Author Topic: Amazing PMOS (Read 9837 times)

CraigHB · « **on:** February 26, 2013, 05:12:17 AM »

So I was looking around for a better P-channel MOSFET to use for my main power switch. The one I'm using now is acceptable, but it's a bit lossy. So I came across this one that's amazing;

http://www.vishay.com/docs/69063/si7137dp.pdf

The R_DS(on) and V_GS(th) are incredibly low, lowest I've ever seen for a P-channel. I put in an order for some samples and I'm going to try one. They're a drop-in replacement for the SOIC-8 packages I'm currently using.

The gate charge is really high so they can't be used as a switch in a converter, but for a main power switch they're perfect. I can probably pick up a percent or two in efficiency with it. Plus the main power switch is my biggest source of heat and this will greatly reduce it.

This is a leadless package, but a small PCB can be easily made to mount the part and connect wires to it.

utak3r · « **Reply #1 on:** February 26, 2013, 07:05:36 AM »

Damn, with those params you surely will go down with your heat... Threshold is also pretty amazing - 0.5-1.4V??

Now, that's the find

Breaktru · « **Reply #2 on:** February 26, 2013, 08:52:16 AM »

Nice find Craig

Very low indeed. I didn't think you can get your PowerBlock to be any more efficient but you did.

Just remember what Craig said guys. "can't be used as a switch in a converter". So don't attempt using it unless you are building you're own boost circuit.

utak3r · « **Reply #3 on:** February 26, 2013, 09:02:40 AM »

He meant it can't be a switching key, but it can be a power switch

CraigHB · « **Reply #4 on:** February 26, 2013, 04:37:47 PM »

Yeah, too slow to be used to switch at any kind of frequency, but unless you plan to switch your mod off and on a couple thousand times a second, it's plenty fast enough for a power switch. Could also be used for a switch in a PWM circuit if you keep the frequency down. The switch would have less resistance than the wires running to it.

BTW, my mods are getting about 93% efficiency at an output of 15W so I can use another percent or two. At that power level, I'm losing about a third of a Watt with my current power switch. That's not to be confused with the other two switches used by the converter. I have a strict limit on gate charge with those, but I found some pretty excellent ones there as well. The on resistance is higher, but that's the tradeoff with MOSFETs, you can have very high speed or very low on resistance, not both.

CraigHB · « **Reply #5 on:** March 07, 2013, 01:18:30 AM »

So, I got my order of these. Amazing parts. I'm measuring 2 mOhms on the bench with a 3.5V gate voltage, amazing. Here it is installed in one of my test boards;

Breaktru · « **Reply #6 on:** March 07, 2013, 08:30:00 AM »

Quote from: CraigHB on March 07, 2013, 01:18:30 AM

So, I got my order of these. Amazing parts. I'm measuring 2 mOhms on the bench with a 3.5V gate voltage, amazing. Here it is installed in one of my test boards;

At 2mOhms, that is amazingly low. Nice going.

vaperkarma · « **Reply #7 on:** March 13, 2013, 06:51:55 PM »

Nice work. This mod will be a beast

CraigHB · « **Reply #8 on:** March 13, 2013, 09:30:52 PM »

Right now I'm assembling a mod with a couple LG 18650 D1 cells running in a parallel pack. That one will be a beast, but I'll still have my little 1400 boxes. I'm upgrading those with the new switch after I finish the big one. I want my big box now but it takes me a while to build one. I'm about halfway through assembling the PCB at this point.