OVP circuit issues

[rakeshm55]

Hi ,
I raised an issue of over voltage protection in one of my previous post.
https://www.eevblog.com/forum/projects-designs-and-technical-stuff/input-overvoltage-protection/


Based on the ideas I have designed one of my own. Please refer to the schematics attached for the circuit .
5Vin is the input supply and OVP o/p is D5VIN. Now this D5VIN power signal also takes power from a li ion battery through a diode (Not shown in the circuit)  .

I used 2 Pchannel MOSFET to prevent the reverse current flow from D5VIN to 5Vin rail when 5Vin is unconnected and battery is connected. I believe this design is flawed as when D5VIn rail is connected to battery through a diode it consumes a current of 45uA approx and the p-Mosfets are turned ON even though Q5 is off since Q7 is conducting and Q6 parasitic diode in action.

How to correct this issue?? I can avoid one P channel and attach one mode diode to D5VIN line making a proper diode Power Oring b/w Battery and D5VIN. Will this help?? How can I make it better

[TerminalJack505]

You might try doing something like the attached.  It's based on my best guess as to what's causing the problem.  Which is, Q7's gate is being pulled to GND when 5VIN is absent.

[rakeshm55]

Hi,
I would like to clarify that Q7 is not pulled to ground . Q5 is still OFF. When a voltage of 4.2V on D5VIN the 5VIN raised approximately to 2.4V similarly the gate potential of Q7 and Q6 rises to 2.4V thus Q7 is turned ON. When I try to short 5VIN rail to ground there is a current flow 200ma (CC of source connected to 5Vin was 200mA) pointing to turn ON Of Q7.

[TerminalJack505]

I kind of figured Q5 was off and Q6 and Q7 where getting pulled down by the ~150k of resistance between their gates and ground.

[rakeshm55]

I think Q6 is OFF as the VGS is 0V, but conducts through the parasitic diode. I was getting a voltage drop of 0.45V across the Q6 when 5VIN is unconnected.

Can I do OVP with just one P channel Mosfet rather than using 2 as the purpose is not solved??. any suggestions welcome.

By the way 150k aspect I couldnt catch up ....please help

[Niklas]

When 5VIN is disconnected, the gates of Q6 and Q7 are connected to GND via R41-R8//R13-R14 (100k+100k//100k+15.4k = 165.4k). Vgs for Q7 is then larger than its threshold voltage and will turn on Q7. Q6 will only act as a diode in this case.

If you are adding a diode, break up the link between R16 and R41-C10-Q6S and place the anode on 5VIN. Perhaps also add a pullup resistor between G and S on Q7 to prevent any floating gates when 5VIN is disconnected.

[rakeshm55]

Hi,
 Actually the circuit has latching issues too. When the device is Switches are ON and when in this state if I plug of the Power input (5VIN) the switches remain ON as the 5VIN rail will raise to battery voltage value (Say 4.2V) this is sufficient to trigger the U7 and latch the switches in turn ON position.

mmm.... Is there a much simpler way to implement this over voltage protection....

[Niklas]

Have you added a pullup resistor between G and S on Q7?

[rakeshm55]

Have you added a pullup resistor between G and S on Q7?

I have tried connecting a 15.4K across G and S of Q7  no diode . This prevented the reverse current flow. When shorting 5VIN to ground the current flow is reduced to 33uA showing current path through R41. Since I have not put a diode in the path as Niklas suggested ,when I short 5VIN to D5VIN with 5VIN unconnected to plug the circuit will latch to Q6 and Q7 ON state.

[rakeshm55]

Hi,
For the OVP can I use a low side switch this will prevent me from using P channel MOSFETs??....

Q5 could used for connecting the Circuit ground to 5V adaptor ground. Q5 has a low current carrying capability so I used N channels in parallel.

Will this be good solution or will there be issues?? which one is better??

[rakeshm55]

Also I am using a boost converter in my design.... Positioning a switch at the low side will it impact the switching characteristics of DC to DC converter ??