This question is related to my previous question about the battery charging and discharging. Here is the link to my previous question.

I have made a small change in my previous schematic, which is the PMOSFET. I have added NMOS in between the Charging IC and the Battery. Below is the schematic, Schematic

The following is the principle of the circuit,

  1. Hysteresis of 3.4V and 4V is set using a Comparator.
  2. Depending on the voltage level of battery, comparator gives the output. If the battery voltage is 3.5V then the output of the comparator is "HIGH"; this then turns "ON" the mosfet M2 (and Turns "OFF" M1) allowing the charging IC to charge the battery. When the batter voltage reaches to 4V, comparator output becomes low. This then turns "ON" the Mosfet M1 (and turns "OFF" M2); allowing the battery to discharge to the load.
  3. The load I am using is 0.25W 1.4K resistor but I have added 10 resistor in parallel. Which then makes the resistance of 140 ohm, so the current Iload=28mA.

Following is what I observe in the circuit.

  1. When the battery is charging and it is 3.7V the charging stops and starts to discharge. At this point I was also monitoring the hysteresis point and it showed 4V, but what I don't understand is why did it stopped at 3.7V?
  2. Also the discharging stopped at 3.48V instead of 3.4V. Also at this point i was monitoring the hysteresis point which was set to 3.4V.

Why is my hysteresis not working properly. I have used M1 as PMosfet so that battery gets connected to load when the output from the comparator is LOW and M2 as NMosfet to disconnect the charging process when the battery is connected to load.


2 Answers 2


Without RC filter caps on the comparator inputs, I might suggest you have 80mV of noise that triggered the low point of 3.4 or you have some input offset on the comparator.

When the battery is charging, again either you have noise or the threshold on (+) input is not 4.0V.

How is the comparator powered? And what is the output swing? And how much noise is on V- and V+ of the comparator. It is either set to wrong threshold or there is noise.

The other problem is the lack of a good bandgap reference voltage supply on comparator output swing and thus input reference voltage for (+) and Vboost +5.. All of these sources & noise + R ratios affect your thresholds.

  • \$\begingroup\$ @Klaus_what do you mean by RC filter caps on the input of the comparator? how can I implement this? I think it's because of the noise that my circuit is triggering faster then it should. Comparator is powered by the output of the boost converter (5V). \$\endgroup\$ May 30, 2014 at 7:04
  • \$\begingroup\$ Try 0.1uF on + input of comparator \$\endgroup\$
    – user43594
    May 30, 2014 at 11:02

I see at least three big problems.

1) You are attempting to provide charging from +5 through a diode, which will produce ~.7v drop. A charge current of 1 mA through the 500 ohm resistor will further reduce that by another 0.5 volts, so the greatest battery voltage the chip can produce is ~ 3.8 volts. Your charger is behaving heroically under the circumstances.

2) The gate voltage for M2 is (at most) 5 volts. If the battery voltage is (for instance) at 3.7 volts, you only have 1.3 volts gate to source. Is your MOSFET really rated to work for that low a gate voltage?

3) In using +5 as your reference voltage, you are at the mercy of any noise on it, and you cannot put a filter capacitor at the divider junction. A noise spike or dip on the +5 could trigger your comparator, and a meter would not see it. And since you have at least one switching regulator being driven from +5, and not a decoupling capacitor in sight, I'd pretty much guarantee you have noise.

You have not identified your comparator. Is it an open-collector output? What are its input offset and bias currents?

  • \$\begingroup\$ @WhatRoughBeast_Thanks for your comment but I don't really agree with what you mention in point 1 and 2.My charging Ic is providing a constant voltage of 4.2V in order to charge the battery and the max charging current,Icharge=(Vin-3.4V)/500=3mA Mosftet is always having either 5V or 0V depending on the battery voltage status as its powered by the boost converter which is 5V. I agree with what you mentioned in point 3, I see some noise when the hysteresis voltage is set, so that can cause the problem. I will try using a decoupling cap at the input to see if it helps.Its LTC6702 that i am using. \$\endgroup\$ May 30, 2014 at 6:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.