Logic-level MOSFET suggestions

[MacGyverS2000]

I'm currently designing with the Fairchild FDS6614A (N-channel) in an SO-8 package, but I'm interested in any suggestions/favorites from everyone. This is for fairly low currents (<<1A/chip).

I came across a handful of logic-level FETs in my search (TI, Fairchild, Panasonic(?), among others), but didn't find nearly as many as I expected. Maybe I'm typing in the wrong thing into Google ("logic level FET").

I would love to find something in a smaller package (need to stuff 30+ of these on a small board), say an SOT-23? Power dissipation would be significantly less than 10mW per package.

 

[felixc]

Have you checked with International Rectifier? They have a nice lineup in the logic-level FET category.
TI has a line of driver arrays that are quite powerful.

 

[Comcokid]

Many MOSFETs don't necessarily advertise themselves as "logic level". You have to look at the data sheets to see the threshold specs.

I went to the Digikey site, and tried a few sorts. It appears Fairchild is good at listing their parts as logic level compatiable.

I don't really have a favorite. Last two projects I used a Fairchild Si4410DY (SO-8) just because it showed-up in the example circuit, and a BSS84 (ON Semi) for a small signal application because we had them in use from a previous project.

 

[MacGyverS2000]

Comco, you may have hit on something...I assumed that "logic level" was such a feature-worthy item, particularly in today's low voltage systems, companies would advertise the components as such. I'll spend some more time at the sites and use a parametric search a bit more...it may take me some time to narrow down exactly what specs I need to pay attention to.

 

[MacGyverS2000]

Funny enough, after roaming through page after page of specs, I have tentatively decided upon another Fairchild piece, the FDN339AN. It's an SOT-23 package that can handle 3A at 20V (more than enough). I'll continue my search as time permits, but if nothing else comes up, I think this piece will more than satisfy my requirements.

 

[nbucska]

You may need a pull-up R to +5 with some FETs.


<nbucska@pcperipherals DOT com> subj: eng-tips

 

[Comcokid]

Just a little bit of past experience-

I once had a logic MOSFET (don't remember the number, but about 50V, 3A) being driven from a Zilog Z8 micro (4 MHz clock) through a 1K resistor to the gate. Solid ground-plane, proper bypassing, etc. MOSFET was hooked to a inductive load with about 350 ohms of resistance to 12 Volts. Mosfet was being PWM controlled at about 150 Hz

Went to the test lab to do emissions, and had a 4 MHz picket fence up through 30 MHz. Removed the load from the MOSFET - still had the same picket fence. Changed the series gate resistor to 10K and the picket fence went away! the PWM had slight risetime then, but it was OK. The micro didn't like driving the capacitive gate of the MOSFET, even through a 1K resistor.

Just goes to show you that driving a MOSFET directly from a micro is never straight forward, though this was more of a micro problem.

 

[MacGyverS2000]

I haven't proto'd yet, but I'm starting with a 500 ohm gate resistor for a 1mA charge, and I'll probably raise it to around 5k if no problems arise. The input capacitance is around 7-800 pF, so intially the rise time is in the area of 0.5us (more than adequate for my purposes). I'll need to switch loads of a few hundred mA (each FET) on or off in the sub-10kHz range. No major inductance in the loads other than stray package and trace inductance. I have a nice DSO, but I've never had the opportunity to play with the FFT function