lower side mosfet keeps blowing

ScottWang

Joined Aug 23, 2012
7,397
The inductor in this buck-converter prevents a direct connection between two voltage sources at input and output (theoretically unlimited current) and allows current control with PWM.
The inductor and C17 became a LC filter to protecting the resistance load from surge or glitch, but for a battery it will almost double the voltage of battery, that will kill the Q3 quickly when Q3 is turn on.
 

Thread Starter

ak52

Joined Oct 15, 2014
230
Thanks guys for all your replies,I have a few more quires on selection of the LC filter.
Can someone confirm the mathematics of it please regarding the selection criteria of the LC filter

L = (Vin-Vout)*Vout / Vin * F * ΔI

C = L*Ilmax*Ilmax /{( (Vout + ΔVout-overshoot) * (Vout + ΔVout-overshoot)) - (Vout *Vout)}

Vin: converter’s input voltage (in Volts)
Vout: converter’s output voltage (in Volts)
F: converter’s switching frequency (in Hz)
ΔI: peak-to-peak inductor current ripple (in A)
ΔVout-overshoot:Maximum voltage overshoot allowed.
Ilmax:Maximum inductor current.
 

GentleRV

Joined Dec 18, 2010
13
I see you emulate a battery, if that is that is the case then set in serial with the coil a diode to prevent current draws back from the battery and blow up the lower FET
 

Thread Starter

ak52

Joined Oct 15, 2014
230
I see you emulate a battery, if that is that is the case then set in serial with the coil a diode to prevent current draws back from the battery and blow up the lower FET
I am using only the High side of the gate driver,so there would not be a Lower FET,as mentioned previously.
 

RamaD

Joined Dec 4, 2009
328
With a lower Mosfet, it is a synchronous buck converter. The low side mosfet, instead of the freewheeling diode, offers lower power dissipation, and hence better efficiency.
However, with a solar battery charger, many operating issues need to be taken care of.
  • When the panel voltage is lower than the battery voltage, (during the night for example, when there is no sun) the upper mosfet's body diode will conduct and the panel will short the battery.
  • During low charging currents, the inductor current will go negative and the lower mosfet will short the battery. This problem will not be there with a freewheeling diode.
If efficiency is not a concern, then a series diode from the output (after the capacitor) to the battery can be used. You need to use a freewheeling diode in place of the lower mosfet.
What is the Panel voltage, battery charging voltage and battery charging current?
 

ronv

Joined Nov 12, 2008
3,770
I think we are all assuming you have voltage and current feedback to control the pwm.
Maybe you could tell us the frequency and the maximum and minimum duty cycle.

PS Where did you find a 150 amp inductor?
 

Thread Starter

ak52

Joined Oct 15, 2014
230
  • When the panel voltage is lower than the battery voltage, (during the night for example, when there is no sun) the upper mosfet's body diode will conduct and the panel will short the battery.
When there is no sun,i would not start pulsing the FETs ,even if the FETS start pulsing there is already a series diode present to block the reverse voltage.
During low charging currents, the inductor current will go negative and the lower mosfet will short the battery. This problem will not be there with a freewheeling diode.
yes sir,i am going to use a fly-back snubber instead of the lower FET.
What is the Panel voltage, battery charging voltage and battery charging current?
Maximum solar voltage (no load voltage) is around 130v,battery voltage is 114v,battery charging current can vary from 0 to 30Amps.
I think we are all assuming you have voltage and current feedback to control the pwm.
Maybe you could tell us the frequency and the maximum and minimum duty cycle.

PS Where did you find a 150 amp inductor?
The frequency of operation is 10KHz,
Maximum and minimum duty cycle varies from 0 to 96% respectively.
I know a 150Amp inductor is a bit over kill ,but i have a guy who winds transformers for a living and he gave me a sample piece :)

About the LC filter, can someone confirm the mathematical formula i posted above.

Thanks and regards,
AK
 
Last edited:

RamaD

Joined Dec 4, 2009
328
As I said earlier, if efficiency is not a concern, then a series diode can be used.
The L calculations are right. As for as C is concerned, they should be rated for the ripple current, and the ESR of the capacitor will determine the pk-pk ripple voltage. This will normally override the calculations and the C will be larger than the calculated one. When charging a battery, this will correspond to ripple current.
Use a freewheeling diode in place of the lower FET, with cathode connected to the source of the upper MOSFET. This will conduct during the off-time. What is a flyback snubber here?
BTW, please check the panel characteristics whether it can supply maximum current at 114V.
 

Thread Starter

ak52

Joined Oct 15, 2014
230
I drew a rough schematic,please confirm if it is correct,Please ignore the LC values as i have to still calculate them.
The diode which i will be using is an STPS60150C which will have two diodes in parallel.
 

Attachments

RamaD

Joined Dec 4, 2009
328
The MOSFET should be N Channel.
A series diode to the battery for protection would be required. Else, the battery will be shorted to the panel when panel voltage is lower than the battery voltage, through the MOSFET's body diode, even if you do not turn on the MOSFET!
 

Thread Starter

ak52

Joined Oct 15, 2014
230
The MOSFET should be N Channel.
A series diode to the battery for protection would be required. Else, the battery will be shorted to the panel when panel voltage is lower than the battery voltage, through the MOSFET's body diode, even if you do not turn on the MOSFET!
I am using N channel MOSFETs only sir.(IRFP4768PbF)
The panels come with inbuilt diodes for protection.
As a precautionary measure, i already have a protection diode in series to the solar voltage
I have used the same diode STPS60150C for protection purpose.(It is the diode in Red color)
 

Attachments

RamaD

Joined Dec 4, 2009
328
It should work. This series diode also should be sufficiently heatsinked as at 30A, power dissipation on this could be over 25W.
 

Thread Starter

ak52

Joined Oct 15, 2014
230
Hey guys,
I am having another weird problem,when i connect a resistive load(i tried with a 7E,10W resistor),the mosfet triggers fine and i see the pulses in the scope,but when i connect a battery for charging i see no pulses in the scope,why dos'nt the mosfet trigger when i connect the battery? Any ideas...

Ps:the charging current is also zero,when i connect the battery,and the current is measured correctly when i connect the resistor.
 

Thread Starter

ak52

Joined Oct 15, 2014
230
Is the FET gate voltage at least 10V above the battery voltage?
My battery voltage would vary from 96v to 112v,(i have 8 ,12v,150AH tall tubular batteries in series). By gate voltage ,do you mean gate to source voltage? if yes then the maximum voltage is 20v as per data sheet.
 

Alec_t

Joined Sep 17, 2013
14,280
do you mean gate to source voltage?
Yes. So if the source is sitting at 96-112V the gate needs to pull up to at least 106-122V.

Edit: The IR2110 uses a bootstrap arrangement. That can't work unless you have a low-side FET to pull down the VS pin.
 
Last edited:
Top