Can a dual Lo side mosfet driver; drive (2) N Mosfets in a half bridge configuration?

Capt. Quirk

I recently purchased some Fairchild dual low mosfet drivers. At the time I thought Lo meant n-mosfet. I have looked at the specs for similar Hi side drivers,
and the specs doesn't make any sense to me despite a illustration showing two n mosfets.

Any help would be appreciated

Thanks

Bill M.

Phil Pilgrim (PhiPi)

No, a half-bridge requires one low-side and one high-side transistor. A high-side MOSFET driver is distinguished by being able to boost the gate drive above its supply voltage, so that it can drive nMOSFETs to source, rather than sink, current. That way, nMOSFETs can be used in the high side of an H-bridge, instead of resorting to pMOSFET transistors.

-Phil

Capt. Quirk

Thanks Phil,

I don't completely understand, but I will take what you said, and work it out.

Thanks

Bill Mazzacane

Phil Pilgrim (PhiPi)

Let me try again. To turn on an nMOSFET, the gate has to be raised a certain voltage above the source. In a low-side configuration, the source is grounded, so the gate voltage is always provided referred to ground potential, or zero. On the high side, when an nMOSFET is turned on, its source is connected to the load and is driving it at the load's supply voltage, say 12V, for example. To keep the MOSFET turned on, then, the gate must be substantially higher than 12V. That's where high-side MOSFET drivers come in. They include voltage booster circuitry to provide a gate drive signal that is higher than their supply voltage. Low-side MOSFET drivers do not provide this facility.

-Phil

PJAllen

http://forums.parallax.com/showthread.php?132721-H-Bridge-Development

Heater.

So why can't we just have a pMOSFET on the high side like we do in some MOSFET audio amplifier power stages?

PJAllen

Nobody said you can't.

GordonMcComb

To put a "git 'er done" spin on Phil's excellent reply, I've seen working -- as in robots moving around -- H-bridges use a simple system of an opto-isolator and extra 9 volt battery on the high side N-MOSFET. It's not pretty, and you have to remember to disconnect the battery (or build in a switch) but it works.

The idea is that you apply the gate signal through the opto-isolator, whose transistor is connected to the 9 volt battery and logic supply in series. It boosts the 5V gate voltage by another 9 volts, without a charge pump circuit. For $1 in parts, including the battery, it's a workable alternative.

As PJ sez in his other thread, designing and building bridge bridge circuits, at least those that are efficient and not prone to burning out, is no walk in the park.

-- Gordon

Phil Pilgrim (PhiPi)

Heater,

You can certainly put a pMOSFET in the high side. There's still the issue of gate drive, though, since the gate must then be at the load's supply potential to turn the pMOSFET off. How complicated this needs to be depends on how fast you need to turn the gate on and off.

-Phil

Capt. Quirk

In PJ Allen's blog, the schematic appears to be supplying "motor voltage" directly thru an opto-isolator to his logic gates?

Phil Pilgrim (PhiPi)

Yes, and that's also what Gordon was referring to. It works, assuming you don't need to switch the high side quickly. Typically, though, any PWMing, for example is done on the low side. You can also use nMOSFETs on the high side, like PJ has done, but with a special kind of optoisolator that uses photovoltaic technology instead of photoconductive methods. This allows you to to switch the high side nMOSFET without requiring a supply voltage higher than that required for the load.

-Phil

GordonMcComb

Capt. Quirk wrote: »

In PJ Allen's blog, the schematic appears to be supplying "motor voltage" directly thru an opto-isolator to his logic gates?

Not exactly. The optos are gating Q1/Q3, and the motor V+ is going through those. At least that's how I read his schematic.

As Phil sez, it's important to get the timings right here, or you stand to have all four transistors on at the same time for very brief periods of time. Timing is less of a problem if you're steering the logic to the four gates via a Prop or other microcontroller, where you can easily control when the gates are turned on.

And yeah, what Phil says about gating the low side for PWM, at least if you want to get the PWM period out of the low hearing range.

-- Gordon

PJAllen

...the schematic appears to be supplying "motor voltage" directly thru an opto-isolator to his logic gates?

+V (12V), a/k/a the "high-side", gets switched through Q1/Q3 (depending.)
Those transistors' gate voltage is switched/enabled through the optos (U1, U2 respectively.)

Q1 presents the motor with a +V and Q2 makes a current path for one direction.
Q3 presents the motor with a +V and Q4 makes a current path for the other direction.

PE -- A few days ago I updated the blog without mention in the subject here (in Projects.) I've been back working on the project that I'd started and interrupted to get on the H-Bridge track.
If you like, feel free to ask questions here or over at the h-bridge subject.

Capt. Quirk

This is an example of how I tend to make things more difficult than they need to be:

I have 2 computer monitors with multiple data sheets open on both screens---including all my notes
from previous questions, Photoshop open for dissecting data sheet graphs, and at least
3 data sheets and an app-note on my desk. After double checking my FDS6670a logic mosfets,
similar mosfets like the IRF3708, the low side mosfet driver, my new 1/2 H bridge mosfet driver, and
at least 2 other mosfet drivers, I came to this conclusion:

(1) If Battery voltage is greater than, my driver should work for a 1/2 bridge because it supplies
Battery Voltage + .3 vdc. to each FDS6670a mosfet.

(2) But for it to be reliable, I think, I would most likely need 24 to 30v batteries,supplying a 12volt r
egulator to supply my mosfet drivers so my speed control doesn't die after a 2v drop below RDS_on.

(3) I should just use a 555 and build a Lo side speed control, for my battery powered lawn mower.

Thanks for your help

Bill Mazzacane