Mosfet gets heated up at 5A or above.

Thread Starter

Vindhyachal Takniki

Joined Nov 3, 2014
487
1. I have designed a circuit using TI's bq24650. Schemtaic & gerbers are attached. (Part of both are removed, only I have shown is high current path & mosfet).
2. I have to operate circuit at 10A. Problem is Mosfet gets too hot at 5A or above even after 2 min of operation. Also Diode D1 PDS1040 gets heated up, but they are not very hot.
3. Actual circuit has only one mosfet but it was rated for lesser current. So I have used 2 in parallel in my circuit. What could be issue with the circuit, are track length too short. I have removed solder msak layer from high current path as in gerbers you can check. Also while soldering I have placed a thick solder on these exposed tracks. Mosfet has high current rating as you can check in mosfet datasheet.
 

Attachments

bertus

Joined Apr 5, 2008
20,317
Hello,

The schematic is barely readable, but where is the boost capacitor and diode (C5 and D2 in the schematic below)?

BQ24650_typical_application.png

Also I do not see any feedback.

Bertus
 

shortbus

Joined Sep 30, 2009
7,451
Have you investigated the actual 'thermal' standards for the mosfet package your using? The data sheet numbers for voltage and current are usually not able to be obtained in real use. I have my doubts that 10A or even 5A is doable in that mosfet package. Each form or package of a mosfet has a wattage limit that usually has nothing to do with the data sheet first page numbers.
 

bertus

Joined Apr 5, 2008
20,317
Hello,

Here are the copper top and copper bottom from the gerberfiles for the others to see:

coppertop.png

copperbottom.png

I am not an PCB maker, but the lines around the fets look quite long.
Bertus
 

ronv

Joined Nov 12, 2008
3,770
Well, I don't know about the FETs but the diode is not rated for that much current when mounted to a pcb.
 

ian field

Joined Oct 27, 2012
6,539
Datasheets don't tell the whole story..
Many assume a "perfect" heatsink in their "maximum" ratings..
Just because a Mosfet may say 150 Amps doesn't mean it won't burn up at 10 Amps if not properly cooled..
AFAIK: the usual term is; "infinite heatsink".

I've often wondered about the current handling capacity of the leads on PC motherboard MOSFETs with Id rating somewhere in the general direction of 80A.
 

Picbuster

Joined Dec 2, 2013
990
1. I have designed a circuit using TI's bq24650. Schemtaic & gerbers are attached. (Part of both are removed, only I have shown is high current path & mosfet).
2. I have to operate circuit at 10A. Problem is Mosfet gets too hot at 5A or above even after 2 min of operation. Also Diode D1 PDS1040 gets heated up, but they are not very hot.
3. Actual circuit has only one mosfet but it was rated for lesser current. So I have used 2 in parallel in my circuit. What could be issue with the circuit, are track length too short. I have removed solder msak layer from high current path as in gerbers you can check. Also while soldering I have placed a thick solder on these exposed tracks. Mosfet has high current rating as you can check in mosfet datasheet.
Is the fet in full saturation? use a scope and observe Gate voltage and its edge(rising and falling). Compare this with data sheet.
Two fets gates parallel will also put a double gate cap onto the controller chip.
 

ian field

Joined Oct 27, 2012
6,539
Is the fet in full saturation? use a scope and observe Gate voltage and its edge(rising and falling). Compare this with data sheet.
Two fets gates parallel will also put a double gate cap onto the controller chip.
One would hope that a made for the job chip is driving the MOSFETs correctly, but its a low voltage application and the chip might expect logic level devices.

If the switching frequency is set by an external capacitor; the frequency could be set too high - if the capacitor was accidentally omitted, it would time on parasitic capacitance and the frequency would be way out of spec.

If the frequency is too high, the MOSFETs will be transitioning through the linear region for relatively more of the cycle.
 
Top