Switching a DD P7 using a MOSFET?

[mudman cj]

Hi everybody,

I know we have some brilliant people here on CPF, and at least a few people can help me with this one, or at least tell me that I need to wear a different hat :tinfoil:.

I want to make a direct drive P7 light in a small host using an Emoli cell, so I only have a relatively low voltage to drive the MOSFET. This host cannot hold a large switch in the tailcap, so I plan to mount a MOSFET on the LED heatsink. I am concerned about this requirement because I have read that (some? most? all?) MOSFETs need at least 4.5V to go into full saturation.

Of course, I only require a continuous current drain of around 4 amps, so I don't need one of the big MOSFETs used for 100W hotwires. In fact, I don't know if I could cram a TO-220 package into the light.

I would like suggestions about which MOSFET to choose for this application (if any) along with specific suggestions for the other components that need to be used with it and a description of component connections (so I don't screw it up :green. Looking through past threads, I came upon a component recommended by Nereus for a different application that seems like it's in the ballpark: IRLI3705. How would this one work?

Thanks in advance!
C.J.

 

[koala]

try trenchfet if not then mosfet gate drivers.

These are all quite low and should handle the P7 drive current
http://www.vishay.com/mosfets/15-rated-on-res/

 

[Nereus]

You can find the irli3705 specifications here. Even at 4,0 volt gate-source voltage, it has 0.018 ohms drain-to-source resistance. I think one li-ion cell is enough to push the mosfet to almost full saturation. You have to check that out: connect the source to li-ion battery minus and gate to battery plus and measure the drain-to-source resistance with a DMM. I suggest that you use almost empty li-ion battery because you want to mimic the under load situation.

WBR,

- N

 

[mudman cj]

Does it matter whether I use an N-channel or P-channel? If P-channel works, then I am looking at SI3499DV.

@ Nereus - how did you get the resistance as a function of gate-source voltage?

 

[Nereus]

Does it matter whether I use an N-channel or P-channel? If P-channel works, then I am looking at SI3499DV.

@ Nereus - how did you get the resistance as a function of gate-source voltage?

From the "Electrical characteristics" table on page 2.

- N

 

[mudman cj]

Thanks! With your help, I have found the si4378dy, which has 0.0034 Ohms D-S resistance at 2.5V G-S. It is one of the new trenchfet parts mentioned by koala. I just wish it was a DIP package instead of SO-8. :sick2:

 

[koala]

Yeah that's one of the problem as well. I am looking for SOT-23 trenchfet package. It's not just vishay that does them a couple other manufacturer like Fairchild has them too.

Trenchfet is awesomely low voltage vs RDSon. Circuit builders please note.

 

[mudman cj]

I agree. That would be preferable. Now I am considering the PMV31XN. It can handle 5.9A @25C and 3.75A @100C. It has a chance of working if I mount it to the LED heatsink. They are available in quantities of 1 from Mouser.