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.

LOG114: Problem getting the circuit to work

Part Number: LOG114
Other Parts Discussed in Thread: REF3040,

I modified the LOG114EVM board to create this circuit:

For the first step I didn't connect Vlogout to R5 in order to just evaluate the first amplification step.

Unfortunately I cannot get the circuit to work.

V+ = 5V, V- = GND

Vref is correct +2.5V, REF3040 generates 4.096V and instead of a photodiode I use a LDO with potentiometer and series out resistance to define the current. Using a Keysight B2985 current meter I measure this input precisely, it ranges between ~5nA and ~1mA.

However Vlogout always shows the same picture:

What am I doing wrong? Or did I destroy the OpAmp during my tests?

Obviously my setup is less than ideal (note that I did quite a lot of modifications to the board, however I'm positive that the input lines I1 and I2 aren't connected to anything else. Additional 100pF caps on I1/I2 didn't improve the result in any way.

  • Hi,

    what happens when you disconnect the 1M62 resistor?

    Kai
  • Hello MMx,

    Sorry to hear you are having difficulties getting the LOG114 circuit to work properly. I use the LOG114 EVM often with a photodiode, or an external power supply/resistor as the I2 source. In both cases the LOG114 functions as expected. Your circuit is the LOG114 single-supply connection shown in datasheet Fig. 2. Since I have used this circuit numerous times I don't find anything about to question.

    A common problem with very low input current circuits such as the LOG114 is PC board flux contamination. Flux residue from soldering is not an ideal insulator, and it can set up all sorts of current leakage paths between points in a circuit. This can result in unexpected electrical behavior. You have been "operating" on the LOG114 EVM and unless you have taken extra steps to clean the boards that might be the problem. 

    The EVM board as received from the manufacturer would have been properly cleaned. We specify solder having a water soluble flux. Assembled EVM boards are ultrasonically cleaned in DI water. When we assemble boards for our internal projects we actually run them through the ultrasonic cleaning process twice.

    Try cleaning the EVM board and see if that makes a difference in the LOG114 behavior. Then, I suggest attaching a DVM at the VLOGOUT pin and observing the VLOGOUT voltage as you adjust the input current. Determine if VLOGOUT conforms to the transfer equation:

    VLOGOUT = 0.375 × log(I1/I2) + 2.5V

    If it still doesn't work, then that points to a circuit hook-up problem or possibly an issue with the LOG114 device.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • hey kai klaas,

    When I short circuit Iin2 and Vref I get Vlogout=4.8V, when I leave it open (i.e. Iin2 floating) I get 4.8V but with 50Hz ripple (930mV voltage drop) from the supply power line.

    When I leave both Iin1 and Iin2 open I get Vlogout ~sine curve between 1.8 and 4.1V with 50Hz frequency

    Michael

  • Hi,

    can you check your wiring? Is there a short circuit? I can see on the photo that there are very thin wires soldered onto the PCB. If there's a short circuit the supply voltage could go out of regulation and you might see the 100Hz rectifier ripple from the input of regulator.

    As Thomas has already recommended, have you carefully cleaned your PCB after the soldering? Take alcohol for the cleaning and rinse with destillated water. Wait for a complete drying of PCB before you power-up the circuit.

    By the way, you should short the inputs to any potential without providing a current limiting...

    Kai

  • After further research I come to the conclusion that the circuit is working, just not in my specific application.
    When I connect various resistors between my 4.096V voltage reference and I2- I get the desired values at Vlogout and Vout.

    Now I try to add a corrosion sensor which outputs a current proportional to the corrosiveness of the environment. When I connect the corrosion sensor to a resistor through a current meter I measure something along 30uA when the sensor is dry and 0.5mA when I pour water on it.

    I connect this sensor between I2- and GND (I also tried Vref instead of GND but it made no difference) but the output always remains Vlogout=4.8V, Vout=0.04V
    If I connect a current meter between I2- and the sensor I measure actually current flowing backwards into the sensor which is weird as when not connected to the circuit the current flows into the other direction.

    When not connected to anything the sensor outputs ~0.6V however if I connect a simple load (e.g. 750R) the output voltage follows the current (e.g. 0.11V @ 150uA) so to all of my knowledge the sensor behaves like a real current source.

    What might the problem be with this specific sensor?
    As it is a custom made sensor there is no datasheet available.
  • Hi,

    sounds like the problems come from the sensor. Can you explain how the sensor works?

    I guess this sensor is not only emitting a current but also adds some insane voltage to the circuit. Or it needs a voltage. A photodiode on the other hand is very simple: You can easily force a zero voltage across anode and cathode and it behaves like a true current source then.

    Kai

  • Hello MMX,

    I am glad to hear that the LOG114 portion of the application appears to be working correctly. However, the inability to get the corrosion sensor to behave as desired is a bit perplexing and probably due to something unexpected.

    I don't know what the sensor look like electrically, or how it electrically connects to the input of the LOG114 in your application. Does it have long, open wires going to the LOG114 input? I am wondering if the sensor is picking up an RF field and rectifying it, producing a current that isn't compatible with the LOG114 current input. The LOG114 can only accept current flow into the input. A wide bandwidth scope, with a low capacitance probe across the sensor may reveal something unexpected.

    If the input current is flowing in the opposite direction that the LOG114 input can tolerate, then the datasheet Figure 7. Current Inverter/Current Source is a good current reversal circuit. I have had better, more accurate results with it than the one in Figure 8.

    Regards, Thomas
    Precision Amplifiers Applications Engineering
  • MMX

    We haven't heard back from you so we assume this resolved your issue. If not, post another reply below.

    Thanks
    Dennis