LM2853: How to handle the enable pin

Hi Varun,

We have verified the operation based on your answers.

1.  When the voltage between the EN pin and GND was measured with an oscilloscope, it was about 2.7V, which is below the EN HIGH input threshold of 3.75V (75% of 5V).

2.  When probing as in 1. after powering on, it outputs 3.3V, but when probing before powering on and then powering on, it only outputs about 1.7V.

Based on 1 and 2, it seems that the enable voltage has dropped due to the input impedance of the oscilloscope probe.
During normal use (no oscilloscope probe connected, PMIC open-drain input to EN pin, no external pull-up), is it possible that the EN pin voltage will drop due to noise or other factors, causing the power supply output to stop?

If the EN pin is internally pulled up, is there a risk that the EN pin voltage will drop due to external factors (e.g., noise or input impedance) if an external pull-up is not used in an open-drain configuration?
Since the EN voltage actually drops when measured with an oscilloscope probe, we are concerned that the voltage may drop further due to input impedance or noise, causing the EN pin to no longer correctly detect a HIGH level.

Thanks,

Conor

Hi Conor,

You are right, the input impedance of the scope acts like a resistor divider with the internal pull up resistor on the EN pin. What is the input impedance of the probe you used? A 10:1 probe should have an input impedance of 10MOhm. 

The impedance of the PMIC open drain FET when off should be >10MOhm for reliably enabling the LM2853. Usually there is a deglitch filter on the input pins to not trigger from noise spikes. But if you want more safety margin, adding an external pull up might be the right option.

If the LM2853 can be always enabled when VIN is high, you can even externally tie EN to VIN. 

Is this for a new or existing project. I recommend using our newer parts for new projects.

Best regards,

Varun

Hi Varun,

Varun John said:

The impedance of the PMIC open drain FET when off should be >10MOhm for reliably enabling the LM2853. Usually there is a deglitch filter on the input pins to not trigger from noise spikes. But if you want more safety margin, adding an external pull up might be the right option.

The PMIC connected to the EN terminal of the LM2853 is the ISPPAC-POWR1014A-01TN48I from Lattice. Below is the datasheet.

ispPAC-POWR1014ADataSheet.pdf

The OUT11 terminal of the ISPPAC-POWR1014A-01TN48I is connected to the EN terminal of the LM2853. As far as I can see from the PMIC datasheet, there is no impedance information for the open-drain output of the OUTx pins, including the OUT11 terminal. If you calculate the apparent impedance in the same way, taking into account the leakage current and output voltage, would it be as follows?

For 5V output:
・When the leakage current is 35µA: Approximately 142.9kΩ
・When the leakage current is 60µA: Approximately 83.3kΩ

From your answer, the impedance is less than 10M ohms, so a pull-up resistor is not necessary. Are these ways of thinking, including the above method of calculating the impedance, correct? I would appreciate any advice.

Thanks,

Conor

Hi Conor,

I think the way you calculated the open drain impedance should be correct. However I would use 10V for the calculation as that is the datasheet pull up voltage. Resistance = 10V/35uA = 285KOhm (typ). Min= 166KOhm. I think these values are quite low compared to the EN pull up of the LM2853.

The EN pull up current of the LM2853 is 1.5uA when pulled up to 5V. This would correspond to a pull up resistor of 3.3MOhm (5/1.5uA). 

The EN pull up resistor and the open drain output of the PMIC makes up a voltage divider network. To get a high (>3.5V) on the EN pin of the LM2853, the open drain resistance of the PMIC output will have to be > 10MOhm. 

I think in this case, it may be best to add an external pull up using a smaller (<50K) pull up resistor. 

Best regards,

Varun