• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 82 subscribers
  • Views 397 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TMS320F280049: Increased ADC leakage current

Urs Wild
Prodigy 10 points

Part Number: TMS320F280049

Hi,

We are observing an issue with ADC inputs when reading circuits with high impedance (~5kOhms).

The problem is high leakage into ADC pin which affects the voltage divider. Failure rate is in percent range (~20Pcs out of ~2000Pcs). At least pins C0/19 and A0/23 are affected.

Following circuit shows the ADC circuit. A voltage divider of 10k plus NTC is fed to ADC.

Following scope plot shows on Ch1 = VDDA=VDDIO and Ch2 = C0/pin 19 voltage.

Interestingly at startup the pin voltage is correct (half of 3V, since NTC=10 at 25°C), but after about 1s (can be anything between 0.2 to few seconds), the voltage drops like some breakdown occurs.

More interesting: After turn off supply and restart within few minutes (all voltages are gone), the pin is directly low at startup as shown here:

But after leaving the DSP unpowered for ~15 minutes, the behavior is again as on first scope plot.

ADC reading corresponds always to voltage applied to pin, so ADC itself is seems ok.

Doing this test with DSP reset XRSn kept low results in the same behavior, so FW can be excluded:

Do you see any potential cause that can lead to such a behavior? Can it be some protection circuit that is mistakenly activated?

It seems the parts affected have it from the start. Parts that are ok remain ok also later.

ESD may be excluded, since production is ESD handling compliant.

Any help here would be appreciated.

Best regards,

Urs

  • 0
    TI__Mastermind 30185 points

    Hello Urs,

    I'm not currently aware of any issue like this and it does not seem like this has been observed in the errata anywhere. As you stated that the ADC itself does not seem to be having an issue, this may be more related to board design.

  • 0 in reply to Omer Amir
    TI__Mastermind 45606 points

    Hi Urs,

    Can you remove R2128 (or whichever respective resistor) to disconnect the C2000 ADC from the NTC, then do your same probe test to see if the behavior remains? This would help to confirm if the issue is related to the C2000 device or not.

    Best,

    Kevin

    Edit: Also, what is the voltage before and after the drop?

  • 0 in reply to Kevin Allen18
    Prodigy 10 points

    Hi Kevin,

    After removing R2128 the ADC input simply is low, since also pull-up is disconnected (only filter cap C2134 present):

    But if adding a 100kOhm from 3.0V reference to ADC input, and also removing filter cap C2134, so only this 100k pull up is present, the situation looks similar as before, initially pin is high, then it's dropping somehow uncontrolled - this is still with RSTn kept low, so no FW running:

    The voltage before the dip is about 1.5V (10k + NTC=10k divider on 3V reference), afterwards the voltage is about. 0.3 .. 0.4V.

    Regards,

    Urs  

  • 0 in reply to Urs Wild
    Prodigy 10 points

    Pulling ADC input to reference voltage removes the leakage for certain time, the voltage divider is not loaded anymore (Ch2 at begin is ~0.2V, then I did short it to 3.0V ref for ~5s, and afterwards it shows correct ~1.5V):

    But some seconds to minutes later the voltage divider is loaded again, reflecting the high leakage current:

  • 0 in reply to Urs Wild
    TI__Mastermind 45606 points

    Hi Urs,

    Urs Wild said:
    After removing R2128 the ADC input simply is low, since also pull-up is disconnected (only filter cap C2134 present):

    I was actually curious about the behavior of the signal on the non-c2000 ADC facing side after removing the resistor. To see if the other circuits on the PCB are affecting the signal someway.

    What is the reference signal you are pulling up the signal with? Is this the ADC reference signal and if so how is it being generated (internal C2000 or external device)? I'm thinking you may be loading the reference signal too much.

    Urs Wild said:

    this is still with RSTn kept low, so no FW running:

    What if some simple firmware is running? Need to check state of ADC / pins during device reset.

    Best,

    Kevin