that this is a novel approach and it cannot be dismissed on a theoretical basis. It comes down to experiment. For those of you who have slogged through the sciencemadness.com threads posted above, I would be happy to answer any questions.

    My background is materials science and my electronics experience is primarily in power and microwave. I am also familiar with PC based data acquisition. I have negligible experience in precision amplifiers and low level signal processing. For those of you who have slogged through the sciencemadness.com threads posted above, I would be happy to answer any questions. I expect this project to take years, and I am setting up a small CNC Taig mill for parts fabrication.

    My immediate design focus is on current-to-voltage conversion for the electrolytic current so that I can get a ~0.1 to 1V input to the SSH-16 simultaneous sample and hold board. Until recently I was planning to use a series current sense resistor, but then I learned that a transimpedance amplifier can be used for current-to-voltage conversion without loading the current source being measured. Speed isn't really an issue, but stability might be, due to the potentially large capacitance in the electrochemical cell.

    I expect that optimization will be largely trial-and-error and involve both the electrochemical cell and the associated circuits, but I know there are people here who could help keep me from getting lost.
     
  9. mdm1955

    Thread Starter New Member

    Nov 19, 2009
    4
    0
    There's something I should add that I hope will clarify my problem. I expect to be driving the electrochemical cell with a voltage of about 0.2-10 volts, so the gain will typically be close to unity, particularly if I use a higher supply voltage op amp and take a 10 volt maximum output to the SSH-16 instead of the 1V I mentioned earlier.
     
Loading...