MOSFET driver - Unused low side, more questions

In the IR2011 mosfet driver, you can use it to drive two mosfets. I'm using an NPN mosfet. I'm not sure what to do with the low side though, I won't be using it, so I think that given there is no supply and return, I can leave these not connected? I plan on just using the high side of the driver, with my NPN mosfet.

My question spawns from the possibility of electrical noise on the LIN pin, with a high HIN pin, and the expected behaviour of these. Should I tie LIN to ground?

From the mosfet driver to the gate of the mosfet, if I put the 10 ohm resistor in, I should see reduced harmonics, but is it really needed in an automotive application? What happens without it, in slow frequency applications, and high frequency applications?

As the IR2011 is noted as having a buffer on the high side, is this the same as a capacitor, or are capacitors needed, would you use capacitors if the switching frequency was low (say 1 - 20 Hz)?

Another Q: If I want a mechanical relay, without the mechanical operation failure possibility of them, a MOSFET is my only choice? (I'm guessing it is, but I've been wrong..)

I'm using an NPN mosfet.

I think you mean N-channel MOSFET.

Don't leave any inputs floating. Connect them High or Low as needed.

Do you have a link to IR2011 and the MOSFET it is driving?

  1. What are you trying to do? Is switching the high side the right way to go? Usually you want to switch the low side when using MOSFETs, all else being equal (you also probably wouldn't need the MOSFET driver in this case, barring things we don't know about the use case - which is one of the advantages to switching the low side).

  2. Never let pins float unless you have a damned good reason for it (like the datasheet telling you to).

  3. You cannot get a mechanical switch without the mechanical operation, otherwise it's a solid state switch, rather than a mechanical one, by definition ;-) . For DC applications, a MOSFET is the best analog to a relay; You can also use an NPN transistor, but those often don't perform as well as switches. For AC applications, you use SSRs or triacs.

You can't use a high-low MOSFET driver like this just on the high side, you drive two MOSFETs in a half-H-bridge configuration and you have to PWM it (you cannot hold the top switch on for more than a few ms since the bootstrap capacitor will discharge).

The lower switch is relied on to regularly pull the top MOSFET source to 0V which allows the bootstrap diode to conduct and refresh the bootstrap cap.

The IR2011 datasheet is here: http://www.irf.com/product-info/datasheets/data/ir2011.pdf The IRF1405 I am driving is here: http://www.redrok.com/MOSFET_IRF1405_55V_169A_5.3mO_Vth4.0_TO-220.pdf

The idea was to use the IR2011 as a gate driver to get the gate to charge fast. I was hoping to simply have: Arduino to IR2011 HIN pin, then 12VDC to the supply of the IR2011 pin, and the HO pin to the gate of the IRF1405, and the load after there.

I don't want the fan running backwards, which is what I thought, the low side was for..

I'm only interested in using the IRF1405 as a switch, on / off every second /half second at worst I suppose, so is IR2011 misplaced in the design?

use a IXDN404PI
this is a dual channel driver (easy to use)
you don’t need to use both channels

the ir2011 seems to be for half bridge setups or dc-dc converters

You can only do low-side switching with an nFET and no bootstrapping. If you want high-side DC switching you need either a pFET or a driver that has a charge pump to provide the gate supply.

The half-H-bridge drivers like the IR2011 rely on a continuous PWM drive to allow diode-capacitor bootstrapping of the high side gate drive. This is normal for the usual load (motors, inverters).

MarkT: You can only do low-side switching with an nFET and no bootstrapping. If you want high-side DC switching you need either a pFET or a driver that has a charge pump to provide the gate supply.

The half-H-bridge drivers like the IR2011 rely on a continuous PWM drive to allow diode-capacitor bootstrapping of the high side gate drive. This is normal for the usual load (motors, inverters).

Lost here.

I am doing low side switching, if I have 12V to load, and then controlling the path to ground with the mosfet, right? So, can't I drive the gate with a transistor, such that base is connected to arduino pin, collector to 12V, and emitter to the gate of the mosfet, with a resistor to ground, to discharge the gate?

With the IRF1405, I am doing low side switching by using it as an N-Channel mosfet?

screwpilot: use a IXDN404PI this is a dual channel driver (easy to use) you don't need to use both channels

the ir2011 seems to be for half bridge setups or dc-dc converters

http://www.irf.com/product-info/datasheets/data/ir25600.pdf This looks like much the same as the IXDN404PI?

Non inverting would give me the result I need too I assume (off = at the Gate of the mosfet).

tocpcs: Lost here.

I am doing low side switching

Then why on earth are you electing to only use the high-side driver in the IR2011??

discard the IR2011

i'm not getting if you want to make low side or not

tocpcs: http://www.irf.com/product-info/datasheets/data/ir25600.pdf This looks like much the same as the IXDN404PI?

Non inverting would give me the result I need too I assume (off = at the Gate of the mosfet).

yep to both

discard completely the IR2011

Contemplating just driving it with a transistor, I tested this with an ordinary 12VDC toggle switch (I'm not expecting to do switching less than 250 or 500ms).

It worked a treat for a long while, I had it running continuously to test for junction temperature getting too hot from passing large amounts of current. I had a quick test of 10 "Off / On" switch cycles in the space of a few seconds, no noticeable heating, i.e. definitely not hot to the touch.

After running it for a bit more after this, I flipped the switch off and left it off for around 1 hour.

Leaving it off, and just driving, I noticed smoke from the container I had it in. No cable shorts found. Just smoke.

I clipped the drain wire and tapped it up. After getting home, I took a look at the damage - the nylon screw holding the heat sink on had melted, leaving the heat sink free in the enclosure. The solder joints from the terminals to the mosfet are fine. There's no short between any of the terminals.

The circuit was fairly simple: 12VDC to load, from load to drain of mosfet, source of mosfet to ground, and the gate to a 12 VDC switch.

What was the cause of smoke failure?

If I were to guess, it would be the wind through the radiator driving the fan free wheel, and it was this that caused the smoke coming out.. although the mosfet has it's own body diode, and it was fine for a long time before the smoke came out.

Any other thoughts?

You had a nylon screw holding a heatsink? :astonished:

Yeh, it was fine for a long while, with the gate switched off... I'm able to change the screw, but I want to know the failure cause, so measures can be put in place..

What kind of switch? SPST SPDT? How did you wire it?

A simple SPST switch, they were soldered at the terminals of the switch, about 30mm space between each side of the terminals. Definitely in the off position.

At the mosfet side, there were a 5mm pitch 3 input terminal, each of the mosfet legs were bent to the terminal and soldered to it, cabling was 2.5mm2 screwed to the terminal.

Hm. No answers.

If you had no resistor to pull the gate low, then the gate may only slowly lose charge when the switch is turned off. This would leave the MOSFET in linear mode, low enough resistance to allow a lot of current, but high enough to drop enough voltage to cause it to heat up a lot.

Never use nylon screws in a heat sink.

polymorph: Hm. No answers.

If you had no resistor to pull the gate low, then the gate may only slowly lose charge when the switch is turned off. This would leave the MOSFET in linear mode, low enough resistance to allow a lot of current, but high enough to drop enough voltage to cause it to heat up a lot.

Never use nylon screws in a heat sink.

Good point, that's interesting though, as when the switch was off, the motor turned off, which isn't what you'd expect if there was a charge held in the gate.

Admittedly, the 'test' was not close enough to a replica of what would be in circuit. But my quick (rushed) test was an attempt to prove the constant current wouldn't heat up the heat sink beyond acceptable (and that seemed to be fine).

I'll do a proper check when I get the PCBs from fab.

polymorph:
Never use nylon screws in a heat sink.

No problem with nylon screws in a liquid-cooled set-up though :slight_smile: