Can't get P-Channel Mosfet to turn OFF!!!

DavidM

HI,

I am having trouble trying to get a FDN5618P LOGIC LEVEL P-CHANNEL MOSFET to work . ( I have blown half a dozen of these buggers so far!)

I need to switch a small load via 48V DC on the high side via a prop output pin.

The load on the DRAIN Side will only be a few hundred milliamps in my final circuit so the current requirement is LOW, (but I do need these volts i.e 48v),

For my test circuit I am just using an LED via a 10K half watt resistor to show me that the mosfet is ON or OFF for my load.

My trouble is that I cannot get the MOSFET to turn OFF?

When I connect my test circuit with the switch OFF, the mosfet is already on ( i believe ) because the LED is ON.

By default I would want the MOSFET to be ON ( no current flow i.e OFF), So I understand that the GATE need to be pulled high.

I am "Simulating" the 3.3V Propellor output with a 3.3V source ( so I don't stuff up a real prop chip!!!)

What I don't understand is the DATASHEET say's that the GATE THRESHOLD VOLTAGE is between -1 and -3V.
What is the "CORRECT" voltage required to turn on the mosfet?

It is supposed to be a LOGIC LEVEL MOSFET!

Should the 10K pullup resistor (R1) just go to 3.3V for the default high state?

I have a feeling I have blown the gate of my mosfet?

I do have an oscilloscope to measure.

Can anyone help?

Thanks

Dave M

lanternfish

Maybe this app note can help

Andrey Demenev

1. This schematic would not work. A P-channel MOSFET requires negative value on gate relative to source to turn on. When you connect this to propeller, output voltage will be in 0 .. 3 volts range, translating to -48 .. -45 volts on MOSFET's gate relative to source - the FET will be always on.
2. Connecting this to to Propeller can damage it. Although in normail conditions, protection diode on Propeller's input comes to play, if Propeller's power is removed, it will receive 48 volts through the resistors.

You would better use the attached schematic. The N-channel MOSFET removes high voltage from Propeller's pin, and resistor divier R1/R2 limits gate-to-source voltage of P-channel MOSFET at 12 volts, as 48 may be too much,

As for LED being always on, make sure you have not interchanged source and drain. If you wire it incorrectly, current would flow through MOSFET's body diode.

Leon

You could use an LTC1693-5 driver. It's intended for use with P-channel high-side MOSFETs.

tdlivings

In your test circuit with a physical switch being open there is only the 10K pulling the gate up to
48v so the Mosfet should be off. So as Andrey indicated maybe you have the drain and source
reversed.
Only in your test circuit would the Mosfet be off as connecting the end of the 1K to 0 or 3.3V will still have
the Mosfet turned on.

Tom

tonyp12

>I am "Simulating" the 3.3V Propellor output with a 3.3V source ( so I don't stuff up a real prop chip!!!)
.
You are leaving the line floating when you not pressing the button, so more of an an open-drain circuit.
But using prop pin to do the sinking of 48v down to 0v (or down to 3.3v same result) is not a good idea.


So you would need a n channel mosfet to sink the p-mosfets gate line to ground when you want it on.
And this nchannel mosfets gate should be pulled down to ground as prop boots up as input pin. (like Andrey show)

Phil Pilgrim (PhiPi)

The schematic with the switch should work if the MOSFET has been wired correctly. With the switch open, the gate will get pulled to the source voltage. The circuit will not work when connected directly to a Propeller pin -- even when the Prop program simulates an open-drain output by toggling DIRA. The reason is that there will still be a conduction path through the pin protection diode to 3.3V. Moreover, with the resistor values shown, there will be enough current conducted via this pathway to damage the Propeller chip.

Use an NPN or nMOSFET to pull the pMOSFET's gate down, and you should have no problems.

-Phil

tonyp12

Here is a high side power switch (300mA Continuous)
I wonder if you could use that with the prop directly when it comes to its enable pin for a 48v load?
its internal resistor R3 is 2.4k and the external should be another 1k min.


http://www.mouser.com/ProductDetail/ON-Semiconductor/NCP3712ASNT1G/?qs=sGAEpiMZZMuCmTIBzycWfIV6CddTkDy92A4nzu%252bxejU%3d
I like the larger .95mm pin spacing on it, low 40cent price and bonus is the spike protection.


.

whicker

When dealing with industrial type devices, use what the industry uses: high side switches, like tonyp12 said.

These devices are value for the money given their functionality: very low on resistance, high current, charge pump for logic level input, and thermal shutdown. The better ones even offer current limiting and notification of overcurrent (high side power switch).

I am biased toward the PROFET series (because of the abuse I've witnessed they can handle), but many other companies make similar types.

DavidM

Hi Everyone, thanks heaps for all the replies.

I tried out an N-CHANNEL Mosfet (logic level) and I could not get it to function, so I need to have another look, I don't think i tried hard enough.

There are some other good suggestions here as well.

RE: HIGH SIDE SWITCH, I am actually using INFINEON BTS452R ( 24 of them) on my board and I have IPS2041L for low side, both of these are smart switches and are pretty tough!

But they are rather large, for the area I need to have this particular high side switch in.

I am going to evaluate these two devices ITS4141N and ITS4040N, maybe a bit hard to get in a hurry, but infineon australia is very helpful), but they are a little BIG for my board but probably easier to use.

I have not been able to find a simple SOT-23 High side switch that can do high volt's and low current ( remember I only need a few hundred milliamps.)

I also tried to see if there is a DARLINGTON that is a combo of NPN ( for the BASE) and PNP for the Collector/emitter, but I have not found one.

part of my problem is that I have some prototype PCB's MADE ( AND THE SOLDER STENCILS!!!) and this switch problem needs to be a mod so i can get the rest of my board working, my original problem as far as the board is concerned, is that I originally got the drain and source mixed up, so I have been carefully soldering on parts ( and blowing stuff up)
I will implement changes in the next revision ( so this will cost me more time and money) . but I have to make do with the PCB I already have ( see attached) the high side switch in question is Q1.

apart from that my 48v boost converter circuit works really well, and my reverse polarity dual mosfet (T11) works exceptionally well. Only the parts on the left hand side of the board have been hand soldered, so I can test out the basic functions, I am getting a few boards machine soldered for further testing, once I get this damn Q1 solved



thanks

Dave M

Sapieha

Hi DavidM.


Nice work on Layout

Mark_T

Phil Pilgrim (PhiPi) wrote: »

The schematic with the switch should work if the MOSFET has been wired correctly. With the switch open, the gate will get pulled to the source voltage. The circuit will not work when connected directly to a Propeller pin -- even when the Prop program simulates an open-drain output by toggling DIRA. The reason is that there will still be a conduction path through the pin protection diode to 3.3V. Moreover, with the resistor values shown, there will be enough current conducted via this pathway to damage the Propeller chip.

Use an NPN or nMOSFET to pull the pMOSFET's gate down, and you should have no problems.

-Phil

No, what will happen with that schematic is that the MOSFET gate oxide will probably punch-through since the 40+V between gate and source is above the abs max rating. Then once the gate oxide has broken down the 1k resistor will have 48V across it and overheat (unless it can take 2.3 watts).

For high-side switching a level-shifting circuit of some sort is required. A simple NPN or n-channel transistor could be used, or a dedicated driver chip. The circuit must ensure the P-channel gate isn't pulled more than about 15V below the source (typical Vgs ratings are +/-20V)

Phil Pilgrim (PhiPi)

Mark_T wrote:

No, what will happen with that schematic is that the MOSFET gate oxide will probably punch-through since the 40+V between gate and source is above the abs max rating.

Right. 'Hadn't considered the V_GS(breakdown) issue. A larger-than-1K series resistor would address this issue in the original schematic; but, of course, that's peripheral to the issue of switching it with a Propeller.

-Phil

Duane C. Johnson

Try this:

Directly driven from a 3.3V Prop pin.
Hi in causes Q1, acting as a 2mA current source, drops 10V across R2 to drive the P-Chanel MOSFET gate.

It's fairly safe for the Prop. Even a direct short from 48V to the collector of Q1 causes no damage. I suggest the use of MPSA42 or MPSA06 NPN transistors. These are 200V and 60V types. This transistor is not very critical and most any that has at least a 50V Collector to emitter rating will do.
2N3904 won't cut it here!

R3 is not really needed but does add a bit of safety

Duane