Power MOSFET Gate Driver

[NCEE]

I am using a gate driver as an interface between a power FET and a microcontroller. The power FET controls a coil that operates at 48V. The gate driver is powered by 12 V with input voltages that are TTL compatible. Any ideas why the gate drivers could be getting hot and exploding? They are decoupled using capacitors as mentioned in the data sheet, the only difference is that I'm using electrolytic capacitors instead of a film and ceramic capacitor.

 

[itsmoked]

Which gate driver?? Hooked to what FET??

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[NCEE]

Gate Driver: MCP1406
FET: IRF1407

 

[itsmoked]

I'm not seeing an obvious reason based on these parts alone.

Got a schematic snippet?

Have you looked at the Driver's output pin with a scope to see if something is coming back at the driver?

Have you done the the device power calculation which includes switching losses, etc?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[IRstuff]

Could it be that the driver is only rated for 18 V Vdd and would probably seriously dislike being run at 48 V?

http://ww1.microchip.com/downloads/en/DeviceDoc/22019a.pdf

TTFN

FAQ731-376

 

[ScottyUK]

Film and ceramic caps have good response to pulses. Electrolytics don't. Why not use the recommended types and see if the situation improves? The guys who write the application notes usually know what they are talking about.



----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Marke]

There are a number of things that you must be careful of:

1) You must use adequate decoupling which provides a very low impedance at very high frequencies. This is due to the very fast switching of the FET and drive circuit. There can be a lot of energy to decouple in a very short period of time. Small Ceramic capacitors provided a low impedance path at high frequencies. Film capacitors (physically small) provide a low impedance path at medium frequencies. Electrolytic capacitors provide a low impedance path at low frequencies. I would suggest that you add 100nF ceramic decoupling capacitors as close to the supply pins as possible.

2) The printed circuit layout is vitally important. It is essential that the drive circuit is not influenced by the load current. That means that you need to be aware of the current paths and how they affect the inputs. Ground plane layouts are a solution if implemented correctly.

3) Heatsinking. These drivers need to be cooled. If you do not apply sufficient heatsinking, they will over heat. The drivers often rely on large copper areas on the pcb to act as heatsinks.

4) Limit the operating frequency and see if the temperature drops.

It is possible that the drive circuitry and FET are oscillating due to poor decoupling and circuit layout and are effectively operating at a very high frequency and this is causing your problem. Check the FET Drain with a scope.

Best regards,


Mark Empson

http://www.lmphotonics.com

 

[NCEE]

When I set up the gate drivers and the power fet on the bench, I don't have problems. I can run it all day long and not heat up. The problem occurs when the FET is operating with the coil.
Heres the overall setup:
4 batteries in series provide 48 volts. The 48 volts powers the coil. The low side of the coil is connected to the drain of the power FET so the FET controls when current is allowed to pass through the coil. 12 V is taken from the first battery to power the gate drivers.

The gate of the FET should be insulated from any effects of the coil on the source and drain, correct?

 

[NCEE]

Another note:

An exact replica of this application is currently in place using MOSFETs as gate drivers. The drain of the FET is connected to 12V via a 10K resistor. The gate of the power FET receives its signal from the drain of the previous FET. Same coil, same voltages. These FETs do not seem to have any problems. They do not get hot while the system is running and they have lasted a very long time. It just not the best way for the system to run because the logic has to be inverted from the microcontroller.

Why does the gate driver have problems but not the MOSFET?
Is it because when the gate of the power FET needs to discharge, there is a clear path to ground through the MOSFET?

 

[Marke]

What is the voltage rating of the FET and what is the transient voltage on the drain when the FET switches OFF?
What overvoltage protection and snubbing protection have you provided for the solenoid?

When the solenoid opens, there will be a very high voltage with a steep wavefront developed across the solenoid to try to prevent it from opening. This can result in a large voltage transient on the drain of the FET and this can be capacitively coupled through into the gate of the FET. The driver has to keep the gate low enough to prevent a turn ON at this point. If the design is not correct, there can be oscillations at this point with the gate driver having to absorb a lot of energy. A scope on the drain of the FET may give a clue.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[NCEE]

There is a diode in parallel with the coil to deal with the voltage transients after the coil opens. I've had a scope on the drain of the FET, but I didn't notice any large transients.

 

[itsmoked]

I too suspect that your lack of proper decoupling and perhaps board layout is causing the Driver to be 'rung' which greatly multiplies its power dissipation.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com