Working H-bridge

[Carl-NC]

Finally, I got an H-bridge to work. Thanks to Dave, for rekindling my interest.

In the past I had tried both NMOS & PMOS top-side devices, but the flyback pulse caused all sorts of havoc. The solution is to put diodes in series with them. But you also have to give them a bit of DC current, so that as the flyback settles, it does so to a consistent level. In this case, a diode below ground. Without the DC current, the diode is in a high-impedance state during decay, and the settled level moves around a lot.

Funny thing is, this circuit still does not work if you put PMOS devices on the top-side. The disabled-side PMOS completely kills the flyback. I.e., if Q1 is "on" and Q4 is being pulsed, then the flyback works as long as Q2 is physicaly removed from the circuit. But when it's inserted, even with the gate tied high, the flyback is completely shut down.

I'll shoot a couple of o-scope pics & post them.

--Carl

[Dave Emery]

Re: Working H-bridge

Carl,

Good going and a very nice circuit. I am wondering if I posted the correct circuit that I used as it was some time ago since I made it. The circuit I ended up with worked fine. The main trick was of course to use a charge pump to provide the required voltage to operate NFET's as both the low side and the the high side transistors. There are a number of H bridge driver chips on the market that we may also want to take a look at.

[Chris]

Re: Working H-bridge

Hi Carl,

a reason of your flyback pulse problem could be the built in protection diode implementet in the MosFet design (across Source & Drain).

Chris

[Carl-NC]

Re: Working H-bridge

I did look at a couple of h-bridge chips, but none were designed for use with high-voltage flyback pulses. I think they are mostly used for motor control, with sub-100v breakdowns.

--Carl

[Carl-NC]

Re: Working H-bridge

>a reason of your flyback pulse problem could be the built in protection diode implementet in the MosFet design (across Source & Drain).

Absolutely, and is also the reason I could not figure out why the damned thing wouldn't work when I built it some months ago. Dave's schematic -- specifically his use of a MOS symbol that explicitely shows the internal diode -- prompted me to re-build it, and add the series diodes.

As an IC designer, I use MOSFETs all the time, but they never have protection diodes, and work just as well when drain & source are reversed. I had forgotten all about the internal diode in discrete MOS devices.

--Carl

[KT315]

ferrite beads

Carl, as you are the designer... is an usage of ferrite beads on the GATE given something of effectiveness?

regards

[Carl-NC]

Re: ferrite beads

>Carl, as you are the designer... is an usage of ferrite beads on the GATE given something of effectiveness?

I expect that would slow down the edges, which you don't want to do.

- Carl