PDA

View Full Version : H bridge driver for 2 Pchan and 2 Nchan mosfets



Zap-o
09-20-2011, 03:12 AM
I have searched for quite some time looking for an H bridge driver that will drive 2 P-chan mosfets (high side) and 2 N-chan mosfets (low side). Anyone heard of such a driver? Most I find drive 4 Nchan mosfets for high and low side.

Thanks in advance

Phil Pilgrim (PhiPi)
09-20-2011, 03:47 AM
Why are four nMOSFETs a problem? That arrangement has advantages.

-Phil

Zap-o
09-20-2011, 04:44 AM
Why are four nMOSFETs a problem? That arrangement has advantages.

-Phil

http://www.datasheetcatalog.org/datasheet/vishay/70611.pdf

Phil Pilgrim (PhiPi)
09-20-2011, 05:23 AM
With the ready availability of high-side charge-pump drivers for nMOSFETs, I'm not sure that Vishay's points carry much water any more. At the very least, such drivers reduce your MOSFET SKUs by half.

-Phil

Ttailspin
09-20-2011, 04:28 PM
I know just enough to be dangerous,
I was just wondering if the TIP's in this circuit could be replaced with Mosfets? or is that completely ignorant?
85247
Just wondering,
like I say, I know just enough to be dangerous..:innocent:

-Tommy

Zap-o
09-20-2011, 04:44 PM
Phil - Thanks for the words of advice, but I am not using the H-bridge like most people tend to use them. I need to use P-chan on the high side.

Ttailspin - That schematic would almost work with mosfets in fact it would work but would require a few adjustments.

All together I suppose there is no driver or at the least its not very common to do as I am intending on doing - thus the shortage of driver.

Phil Pilgrim (PhiPi)
09-20-2011, 04:51 PM
Can you explain what's different about your approach that precludes using n-channel devices in the top half?

-Phil

alex123
09-20-2011, 04:56 PM
I agree with Phil. It'd be better to use 2 n-MOSFET's. A pair of NMOS and PMOS are rarely used nowadays because the p-channel is approximately three times weaker that n-channel with the same physical dimensions hence it's hard to match both transistors' parameters.

Here is the Fairchild MOSFET driver:
http://www.fairchildsemi.com/ds/FA/FAN3268.pdf

Alex

alex123
09-20-2011, 05:11 PM
Tommy,

The optocoupler solution would be ok for very low switching speeds. When you switch the MOSFET gate you're switching a large capacitor. The higher the speed the more power you need. That would require a large transistor in the optocoupler. Plus the push-pull instead of one transistor and a resistor to better control the on/off conditions hence to avoid the state when both MOSFETs are on at the same time.

Zap-o
09-20-2011, 05:45 PM
Phil - I suppose the reasons are as follows:

My circuit needs to be cheap (currently I am driving the h-bridge with Comparators)
10 Amps at 15 Volts running through a 10 foot cable means NO PWM.

Perhaps this afternoon ill post a schematic

alex123
09-20-2011, 06:10 PM
If it's 15V it may be better to use this IC instead: http://www.fairchildsemi.com/ds/FA/FAN3278.pdf

Why 10A/15V means no PWM?

Zap-o
09-20-2011, 06:19 PM
Why 10A/15V means no PWM?

Well PWM 'ing 10amps through a 10Foot cable will radiate some noise. You ever hear a power line running, of course that is a hell of a lot more power, but its the same idea. I also have very sensitive analog signals running in that same cable so I want to keep noise minimal.

Not to mention my PCB is small and I cant afford to heat up the ICs on that board. So I use a Mosfet located off the PCB to control the current in an analog fashion (no PWM). The H-bridge just sets the direction of current.

Capt. Quirk
09-20-2011, 06:31 PM
For 10 to 20v These drivers and object from the OBEX will do it. (http://obex.parallax.com/objects/527/) The n-mosfets for the high side is just a normal option.

For 5 to 15v look at the IR2301 as a dip, or the IRS2301 as a SO-8

Or possibly a LMH6644 and save 3 pins on the micro

alex123
09-20-2011, 06:54 PM
I heard the power lines running but they are 100kV (or more) hence the buzzing if that's what you're talking about.
Obviously you'll have coupling to your other wires due to the mutual capacitance and inductance in your cable. It's not the voltage or current that cause the coupling but the edges when FETs switch. You could de-Q it a little by a series power resistor...

If I understand it correctly you want to control these MOSFETs in the linear region? Do you want to control the current and retain the voltage level or the other way around?

Zap-o
09-20-2011, 09:09 PM
alex123

No I want to use my h-bridge just simply to set the direction of current. So that the mosfets in the h-bridge will be either fully on or fully off. This keeps the heat off the PCB. The current is controlled by another mosfet located off the PCB and yea its used in the linear region. Have a look at the crude drawing.

85252

alex123
09-20-2011, 11:39 PM
Gatcha. If switching the current direction is slow I'd suggest a simple DPDT relay. No need to drive the bridge FETs at all this way.
If not the Fairchild part seems good but I'd check the FETs' datasheets to see if it's strong enough to drive the gates. I'm not sure if the NMOS high side driver Capt Quirk suggested can be used to drive the PMOS though. It would definitely have to be connected differently than the datasheets suggest for the high side NMOS.