• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 48 subscribers
  • Views 1686 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.

ACF2101 Problem with low currents

Yurii Morozov
Prodigy 40 points
Other Parts Discussed in Thread: ACF2101

Hi!

I want to use ACF2101 as a photodiode amplifier.

I connect the ACF2101 as in typical connection on Fig 1a in datasheet, photodiode connected only to one channel. I am not using select switch, the supply voltage is +-15 V.

Typical hold pulse duration was about 5 mS, reset 1.5 mS and the reset pulse is shifted by 2.3 mS from the beginning of the hold pulse,  pulse repetition rate 10 Hz, and the integration duration 95 mS.

At low current levels (photodiode was in complete dark, expected dark current about 250 pA) the output of the ACF2101 after starting integration cycle goes up (for about 5 mV in 100 mS), instead of going down to negative voltages, see figures below

Zoomed portion around the end of integration cycle:

Green dashed line is scaled hold signal and red - reset.

When I increase the current the slope of the output signal is decreasing, so at some point it can be flat and when you increase it more it looks something like this:

Zoomed portion around the end of integration cycle:

And the beginning of the integration cycle:

Also here are the timing diagrams of the hold and reset separately:

So my questions are:

1. why does the output voltage goes up at low current levels?

2. why does the voltage drop at the beginning and at the end of integration cycle is so big (around 15 mV)?

and why the drop at the beginning of the integration cycle is positive? (if I understand it correctly it should be negative)

So far I have tested two devices and both channels on those, qualitatively they all behave similar way.

I will be appreciate for any help with this issues,

thank you!

  • 0
    TI__Guru* 93105 points

    Hello Yuri,

    I don't have answers to your specific questions yet because we don't know if the behavior is specific to the ACF2101 devices you have, or something being caused by the physical, or electrical, implementation of your circuit. Their performance is vulnerable to extraneous circuit currents and capacitances, and coupled fields. Good board layout, and sometimes shielding, are critical to attain the expected performance. The output voltage level shifts appear to coincide with the hold and reset commands indicating the input current is changing abruptly when they change state. The shift could be related to charge transfer.

    You mention that you are using the basic circuit of figure 1A. That figure imparts the basic ACF2101 connection information, but lacks the details about input guarding shown in figure 1C. Is your circuit constructed on a laid out PC board and does it incorporate input guarding? Also, is the circuit setup open to the environment, or shielded from it? You may want to try placing the unused section of the ACF2101 in a know state rather than letting pins float. That probably won't make a difference, but at least rules out it as being a cause.

    Would it be possible for you to share your schematic?

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Prodigy 40 points

    Hi Thomas,

    thank you for your reply,

    here is the schematic of my device

    Zero Ohm resistors in schematic for working with external capacitors, I tried with external capacitor and had the same behavior.

    and the board layout I have so far:

    I hope that is clear.

    Reset and hold inputs are connected to the functional generator through coax cables.

    I tried to apply hold and reset to one and both channels, seems like no difference, as well as using one and both photodiodes, though SELECT A , SELECT B, SW OUT A, B, SW COM A, B are not connected.

    The board was shielded, I put it to the grounded metal box which is a bit larger then the board itself, when the board is shielded the output is less noisy but the same behavior.

    My first device I made on soic breadboard and the same issues with that one.

    Regards,

    Yurii

  • 0 in reply to Yurii Morozov
    Prodigy 40 points

    Forget to tell that PCB board is single sided

  • 0 in reply to Yurii Morozov
    TI__Guru* 93105 points

    Hello Yurii,

    Thank your for the schematic and PC board information. Everything looks to be straightforward so the problem is not on the surface. Here are some of my thoughts and things you may want to evaluate one at a time:

    • Are the output voltage shifts you are observing identical for different ACF2101 devices, i.e. do the follow the same pattern and have approximately equal levels?
    • The ACF2101 data sheet does call for 1 uF power supply decoupling capacitors and according to the schematic you are using 10 uF capacitors. I suggest changing them to 1 uF ceramic capacitors. Also, if you could add a much lower value capacitor such as a 10 nF, or 100 nF, closer to the device to pin, to ground, I think that will put to rest any questions about power supply decoupling.
    • I have not seen an ACF2101 application before where a series resistance has been added to the photodiode circuit. Try replacing the 200 Kohm resistors with 0 ohm jumpers.
    • You mention that the Reset and Hold pins are connect to the function generator by coaxial cables. Even though the switching rate is slow, the edge rates appear to be fast. The generator inputs do not appear to be terminated at the board. Try adding 50 ohm surface mount resistors from the Hold and reset Inputs to ground.
    • See if changing the generator edge rates has any effect on the output shift.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Yurii Morozov
    TI__Guru* 93105 points

    Yurii,

    Thank you for the additional information.

    Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Prodigy 40 points

    Hi Thomas,

    thank you for looking at my issues,

    some quick answer now,

    Thomas Kuehl said:
    I have not seen an ACF2101 application before where a series resistance has been added to the photodiode circuit. Try replacing the 200 Kohm resistors with 0 ohm jumpers.

    what about this application note, Figure 7: http://www.ti.com/lit/an/sboa032/sboa032.pdf

    I will answer your other questions and try your suggestions a bit later,

    Thank you!

    Yurii

  • 0 in reply to Yurii Morozov
    TI__Guru* 93105 points

    Hi Yurii,

    Thanks for bringing that ACF2101 application note to my attention. I had seen the note somewhere before, but I am not intimately familiar with its content.

    I understand how the series resistor in conjunction with RHOLD and capacitances creates the pole/zero pair. The benefit being the added low-pass response and noise filtering, but I am just not sure it doesn't introduce delays that could be causing the problem you are seeing.

    Please do let me know what you find.

    Regards, Thomas

    PA - Linear Applications Engineering