N-channel or P-channel mosfet needed?

Dave E

Solar panels and battery chargers seem to be a big topic lately.

I have a basic mosfet question:
I have a charger that uses an N-channel mosfet (only mosfet I could get locally) to control the power from the solar panels to the batteries.
Any time there is not enough sun light to give me full or near full power output from the panels I noticed that there is always about 2.1 volts difference between the panel voltage and battery voltage. When the output is near full or the batteries are at full charge voltage then the BS2 backs off on the gate voltage and so the difference between Vsolar and Vbattery increases anyway. I assume that this is due to the fact that an N-channel mosfet must have a Vgs greater than Vds to turn on. Would I be able to get that last little bit of power from the panel if I use a P-channel where Vgs needs to be less than Vds. In other words, will I be able to turn the mosfet on harder so that I will be able to get the power from the last few volts?

My circuit is very basic. The panel output is connected to the drain pin, the battery positive terminal is connected to the source pin. The gate is controlled with an analog voltage so I can regulate the current once the batteries are at full charge voltage. At this point, the BS2 / mosfet does a great job and can regulate the battery voltage to one or two hundredths of a volt.

This may not be the best way to use a mosfet but I eventually will use direct PWM and MPPT but when I made this circuit for proving the BS2 control concept, I noticed that I could not get the mosfet to turn on hard enough to get the solar voltage any closer to the battery voltage.
I know why this is now, Vgs must be greater than Vds for N-channel,·but it buggs me enough to investigate it further even though this will not be my final design.

Am I on the right track or did I miss the train completely on my mosfet theory?·

Dave

·

Phil Pilgrim (PhiPi)

A complete schematic would be helpful here...

-Phil

william chan

Is Mosfet switching the Positive supply or the Negative supply.
My n-channel mosfet is switching the negative supply and it works well with an optocoupler.
What Mosfet model are you using?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.fd.com.my
www.mercedes.com.my

Beau Schwabe

Dave E,

As William asks, "What Mosfet model are you using"? The Rds "ON" of the MOSFET could be playing a big role here, in which case it wouldn't matter much if you were using an NMOS or PMOS. Finding a MOSFET with a lower Rds "ON" might solve your problem... that or you can parallel multiple MOSFETs to lower the effective Rds "ON".

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Dave E

Thanks for the input. I am using an IRLZ34N to switch the positive side. If anyone knows of a better model, let me know.

I am working on a full schematic and hope to post it this evening.

Dave

Beau Schwabe

Dave E,

Here are a couple of other options with lower Rds "ON" values


I'd get the TO220AB version

http://www.irf.com/product-info/datasheets/data/irlz44zpbf.pdf
Rds = 13.5 mOhms
Ids = 51A
Power Dissapation = 80Watts

http://www.irf.com/product-info/datasheets/data/irl3705zpbf.pdf
Rds = 8.0 mOhms
Ids = 75A
Power Dissapation = 130 Watts

http://www.irf.com/product-info/datasheets/data/irl2505pbf.pdf
Rds = 8.0 mOhms
Ids = 104A
Power Dissapation = 200 Watts

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

william chan

Beau,

Seems like there are so many types of mosfets and most are about the same price even though some have much lower "on resistance".
Wouldn't irl3705zpbf obsolete the irlz44zpbf?

In your opinion,

1. Can the small leads of the T0-220 take 86 amps continuously?
2. Can the irl3705zpbf be used to start a car engine directly?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.fd.com.my
www.mercedes.com.my

Beau Schwabe

william chan,

"Seems like there are so many types of mosfets and most are about the same price even though some have much lower "on resistance".
Wouldn't irl3705zpbf obsolete the irlz44zpbf?" - depends on the application and availability

1) This rating would be package limited as indicated in the PDF... the D2 package version has pin widths of about 2.28 millimeters where as the TO220 is about 1.4 millimeters
2) You need a heat sink of some kind if the package temperature goes outside of the chart specs. There is a chart showing current verses case temperature that indicates the package limitation.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 6/6/2008 1:44:28 AM GMT

Phil Pilgrim (PhiPi)

There's usually an inverse relationship between R_DS(on) and gate capacitance. Since the latter affects switching speed and/or needed drive current at high switching speeds, it introduces a tradeoff that doesn't necessarily obviate devices with higher R_DS(on). As with most electronic design work, you have to consider the tradeoffs, as there is no universally optimum device for all apps.

-Phil

Dave E

Thanks.
I will get a couple of them on order in the morning and give them a try.

Is there some reason to stay with Nmos? Wouldn't Pmos be easier to use?
It seems it would be easier to generate a gate voltage < drain voltage for a Pmos rather than generating a gate voltage > drain voltage for an Nmos.

Dave

william chan

n-channel is cheaper and can take higher currents.

That's why you need to switch/block the negative supply.
Use an optocoupler.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.fd.com.my
www.mercedes.com.my

Dave E

William, you're right about the current handling capacity of the Nmos. Some that I saw were rated in the hundreds of amps.

The Pmos transistors I ordered came in and I gave them a try. Looks like they will do the job. Now when the panel output is low I get less than 0.1 volt across the transistor where before the minimum I could acheive was just over 2 volts with the Nmos I was using.

Beau, I will get a couple of the Nmos transistors you suggested and give them a try.

Thanks for the input.
Dave