From: Winfield Hill
Subject: Re: MOSFET "Normalized On-Resistance" question
Date: 17 Dec 2002 03:58:29 -0800
Organization: Rowland Institute
References: <email@example.com> <firstname.lastname@example.org> <email@example.com>
X-Newsreader: Direct Read News 4.20
John Muchow wrote...
> Cool...and even better, it makes sense. It looks like a Vgs
> of 5.5 or above lets me get a positive tempco (5.0V is *almost*
> there but not quite).
> It seems that you can also use these curves to select the range
> of Vgs that you want to have when operating linearly. Knowing
> that the IRL1404 (according to its data sheet) goes positive
> tempco (and stops acting linearly) at or above a Vgs of 4.3V
> seems to be a good fit for a op-amp output that swings up to
> that point with a 5V supply...like the LT1013 you recommended
> (seems like 4V min, 4.4V typical with the almost zero current
> of a MOSFET gate).
Of course you can't pick the operating voltage for a FET gate,
it is what it is, as they say. Extrapolating the 1404's fig 3
Typical Transfer Characteristics curves implies a typical part
will be about 4V for 20A 1V drain when cold, decreasing toward
3V when hot. FETs have a rather poor part-to-part tolerance as
well, plus you'd like to keep the ability to substitute FET parts
one day, so you'd be well advised to use a higher supply voltage
for the opamp, at least 6V, but 12V, etc. even better, allowing
you a maximal ability to turn on the FETs at high currents and
low drain voltages in some future application.
> Time for me to read Chapter 3 of the AofE again...
Good idea, especially the FET formulas!