Quick question on FETs

We use IRLML2502s for general purpose stuff, from driving LEDs to relays.

On a sunny day (Fri, 10 Jul 2009 05:58:18 -0700 (PDT)) it happened Andrew wrote in :

IRLC34N will work from 3 V up, no resistors or diodes needed,

30 A, 55 V, .035 Ohm on, TO220, Vgs on max 2 V (min 1 V), PD 68 W.

On a sunny day (Fri, 10 Jul 2009 13:40:21 GMT) it happened Jan Panteltje wrote in :

Make that IRLZ34N

Yikes! A buck a shot?

A typical N-channel logic FET in a TO-220 package capable of doing 10A of switching would be a IRL520. It will cost more ( a buck or two ) than a TIP122.

More likely, a driver chip with multiple FET or bipolar switches driven by a microprocessor bus would be used in most applications.

Darlington drivers are still extremely commonplace... I don't think they'll stop making the ULN2003 anytime soon. It sounds like this class of part would meet most of your needs without making the jump to FET technology.

Tim.

On a sunny day (Fri, 10 Jul 2009 06:56:48 -0700 (PDT)) it happened " snipped-for-privacy@gmail.com" wrote in :

Oh well, for small series or one of the cots of the components is neglible to the other costs. Sure, if it is for large series, then a bit of google will find cheaper ones. This one is very nice though. Interfaces nicely to my PIC projects.

If you are running a 5-volt processor, a 2N7000/7002 is a nice super-cheap switch for moderate currents. Something like FDV301 is better if all you have is 3.3 volts to drive the gate.

There are bigger "logic level" fets, up into the 10's of amps, but at some point it often makes sense to use a regular power fet and a gate driver chip.

John

Hmm. I hadn't seen it before. Vgs(th) max of 1.2V and can cost under

20 cents in some modest qty. Still, I got pn2222 for well under one cent. Of course, other components are usually needed. So I'm keeping this one in mind and buying some soon to try out.

Thanks, Jon

l

lf

ut

rt

r g

With an Rds(on) of ~50mohm, it can handle a fairly hefty load, too.

With a base drive that many micros can handle, a pn2222 can do several hundred mA with Vce of a few tenths volt. I usually experience a beta of about 50-100 at useful saturation levels (1/4 volt Vce.) More like an ohm, though, I suppose. Cheap.

But yes, I like the idea where handling more current is required and this part looks very nice because it works with a variety of low voltage micros and is not over-priced.

Thanks, Jon

That was sorta my point. The '2502 will do something like 4A without glowing. I've never used it at anything like that though. I could easily use 2222s, but that's another part. ;-)

MOSFETS are not useful in low voltage applications; one can drive a BJT with 800mV but the minimum drive for a normal MOSFET is 3-4V for a "logic" FET which cannot handle high voltages that BJTs can. And if one wants decent turn-on and turn-off speed, MOSFETs need large gate currents during the transition. So, as long as your applications are not in those regions, have fun.

or

mall

yself

=A0But

part

nder

ping

When you have to switch low voltage loads in the Ampere range, you can use the "low VCEsat" transistors (NXP/philips, Zetex/diodes). These devices have high HFE (>200), very low VCEsat and can be driven from

1.8V logic.

Best regards,

Wim PA3DJS

please remove the obvious three characters in case of PM.

Nonsense!

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
 I love to cook with wine     Sometimes I even put it in the food

The above IRLML2502 has a Vgs of 1.2V. It doesn't like high voltages though.

Yes, they still sell BJTs.

I've not looked, but are any priced near the 20 cent area I found for the IRLML2502?

Jon

essor

g small

n myself

. =A0But

e a part

t under

one

keeping

ta

ke

Hello John,

There you might have a point. The voltage and current range is limited for the low VCEsat devices, for example: ZXT11N15DF (15V/3A) is about 0.17EUR, but this is about 0.23$. FMMT618 (20V/2.5A) is about 0.21$. Of course, large quantities.

Your mosfet is cheaper for supply voltage > 2.5V.

Best regards,

Wim PA3DJS

myself

 But

part

..Jon

That 20 cent figure I reported was taken for small qtys.

With a Vgs(th) of 1.2V, it may work below 2.5V, as well.

But nothing gets close to what I pay for pn2222s. It costs me almost nothing to get a lifetime supply of them. (Speaking as a hobbyist.) I go though hundreds of them, maybe a thousand, but not much more than that over periods of years. At 50 cents (just picking a number out of my hat), this may translate to 1000*.5 = $500. For one part, that is too much to spend. But if I invest my time in them, learning them, and using them... I will want to have perhaps many hundreds around to give away, play with, etc. At the well less than 1 cent I pay for pn2222s, $500 would buy me almost 100,000 pieces!! I could wade in them, hip deep. ;)

When I buy something, I buy a lot so I don't have to bother again. But I buy something generally useful, too. At 15-20 cents, I'm right at that point where I _may_ or _may not_ choose to use it as such a part.

100 would be $15-20, which is fine. 1000 would be getting into $200 territory, which is scary much for just one part to have around. If the particulars are good enough and the utility broad enough, it may be worth it. Much more, though, and its out of my territory.

I'm pretty much a 1 cent to 3 cents buyer. That is where I'm comfortable picking up things like this. That's one reason mosfets have been mostly off the table for me -- except where they represent a serious advantage. But then, I just buy a few for the immediate need and a little extra, just in case.

Thanks, Jon