• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 49 subscribers
  • Views 1109 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TL3016 comparator functionality

sebastian wartmann
Prodigy 20 points
Other Parts Discussed in Thread: TL3016, TLV3011, LM397, TLV3502

Hi,

within my master thesis I´m using the analog comparator TL3016 to compare the the voltage of a solar cell to a reference voltage. The solar cell is directly connected to the positive input of the comparator. The reference voltage is connected with a 100nF capacitor in parallel. I didn`t connect a capacitor at the positive input, because I want to measure very fast changes at the solar cell due to high velocity impacts.

I tested the circuit before with an external voltage source and everything worked. At a difference voltage of 15mV between the inputs, the output switched.

Now, with the solar cell, the output switches at a voltage difference of 300mV between positiv an negative inputs! Is this a problem, because the solar cell produces less current (100µA)? And between these voltages the comparator has not a definde state (looks like floating).

Another problem occurs, if I`m placing a probe at the output of the comparator. The output seems to influence the inputs and its starts so oscillate a little. Is this because of the capacitance (0.9pF) of the probes? How can I avoid this?

Or maybe I have to add some more components to stabilize the device? The dadasheet gives no information about that.

Thanks!!! Kind regards,

Sebastian

  • 0
    TI__Guru* 93105 points

    Hi Sebastian,

    The TL3016 is an ultra-fast comparator and it may be oscillating in your circuit. The 300 mV threshold it is requiring to change output state with with the solar cell input may be a the result of oscillation. Normal dc behaviors may not be achieved when a device oscillates. The oscillation is often the result of noise being injected into an input and the comparator switches very quickly back and forth in response to it. The solar cell may be picking up noise from surrounding sources and coupling them into the comparator. Removing the capacitor allows for almost any noise frequency to couple into it.

    You mention the solar cell's output current capability, but you do not mention anything about its output voltage levels. Are you sure that the output range conforms with the TLV3011 input range relative to the supply voltage and any input reference voltage you have established? The TL3016 input current being in the range of 6 to 10 uA does represent a 6 to 10%  loading of the cell's 100 uA capacity. How is the input bias current provided when the cell is dark and not producing output current?

    Did you observe high-speed layout practices with your PC board; low impedance ground returns, proper power supply decoupling, etc? If you did, then you may want to consider adding hysteresis to your circuit to reduce the comparator's likelihood for switching on noise. Take a look at another comparator's data sheet such as the LM397. That data sheet provides information about comparator oscillation and how to apply hysteresis to help improve a comparator's immunity to it. There are many other resources on line about comparator hysteresis.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Prodigy 20 points

    Thanks for the answer!

    The output voltage of the cell is max. 2.3V. So its in the "allowed" range of the comparator input voltage. Now, I`m trying to built the circuit as simple as possible. It means that I have a reference voltage and I try to move this voltage under and over the voltage from the solar panel. The output of the comparator changes at almost +-300-500mV. So I dont have very high frequency changes at the input, but somehow it doesnt work.

    Im not sure if there is a problem by connecting the solar cell directly to the pos. input. Maybe there are some parsitic effects (except noise)?!

    I considered power supply decoupling etc.

    Regards,

    Sebastian

  • 0 in reply to Thomas Kuehl
    Intellectual 540 points

    Hi Thomas,

    I'm looking for a high-speed dual comparator for comparing the composite video signal coming from an analog monochrome camera against two thresholds (one comparator for each threshold). The pixel clock of the camera is 28.6MHz. For prototype I used KA319 and noticed oscillations and propagation delay issues at the output. Spec. of TL3016 seems to be OK. But I need a dual channel part.

    1. Has there any dual version of TL3016 in 8-pin package? I'm OK w/o latch and complementary o/p.

    2. The actual amplitude of video information shall vary b/w 0-700mV. And the thresholds against which I'm comparing it shall be from 2-400mV. From TL3016's datasheet 0.5 to 3.0mV of hysteresis would be OK? Does this hysteresis cause any other problem since the lowest threshold which I need to compare is 2mV?

    Kindly suggest!

    Regards, Sinoj

  • 0 in reply to Sinoj B S
    TI__Guru* 93105 points

    Hello Sinoj,

    The TL3016 is only available as a single comparator. What your application appears to require is a precision, high-speed comparator having both low voltage offset and minimal internal hysteresis. We do have one dual, high-speed comparator, the TLV3502, that is even faster than the TL3016. However, it has typical voltage offset of +/1 mV and a hysteresis voltage of 6 mV (typ.) which in some combinations would be an issue when you are trying to compare a difference of 2 mV.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Intellectual 540 points

    Hi Thomas,

    Thanks for your reply.

    I checked datasheet of TLV3502. Speed is OK and I would be powering it from 0-5V. Suppose the lower threshold is now 10mV instead of 2mV. Then would you forsee any issue with the 6mV hysteresis?

    Regards | Sinoj

  • 0 in reply to Sinoj B S
    TI__Guru* 93105 points

    Hi Sinoj,

    A 10 mV lower threshold should be large enough to overcome the full voltage-offset range and the built-in hysteresis of the TLV3502.

    Regards, Thomas

    PA - Linear Applications Engineering