Post Go back to editing

adm1087

Hi,

I’m using ADM1087 and in general it works as expected.  However, the timing seems to shift dramatically from time to time and becomes somewhat unpredictable.

As attached, the circuit is to generate a 60-second delay.  What we are seeing is that it ranges from 40~65 seconds all in room temperature.  What makes it strange is the timing typically decreases from longer delay (65 sec the first operation) to shorter delay (40 sec after few operations).  The capacitors used are 10% tolerance (P/N shown below).

GMK316AB7106KL-TR

GRM188C71E225KE11D

I understand the cap’s 10% tolerance can affect the initial timing, but can’t figure out why it changes over time.  Does the circuit require discharging the timing capacitor?

Thanks.

  • Hello,

    It seems like the capacitors on the CEXT pin are not getting discharged completed, as a result you are seeing the variation in delay timing after the first power up. How long do you wait between subsequent power cycle?

    To verify this is the issue, after the first power up could you completely discharge the capacitor before powering up and check the time delay.

    Regards,

    Opal

  • Hi,

    We suspected the same thing and have confirmed that it was the case.  Once we bleed off the charge, the timing is more consistent.

    However, I thought the internal MOSFET is supposed to bleed off the charge.  Is that not true?

    Thanks.

    Shang

  • By the way, the capacitor discharges fairly quickly until ~0.27V and stays there for minutes.

  • Hi Shang,

    Yes the internal mosfet will discharge the delay capacitor if the Vcc to ADM1087 is still preset.

    If you refer to the capture below, where Pink is the Vcc, and yellow is the CEXT. You can see that the 10uF cap gets discharged in ~1mS.

    If you refer the the capture below, where the Vcc to ADM1087 is removed, and you can see it takes multiple sec for the delay capacitor to discharged completely after the initial discharge. 

    Regards,

    Opal

  • Humh...I think I understand what you said.

    The following is how we operate the ADM1087.  VCC is present the entire time few minutes after the timer is up.

    Each time we exercise that, t_EN gets shorter and shorter.

    If the MOSFET were to discharge the capacitor, it should happen at t=60 right?  I would imagine it discharges the capacitor immediately after the timer is up (t=60).  VCC is still present even till minutes later.

    Does require ENIN to change the state in order activate the discharge?

    Thanks.

    Shang

  • No, the capacitor does not discharge after t = 60, it only discharges when ENIN goes high (active low) or Vin drops below 0.6V. In your test when t = ~3 min you're removing Vcc while ENIN is low as a result the capacitor is not getting completely discharged, and you're seeing the variation in time delay next time you're powering up. 

    As a test, if you drive ENIN high before powering down, the capacitor should be fully discharged and you should see the expected time delay (~60s) next time you're powering up.  

    Regards,

    Opal

  • OK.  I understand now.

    That makes sense.

    Thanks for your help.

    Shang

  • This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

    Thank you,
    EZ Admin