I am facing issues with the synchronous buck convert IR2104 Mosfet driver

[charan_6]

I am building a MPPT Solar charge controller.I am facing issues with the synchronous buck convert where we used IR2104 Mostet Driver and IR540 Mosfets in the charger controller. When I connect battery to the circuit with in fraction of seconds the the low side mosfet is getting blown and making battery short circuit. Can anyone help me in solving the issue. If you want I am ready to share circuit here.

Thanks in advance.

 

[Diver300]

Please share the circuit. It makes it much easier for anyone to picture the circuit if they aren't familiar with the MOSFET driver you are using.

 

[charan_6]

Hi here is the circuit attached below for the above issue raised... In this Q3 mosfet is getting heated up and blown off while battery is connected to the circuit and and there is a issue with when i connect solar to the circuit then esp32 is been getting paused Or switching off and there is a intrupt to the serial monitor in the arduino after i disconnect solar and battery again it is getting to normal state... I am unable to know why this error happened with the circuit.

Mosfer used: N Channel Mosfets
Controller: Esp32 Wroom
Battery Connected: 12.8volts/30ah
Solar Connected: 18volts.

Thanks in advance.

 

[alec_t]

Per the datasheets, the IRF540 can have a total gate capacitance of ~70nC and needs a Vgs of 10V for its minimum 0.077 Ohm Rdson. The IR2104 can source only 130mA, which means it will take about 70nC*10V/0.13A = ~5.4 microseconds to charge the gate capacitance up to 10V to turn the FET on fully. However, the dead time provided by the IR2104 is only ~500nS, so it is likely there will be considerable current shoot-through between Q2 and Q3.
I think you need a gate driver with a much higher current rating and/or a MOSFET with a much lower gate charge. The IRF540 also has a rather high Rdson by modern standards, implying a large heatsink is needed for significant power handling.

 

[rjenkinsgb]

I'd suggest an IR2110 instead of the 2104; that has a drive current of +/- 2A, so much faster switching with high gate capacitance MOSFETs.

Also, the gate resistors should be as low as possible without overloading anything - again, to minimise switching time.

 

[rjenkinsgb]

Re. the dropout, the AMS1117 data suggests a minimum 22uF tantalum on the output, for stability. You only have 4.7uF?

If you are using electrolytics for bulk decoupling, they need to be considerably larger, plus ceramics close to the MCU.

 

[danadak]

Calculating Gate R's, attached.

 

[charan_6]

Per the datasheets, the IRF540 can have a total gate capacitance of ~70nC and needs a Vgs of 10V for its minimum 0.077 Ohm Rdson. The IR2104 can source only 130mA, which means it will take about 70nC*10V/0.13A = ~5.4 microseconds to charge the gate capacitance up to 10V to turn the FET on fully. However, the dead time provided by the IR2104 is only ~500nS, so it is likely there will be considerable current shoot-through between Q2 and Q3.
I think you need a gate driver with a much higher current rating and/or a MOSFET with a much lower gate charge. The IRF540 also has a rather high Rdson by modern standards, implying a large heatsink is needed for significant power handling.

alec_t thanku for that info I will try and update you soon... what I am going to reach for my next move to solve the issue.

Thank you for fast reply.

 

[ChrisP58]

I am building a MPPT Solar charge controller.I am facing issues with the synchronous buck convert where we used IR2104 Mostet Driver and IR540 Mosfets in the charger controller. When I connect battery to the circuit with in fraction of seconds the the low side mosfet is getting blown and making battery short circuit. Can anyone help me in solving the issue. If you want I am ready to share circuit here.

Thanks in advance.

The timing diagram for that driver has the LO output high whenever the IN pin is low. And, while that's mostly OK during the charge operation, that is what's killing Q3. Because the LO output is

You must make sure the the LO output of the driver is never high except for the short freewheel period following the High side on time. I'm not sure that you can reliably do that timing with that mosfet driver. I would choose a driver with separate Hi and Lo enables.

Making a battery charger with a synchronous rectifier is always tricky.

 

[aliarifat794]

High-voltage spikes due to inductive kickback from the inductor can also damage the MOSFETs. A snubber circuit across the MOSFETs can help absorb these spikes.

 

[Flyback]

Hi,
Whenever you use synch rects, always first get it working with the synchronous fet disabled...then add it in when you have that working.
As above, beware of shoot through. Also spurious turn on of the lower fet when the top one turns on......get round this by using lower fet with bigger gate capacitance...but not so big that you cant turn it off quick enough. (QGD should be smaller than Qgs for the lower fet)
Also, beware inductor current reversal in light load....if necc, use a load detector and disable the sync fets when the load goes below 1/3rd max...or whatever load makes the inductor curent go discontinuous.