mosfet drivers 4422 in an H bridge

amps.http://ww1.microchip.com/downloads/en/DeviceDoc/21420d.pdf

check the output resistance.... that is the problem with these devices for non capacitive loads.....

I've used them (4427) for exactly that application, but I was only driving LEGO motors.

Thanks for feedback, you are right, 1.4 ohms per channel is not good, so many better mosfets out there. I still want to use the

I cant find any mosfet H bridges using 2142 devices circuit diagram on net.

Here is a norml transistor driven mosfet h driver

Do I just replace the transistors and connect output of 2142 to each mosfet?

thnaks

That's NOT a "normal" h-bridge driver. In fact it's an example of how *not* to do it properly. Consider, the use of driver outputs to drive the out-of-phase drivers can be a disaster. That's because the driver (and I use that word advisedly), with its 10k pullup resistors against the high IRF9Z30 gate capacitance (900pF), slows the gate-drive signal (t = RC = 10us) enough to cause cross conduction. That's when the p-channel MOSFET conducts through an n-channel MOSFET from the supply rail to ground. Ouch! We're talking about 20A or more during the up-to-10us supply-short time. That's not my definition of normal. :-)

That isn't necessarily a problem. The MOSFET turn off is slow but that doesn't give cross conduction unless you turn on the other half of the H too soon.

technically true, but what sort of PWM frequency can you get with ~ 50us of dead time?

on the bright side, that makes it slow enough for the 1N400x diodes.

besides, the requirement for a LOT of deadtime is *crucial* information, and just aint there.

Cheers Terry

The OP didn't say he needed PWM, The bridge is OK for simple on/off forward/reverse motor control providing whatever drives it gives sufficient dead time.

and thats about all. except they dont bother mentioning the dead time anywhere. that must make it fun for novices.

Cheers Terry

Thanks for all the feedback, Im now a even more confused novice.

Basically I have a 4 amp motor for a robot . Will have a pic giving it control logic and PWM. Was going to build this

but upgraded the driver to TO-220 (5pin) 4422 but now concerned the

you probably can find a suitable H-bridge IC for your motor.

if you DIY with FETs you need to ensure a reasonable amount of DEAD TIME between the drive pulses. the original circuit you posted a link to was really, really bad, in that it required dead times on the order of

Cheers Terry

I have to ask one more dumb question, to avoid issues with dead time etc with mosfet H bridges, could I not just parallel the 4422 mosfet drivers effectively giving me, 4amps cont,

the answer is a definitive maybe. within the chip they will have some way of dealing with shoot-thru (eg with deadtime). if you try and parallel these, you would want individual chips to switch at exactly the same time, to avoid shoot-thru between adjacent devices. that may well be a nsaty trick.

the real problem is you are playing with a little bit of power, so the Rdson is fairly low, so shoot-thru is bad. if OTOH you want to parallel individual HC14 gates, go right ahead, it works real well. I havent tried paralleling different chips though, but ultimately they have much higher Rdson, so the problem is a lot less severe.

I suspect you are going to have to do it properly, if you want it to work reliably. OTBS its a great learning experience. just buy plenty of FETs, and use a current limited supply until you are convinced it works. no scope = no hope.

Cheers Terry

This circuit uses a single power supply (+9 to +18 volts). With such a setup you can use one kind of MOSFET, probably N-channel to connect your supply to the motor.

If you use dual (+ and -) supplies, you have to use complementary pairs of MOSFETS (N- and P-channel) to conduct each supply to your load (N for the +, P for the -).

Just something I think no one has pointed out yet...

"John E." schrieb im Newsbeitrag news: snipped-for-privacy@news.sf.sbcglobal.net... | >

Thanks, interesting circuit and very well documented. But I work about one million times higher in frequency. Obviousely it seems that you overlooked that.

cheers - Henry