Using P-Channel Mosfet as Switch

[nura100]

Hello Forum,

I am building a 24Volts Lead acid battery charger (28V output, 6 Amps) and have a micro which senses the battery full charge voltage and disconnects the charging current, right now i am using a relay but want to use P-channel mosfet like IRF9540 for controlling the charging voltage/current. i cannot use N-channel mosfet. and i am not using any PWM. only simple on/off

i need help as to how to turn on the P-channel mosfet without exceeding the Vgs limit, i know a 10V zener can be used but not sure how to do it for a P-ch mosfet.

Regards,

Arun

 

[MikeMl]

Hello Forum,

I am building a 24Volts Lead acid battery charger (28V output, 6 Amps) and have a micro which senses the battery full charge voltage and disconnects the charging current, right now i am using a relay but want to use P-channel mosfet like IRF9540 for controlling the charging voltage/current. i cannot use N-channel mosfet. and i am not using any PWM. only simple on/off

i need help as to how to turn on the P-channel mosfet without exceeding the Vgs limit, i know a 10V zener can be used but not sure how to do it for a P-ch mosfet.

Regards,

Arun

Here is a PFet switch which can be controlled by a ucontroller port pin. Gate voltage is limited to ~12V

 

[Hero999]

Is this a fast charger?

I'd recommend charging to more than 28V

Some lead acids can be charged to 29V or more depending on the type of battery. I recommend looking at the datasheet.

 

[nura100]

Thank you Mike ML, Once again !

you saved me last time too ( in my high side current sense post)

O.K the peak charger output is 30.2V, 6Amps max. In your diagram i think you have used a Voltage divider ( 10K pull-up and 20L pull down Resistors), is this the best method to use or using a 12V zener would be safe. will the PFet be turned on fully to handle 6 amps current by using such large value resistors ? ( i know that mosfets are voltage driven devices unlike BJTs which are current driven)

also my micro is a poor sourcer of current (can just supply 1mA per port pin), but can sink 20mA, is there any way i could use a PNP transistor like BC327 to control the Pfet.

about the heat dissipation, IRF9540 has a ON-RDS of 0.2 Ohm, and when it carries 6 Amps then the power wasted/dissipated would be 7.2W (I2 x R), will it make more sense to use a Relay or a Pfet ?


Regards,

Arun

 

[MikeMl]

The voltage divider on the PFET gate in the previous diagram applies a Vgs of -9V, which should turn it on well. Changing R3 to 15K makes the Vgs = -11V. A Zener wouldn't add anything.

When the port pin is at 5V, the current being sourced is only 420uA, so that has a good margin. I cant see how you would use a PNP.

I'm not familiar with a IRF9540, nor do I have a Spice model for one. I think you need a better PFET than the IRF9540. It should have a Vds of -50V or so, and an ON resistance of < 100mΩ. If you cant find one, a relay might work.

I built a charger for 14V lead-acid battery in my RV. When the battery is charged, I just lowered the charger voltage from 14.8V to 13.8V (float voltage) so I dont have to switch the high current output of the charger. Can you do that?

 

[nura100]

thanks Mike for your replies,

the only Pfet i can get here is IRF9540 which has VDS of -100V and rated at 19 amps at 25*C. yes it has a very high ON RDS of 0.2 Ohm bcoz of which it dissipates 7.2 watts when carrying 6 amps (using relay seems to make sense)


i have been using Zeners to limit VGS of N-Fets and not for Pfets, but i have seen posts from other forums where they have used Zener to limit the VGS of Pfets, i am attaching them, may be you can comment on these ;


regards,

Arun

 

[MikeMl]

Using the Zener, you can get a little faster Turn-On of the PFET, but the Turn-Off is just as slow as the purely resistive network I showed. Since your battery charger is turning ON/OFF only once in a while, then the speed of the PFET switching is not critical like it would be if were being used in an HBridge or SMPS. My resistive network wastes a lot less power than the Zener network.

On lowering the On resistance of the PFET, why not parallel two of them. A low value resistor in series with their respective sources should do it.

 

[mneary]

Putting two MOSFETs in parallel really keeps them a lot cooler. Not only does it cut the voltage drop in half, but the resulting (reduced) power is shared between two devices. So 7.2W becomes 3.6W, which is 1.8W per device. Since the MOSFETs have a positive temperature coefficient, a source resistor isn't necessary. (If one MOSFET begins to get warmer, its resistance increases and the current shifts to the other.) They stay balanced, just the opposite of BJT's.

BTW, 1.8W in each MOSFET will be pretty hot; three in parallel would be more sensible. Or a relay, of course.

 

[eblc1388]

Just use the IRF5210 with 60mΩ Rds ON and that's a maximum.

At 6 amp current, it wastes only 2W or less as heat.