GPIO Input level with external Mega Ohm resistor pull up

[iharyadi]

Hi Support,

I am observing something that is quite strange on the GPIO input.  I am developing a Zigbee device with Energy harvester with NRF52840.   I need to use the state of the harvester using a GPIO as input.  The current consumption in most cases is good.  

I do have hard time using the GPIO with external pullup.  The internal pullup resistor are in the ~tens of KOhms.   It is not suitable for low energy usage.  I am replacing with external pullup resistor.  I uses 3.3M Ohm to test.  I open pin 1.13 (It does not really matter which pin.  I am including here for completeness only) as input and NOPULL.  I measure the high voltage with multi meter.   The value of the voltage at high is approximately 0.95 volt.  VDD is 3.3V.  The pull-up resistor is pulled to 3.3V source that is quite powerful.  There is no issue the voltage dip.  I uses the PPK to measure the current.  The current consumption is low (which is good).

Here is the first strange thing.  I read the Electrical specification.  The input voltage for high level should 0.7*vdd.  In reality,  when I read the pin,  I got the correct result.  The 0.95v is read as 1.  This is a good result.  But,  I am worried that this is not within specification.  Is this something to worry?  Or,  it just happens to work correctly by some unknown reason.  Is the noise of the high resistor is very high such that my multimeter does not read the peak?  I understand that there will be more noise.  But,  is the noise that high that it is enough to hit the 0.7*vdd threshold?  

I also worry that the resistance of the pin is reasonably for NOPULL is not quite high.   It is probably somewhere around 900K to 1M ohm.  That is why the pull only up to 0.95v.   The pin is driven by open drain pin.  That is why I need to use the pullup.  If you have suggestion to overcome this,  please let me know.  I am not reading a fast signal.  It is more like a long term status of whether the harvester is working or not.  I consider making a push-pull buffer.  It feel like an overkill solution for the problem.  Do you have suggestion on how to overcome this issue?

Thanks in advance for the help.

br

Iman

[ketiljo]

The value of the pull up depends on the source. The input resistance is very high, >10 Mohm. So if the resistor isn't able to pull the signal high enough, there is something external that pulls it down. You asy it's open drain. Are you sure it's really open or is there something that drives the transistor that pulls dowh the signal?

If you measure only 0.95 V on the pin, it can read as either 0 or 1 as the level is not defined. You need to be below 0.3 V or above 0.7 * Vdd. 

[iharyadi]

Thank you for the quick response.  I understand.