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.

Op Amp with low "low-level output voltage"

Other Parts Discussed in Thread: TLV4110, OPA569

I am trying to design an inexpensive solar cell current-voltage tester for educational modules related to solar energy.  The purpose of the circuit is to sweep an applied voltage across a solar cell and measure the current out.  At the moment, I'm using an Arduino Nano microcontroller to program a DAC for the voltage sweep and to read the current.  I am measuring the current as the voltage drop across a 0.1 ohm resistor between the solar cell and ground.  I am using TLV4110 op amps to buffer the current in the solar cell from the DAC and to transform the voltage across the resistor to a useable value for the 0-5V analog inputs of the microcontroller.  I am trying to attach a schematic of the diagram.  The "DUT" in the diagram is the solar cell, and solar cells can be modeled as a current source (current is proportional to light intensity on the cell) in parallel with a diode.

Unfortunately, the TLV4110 has a pretty high "low-level output voltage", and it is very important that I be able to apply voltage to within mV of zero at high current.  Is there an op amp that can handle 200+ mA and has a low-level output voltage in the mV range?

 

Thanks very much,

David

  • Hello David,

    The TLV4110 is remarkable in that it can deliver hundreds of milliamperes of output current from an SO-8 package. The increase in low-level output voltage (Vol) with high output current (Io) is normal and is due to the voltage developed across one of the output transistors as higher current is demanded by the load. To reduce Vol at a specific current level such as 200mA, a higher output current device is required.

    An operational amplifier that will pull very close to the rails at 200mA is the OPA569. The figures shown below are from the OPA569 datasheet Typical Characteristics curves section. It can be seen from them that the OPA569 will swing to a level about 20mV from the rail at 200mA. That is about as good as it gets for an operational amplifier at these high current levels.

    The OPA569 isn't in an SO-8 package like the TLV4110 because of its high 2A output current rating. It is available in a SO-20 PowerPad package which is very small as well. It includes a number of built-in safety characteristics such as current limiting and a thermal flag which are very desirable in power applications. Here's were you can find more information about the OPA569:

    http://focus.ti.com/docs/prod/folders/print/opa569.html

    I hope this helps resolves your circuit.

    Regards, Thomas

    PA - Linear Applications Engineering

  • I realize I was not very clear about my other requirements.  It would be ideal if the op amp was through-pin, could operate off USB power (0-5V supply), and could handle 0-5V in and out (I plan to use it as a buffer amp, so it will be unity gain, but I'd like to apply 0-5V to the solar cell).  Does such a device exist?

     

    Thanks again,

    David

  • David,

    If I understand your application correctly, you may need to source current at close to ground potential. These op amps should be able to do this. The specs regarding output voltage close to ground are for sinking current.

    If my assumption is correct, the issue may be power dissipation under this condition.

    Please clarify and we can help further.

    Regards, Bruce.

  • Hi Bruce,

     

    The op amp needs to sink current near zero volts (this is where the illuminated cell is producing the most current). I will only source high current at high(ish) voltages, and I am not worried about how close I can get to 5V.

    It would be ideal to find an op amp in a through-pin package that runs off 0-5V and can get me within millivolts of zero while sinking 100-200mA of current, but if such a component does not exist, could I convert my 0-5V supply to +/-5V with a differential amplifier and then use an op amp that runs of +/-5V?  My understanding is that my voltage is only limited near the limits of my supply voltage, so an op amp that can sink 200mA and runs off +/-5V should be able to sink 200mA while supplying 0V, right?  If this is the only solution, what would be a good component to select?

     

    Thanks,

    David

  • David,

    Okay. My conclusion that the op amp was sourcing current was based on the fact that the op amp to the right of the DUT can only measure current if the high current op amp is sourcing current. If the high-current op amp is sinking current, the current measurement op amp's output will be saturated at near ground potential. The measurement op amp will have a negative input voltage.

    I've discussed this with Thomas and we agree. If you need to sink 200mA very close to ground, we see two possible solutions.  1) Some type of negative power supply.  2) A discrete power MOSFET circuit using a very low on-resistance power MOSFET. To sink 200mA at 10mV you would need a MOSFET with an on resistance of 0.05-ohm. Consider also, that you are losing 20mV across the shunt resistor at 200mA.

    If we misunderstand, please come back with some more information.

    Regards, Bruce.

  • Hi Bruce,

    Thanks for pointing out my mistake with the op amp to the right of the DUT. 

    I'd like to try to implement the MOSFET solution, but I don't know much about FET's.  If I want to continue to power this circuit using a 5V power supply and sink 200mA close to ground and source 200 mA around 1-2V, what specs do I need to take note of when selecting an FET?

     

    Thanks,

    David

  • David,

    I've shown a possible implementation. The DUT is shown as a diode and current flows the "wrong way" in the diode. Note the reversal of inputs on the op amp because the MOSFET provides an inversion at the output. It may also add gain to the loop creating a stability issue. That's the reason for R3/C1, which may help. It all depends on the impedance the DUT creates as a load. It would require experimentation.

    Key spec for the MOSFET would be the Ron which needs to be less than 50mOhm with 5V gate drive.

    Regards, Bruce.

  • Bruce,

    Just to clarify, in the configuration in your diagram, the DUT is reverse biased?  Would the circuit still work if I put the DUT into the circuit the other way?  And are you assuming the DUT is sinking or sourcing current in this configuration or could it do both?  I am guessing that I would measure the voltage between R2 and the DUT and measure the current as a voltage at VF1.  Is that correct?  And assuming I am both sourcing and sinking current in the DUT in a single voltage sweep, wouldn't IOP2 have to be bipolar? 

     

    Thanks again,

    David

  • David,

    I probably should not have shown the DUT as a diode as everyone seems to get confused by this model of a photovoltaic cell.

    The DUT gets connected the same way it does in your original circuit. The positive output is on the top and the MOSFET sinks the current generated by the cell. The voltage input applied at VG1 is forced on the DUT (not accounting for the voltage loss on the shunt resistor, R1). VF1 measures the voltage on the shunt with a gain of 20.

    I think this is what you want. You are trying to force a voltage on the cell and measure the current it produces under that condition... correct?

    Regards,  Bruce.

  • Bruce,

    This is what I want.  My only concern is that under illumination, at low voltages, the solar cell will produce current and the MOSFET will have to sink it, but at higher voltages (above 0.5V or so), the solar cell will draw current.  Will the MOSFET/op amp circuit to the left of the solar cell be able to source this current?  And will the current measurement op amp be able to handle this situation if it's not bipolar?  At the moment, I am using a differential amplifier to convert my bipolar current measurement into a unipolar measurement for the ADC.  My understanding is that this does not require a bipolar op amp, since the output is unipolar even though one of the inputs is bipolar.

     

    Thanks,

    David

  • David,

    My circuit can only sink current. Sorry. It gets more complicated to source current as a second P-ch MOSFET would be required.

    Also, your circuit and mine can only measure one polarity of current. You could maybe get fancier with both a inverting and non-inverting amp on the current shunt.

    Bottom line--all this gets very messy in order to avoid a negative supply. Yet another possibility is to create a pseudo ground at about 2V and float the cell at this voltage. This would require a second high current op amp and you would need a difference amplifier to remove the common-mode voltage from the current measurement.

    Regards, Bruce.

  • Bruce,

    If it's only as complicated as adding a p-ch MOSFET, I'd still be very interested in how this is done.  I'm attaching an updated circuit diagram which I believe should be able to (and seems to) measure current from -250mA to +250mA using a 0-5V ADC as well as an example of the IV curves I'm getting from it.  The second channel on the DAC (the one not sweeping voltage on the DUT) is putting out .089V to the differential amplifier.

    I also thought just using a bipolar supply would be easier, so I tried using a benchtop one I had as well as a 0-5V to +/- 5V converter to test replacing the TLV 4110's with a bipolar op amp with no success.  The converter ended up only being able to supply about +/- 4V, so it wouldn't power the op amps, and the bench supply burned out the op amps.  This experience made me think finding components for a bipolar supply would be equally if not more difficult as well as adding bulk to what I am hoping can be a small device.  Hence, looking into how difficult the MOSFET solution would be.

     

    Best,

    David

  • David,

    In concept, adding the ability to source current is as simple as adding a P-channel FET driven by the same op amp and with its source connected to the 5V supply. The devil is in the detail. The problem is in controlling the gate voltage drive for both the N and P so that both FETs are not "on" at the same time. This involves careful selection of the FETs for the proper threshold voltage or additional circuitry to control the gate voltages with separate signal. Not so easy.

    Regards, Bruce.